■^3 1S> j» -^"31* ^ -^--^^ )V-»:>-> :>iap ^>>^^il:^> i^-.^' ^^;^»>. B:^)>> l>> ■^ ,y^J£>y> y' '3) .iJS^ ■)>)_) JJl*^;''^^*) > ^ ^l v^^^3 |»i^^^^-^X> :3fe^3»^ 5>>^:^ ? .' '^ ^4a«.^,i*^^^^'^'^^ naiflp? ';:sO^>> ^j>^jQM» Xmy 2^1 VU; JU f ibvarg 0f tlje Pxiseum OF COMPARATIVE ZOOLOGY, AT HARVARD COLLEGE, CAMBRIDGE, MASS. iFoutiDetr tj prfbate sufiscrfption, m 1861. The gift of ^kt ^oviAvtC/^v^C/^^ylr No. H^JO, jiocc^i^''^^- TRANSACTIONS CONMCTICUT ACADEMY ARTS AND SCIENCES. VOLUME II. NEW HAVEN : PUBLISHED BY THE ACADEMY PRINTED Br TUTTLE, MOREHOUSE & TAYLOR. ^1871 TO 1873. > CONTENTS OF VOL. II, PART 1 Page List of Additions to the Library, i Art. L — Notice op the Crustacea collected by Prof. C. F. Hartt on the coast of Brazil in 1867. By Sidney I. Smith, 1 List of the described species of Brazilian Podoph- thalmia, 31 Art. II. — On the Geology of the New Haven Region, WITH special reference TO THE OrIGIN OF SOME OF ITS Topographical Features. By James D. Dana, . 45 Art. III. — Notes on American Crustacea, No. I, Ocypod- oidea. By Sidney I. Smith, 113 Art IV — On some alleged specimens of Indian Onomato- pceia. By J. Hammond Trumbull,.. 177 Art V — On the Molluscan Fauna of the later Tertiary OF Peru. By E. T. Nelson, 186 CONTENTS OF VOL. II, PART 2 List of Additions to the Library, v Art. VI. — On the Direction and Force of the Wind, with the Fall of Rain and Snow, at Wallingford, Con- necticut, FROM Observations made by Benjamin F. Harrison, M.D., and reduced by Francis E. Loomis, Ph.D., 209 Art. VII. — Design for a Bridge across the East River, New York, at Blackwell's Island. By William P. Trowbridge, 263 Art. VIII. — On the Mean Direction and Force of the Wind at New Haven, Conn. ; from an extended se- ries OF Observations reduced by Francis E. Loomis, Ph.D., 269 IV CONTENTS. Pagb Art. IX. — Notes on the Geology op the Island of Yesso, Japan, from Observations made in 1862. By W. P, Blake, 293 Art. X. — Comparison of the Muscles of the Chelonian and Human Shoulder-girdles. By Henry S. Williams, 301 Art. XI. — Graphical Methods in the Thermodynamics of Fluids. By J. Willard Gibbs, 309 Art. XII. — List of Marine Alg^ Collected near East- port, Maine, in August and September, 1873, in con- nection with the work op the U. S. Fish Commission UNDER Prop. S. F. Baird. By Daniel C. Eaton, 343 Art. XIII. — The Early Stages of the American Lobster {Hohiarus Americanus Edwards). By Sidney I. Smith, 351 Art. XIV. — A Method op Geometrical Representation OP the Thermodynamic Properties op Substances by MEANS of Surfaces. By J. Willard Gibbs, 382 TRANSACTIONS COMECTICUT ACADEMY ARTS AND SCIENCES VOL. II, PART 1. NEW HAVEN: PUBLISHED BY THE ACADEMY PKINTED BY TUTTLE, MOREHOUSE & TAYLOR. 1870. CONTENTS. Page. List of Additions to the Library, i Art. L — Notice of the Crustacea c;ollected by Prof. C. F. Hartt on the coast of Brazil in 1867. By Sidney L Smith, 1 List of the described species of Brazilian Podoph- thalmia, 31 Art. IL — On the Geology of the New Haven Region, with special reference to the Origin of some of its Topographical Features. By James D. Dana 45 Art. IIL — Notes on American Crustacea, No. I, Ocypod- oidea. By Sidney L Smith, 113 Art. IV. — On some alleged specimens of Indian Onomato- poeia. By J. Hammond Trumbull, 177 Art. V. — On thic Molluscan Fauna of the later Tertiary OF Peru. By E. T. Nelson, 186 OFFICERS OF THE ACADEMY. President, CHESTER S. LYMAN. Vice-President, ELIAS LOOMIS. Secretary, WILLIAM H. BREWER. Librarian, ADDISON VAN NAME. Ireasurer, HENRY C. KINGSLEY. PiiMish i:i(/ ( ^(rmmittee, HUBERT A. NEWTON, GEORGE J. BllUSH, ELIAS LOOMIS, CHESTEll S. LYMAN, JAMES HADLEY, DANIEL C. OILMAN, x\DDISON E. VERRILL. Auditing Committee, ELIAS LOOMIS, HENRY T. BLAKE, WILLIAM D. WHITNEY. ADDITIOI^S TO THE LIBRARY, From January 1, 1867, to August 1, 1870. AMERIOAX. Albany Institute. Transactions, vol. v, 1867. 8°. Manual, March, 1870. 8°, pp. 48. Dudley Observatory. Annals, vol. i, 1866. 8°. American Association for the Advancement of Science. Proceedings at 15th, 16th, and 17th Meetings, 1866-8. Cambridge, 1867-70. 8°. Boston Society of Natural History. Memoirs, vol. i, parts 1, 2, and 4, 1866-9. 4°. Proceedings, vol. xi (wanting sig. 2), xii, xiii sigs. 1-17, 1866-70. 8°. Occasional Papers, vol. i, 1869, 8°. Annual Reports, May, 1869. 8°, pp. 76. Chicago Academy of Science. Transactions, vol. i, 1867-9. 8°. Cincinnati Observatory. Adams, J. Q., Oration before the Cincinnati Astronomical Society, Nov. 10, 1843. 8°, pp. 72. Mansfield, E. D., Annual Address. June, 1845, with Reports. 8°, pp. 55. Reports of the Director, June, 1868, May, 1869. 8°, pp. 48. Montreal. — The Canadian Naturalist, with the Proceedings of the Natural History Society of Montreal. New Series, vol. iii. 1866-8. 8°. Philadelphia. — Historical Society of Pennsylvania. Memoir.-^, vol. i (reprinted), iii, part 2, IV, part 1, v-vn. 1836-60. 8°. G-arrard, L. H., Chambersburg in the Colony and the Revolution. 1856. 8°. Coles, E., History of the Ordinance of 1787. 1856. 8°, pp. 33. Mayer. B., Calvert and Penn. 1852. 8°, pp. 50. Ingersoll, J. R., Memoir of Samuel Breck. 1863. 8°, pp. 40. Salem. — Essex Institute. Proceedings, vol. v, parts 2-4, 7, 8, vi, part 1. 1866-8. 8°. BuUetin, vol. r, 1869, ii, Nos. 1-3, 5. 8°. Peabody Academy of Science. Memoirs, vol. i, No. 1. 1869. 8°. Savannah. — Geo gia Historical Society. Collections, vols, ii, in, part 1. 1842-8. 8°. Stevens, W. B., History of Georgia. New York and Phil, 1847-59. 2 vols., 8°. Toronto. — Magnetical Observatory. Monthly Meteorological Register, 1869. 4°, pp. 12. General Meteorological Register, 1868, 1869. 8°, pp. 7, 6. Meteoro- logical Summary, Nov., 1869. p. 1. Monthly Value of Magnetic Elements, 1 865-8. 8°, pp. 2. On changes of barometric pressure, by G. T. Kings- ton. 8°, pp. 5. ii Additions to the Lihrary. EUROPEAiSr. Altenburg. — Mittheilungen aiis dem Osterlande, von den Gesellschaften zu Altenbiirg, Bd. XVIII, Heft 1-2. 1867, 8°. Bologna. — Accademia delle Scienze dell' Instituto di Bologna. Rendicdnti, 1865--6, 1866-7, 1867-8. 8°. Galvani, L., Opere edite ed inedite, Bologna, 1841; Aggiunta, 1842. 4°. Brussels. — Academie Royale des Sciences, des Lettres et des Beaux-Arts de Belgique. Memoires, Tome xxxvn. 1869. 4°. Memoires Couronnes et Memoires des Savants Etrangers, Tome xxxiii. 1865-7. 4°. Memoires Couronnes et aiitres Memoires, Tomes XIX, XX. 1867-8. 8°. Bulletins, 2 e Serie, Tomes xv-xxi, xxiii-xxvL 1863-8,8°. Annuaire, 1864-9. 12°. Observatoire Royal. Annales, Tome xvin. 1868. 4°. Annales Meteorolo- giques, lere Annee. 1870. 4°. Observations des Phenom6nes Period- iques, 1861-2, 1864, 1865-6. 4°. Chemnitz. — Naturwissenschaftliche Gesellschaft. ErriterBericht, 1859-64. 1865. 8°, pp.30. ZweiterBericht, 1864-8. 1868. 8°, pp. 55. Cherbourg. — Societe Imperi.qle des Sciences Naturelles. Memoires, Tomes xviii, xix. Paris, 1868-9. 8°. Christiania.— Videnskabs-Selskab. Forhandlinger, 1867. Registre, 1858-67. 8°. Observatorium. Meteorologiske lagttagelser, 1867. 4°. Norske Meteorologiske Institut. Aarbog, 1867. 4°. Kongelige Norske Universitet. Aarsberetning. 1867. 8". Index Schola- rum, Feb., Aug., 1868. 4°. Sars, M., Memoires pour servir a la connaissance des Crinoides vivants. 1868. 4°. Broch, 0. J., Traite elementaire des Fonctions elliptiqnes, second fasc. 1867. 8°. From the University of Christiana. Copenhagen,— Kongelige Danske Videnskalicrnes Selskab. Oversigt over Forhand- linger, 1866-8, 1869, Nos. 1, 2, 8°. Danzig.— Naturforschende Gesellschaft. Schriften. Neue Folge, Bd. i, ii. Dresden. — Naturwissenschaftliche Gesellschaft Isis. Sitzungs-Berichte. Jahrg. 1868, No. 4-6. 8°. Edinburgh Geological Society. Transactions, vol. i, 1868-70. 8°. Harlem.— Miisee Teyler. Archives. Vol. i, ii, Fasc. 1, 2, 1868-69. 8°. Catalogue Systematique de la Collection P.^]eontologique, par T. C. Winkler. Livr. 1 6, 1863-7. 8°. Helsingfors.— Societas Scientiarum Fennica. Acta, Tome viii. 1867, 4°. Ofver- sigt af Forhandlingar, Bd. ix, x, xi. 1867-9. 8°. Hjelt, 0. E. A., Gediichtnissrede auf A. v. Nordmann. 1868. 8°, pp. 61. Krakau. — K. K. Sternwarte. Materialy do Klimatografii Galicyi. Rok 1867, 1868. 8°. Karlinski, F., Mittlere Temperatur zu Krakau 1826-1865. Wien, 1868. 4°. Leeds.— Philosophical Society. Annual Report, 1867-8. 8", pp. 32. Leyden. — Sternwarte. Annalen. Bd. i. Harlem, 1868. 4°. London. — Museum of Practical Geology. Portlock, J. E., Report on the Geology of the County of Londonderry, and of parts of Tyrone and Fermanagh. Dub- lin, 1843. 8°. Additions to the Librari/. iii Madrid. — Real Observatorio. Observaciones Meteorolo.^icas, 1865-6, 1866-'?. 8°. Observaeioues Meleovologicas effectuadas en la Peuiusula, 1865-6. 8°. Informe del Director. 1867. 8*^. Manchester. — Literary and Philosophical Society. Memoirs, Third Series, vol. in. London, 1868. 8°. Proceedings, vols, v-vii, 1866-8. 8". Mannhedi. — Yerein f iir Naturkunde. Fiinfunddreissigster Jaiiresbericht. 1869. 8°, pp. 75 Milan. — Reale Institute Lombardo. Rendiconti. Serie II, vol. i, ii, fasc. 1-1 2. 8°. Amati, A., Dell' Australia e della fondazione d' una colonia con bandiera Italiana. Milano, 1868. 8°, pp. 50. Reale Osservatorio di Brera. Effemeridi Astronomiche, 1868, 1869. S°. Schiaparelli, G. V., and Celoria, Gr., Sulle variaziooi period, del barometro uel clima di Milano. 4°, pp. 31. MoNCALiERi. — Osservatorio del R. Collegio Carlo Alberto. Bulletiuo Meteorologico, Vol. I, II, m Nos. 1-9, 11-12. Deuza, F., Le Slelle Cadenti dei periodi di Agosto e Novembre osservate in Piemonte, 1866-8, Memorie, i-iv. Torino, 1866-8. 8°. Sopra gli AeroUti caduti, 29 Feb., 1868. Torino, 1868. 8°. Moscow. — Societe Imperiale des Naturalistes. Bulletin, 1867, No. 4. 1868, Nos. 1-4. 8°, Munich. — Konigliche bayerische Akademie. Sitzungsbericlite. 1866-1869, n, Ileft 3 (wanting 1866, i. Heft 3). 8°. Almanach, 1867. 16°. Sternwarte. Annalen. Bde. xrv-xvii, 1865-9. 8°. Supplementbde. v-ix, 1866-9. 8°. BeitrJige zur Geschichte der westlichen Araber, herausg. von M. J. Miiller, Heft I. Miinchen, 1866. 8°. Bauemfeind, C. M., Die Bedeutung nioderner Gradniessungen. Miinchen, 1866. 4°, pp. 41. Bischoff, Th. L. W., Resultate des Recrutirungs-Geschiiftes. Miinchen, 1867. 8°, pp. 65. Giesebrecht, W. v., Ueber einige iiltere Darstellungen der deutschen Kaiser- zeit. Miinchen, 1867. 4°, pp. 20. Vogel, A., Denkrede auf H. A. von Vogel. Miinchen, 186S. 8°, pp. 72 " " Ueber die Entwickelung der Agriculturchemie. Miinchen, 1869. 4°, pp. 49. Voit, C, Ueber die Tlieorien der Ernahrung der thierischeu Organismen. Miinchen, 1868. 4°, pp. 37. Meissner, C. F., Denkschrift auf C. F. P. von Martins. Miinchen, 1869. 4°, pp. 28. Sclilagintweit, E., Die Gottesurtheile der Indier. Miinchen, 1866. 4°, pp. 36. Brunn, H., Ueber die Sogenannte Leukothea in der Glyptoihek. Miinchen, 1867. 4°, pp. 25. I^vm the Royal Academy. Nuremberg. — Naturhistorische Gesellschaft. Abhandlungen, Bd. iv. 1868. 8°. Paris. — Observatoire Meteorologique de Montsouris. Bulletin, ]er Juillet, 1869 — 27 AvrO, 1870. 4°. Societe d' Ethnograpbie. Expose Generale, 1869. 8°, pp. 24. Prague. — Konigliche bohmische Gesellschaft der Wissenschaften. Abhandlungen. Fiinfte Folge, Bd. xrv. Sechste Folge, Bd. i. 1866-8. 4°. Sitzungs- berichte, 1865-7. 8°. iv Additions to the Library. Riga. — Naturforschender Verein. Arbeiten. Neue Folge, Heft 2. 1868. 8°. St. G-allen. — Naturforschende Gesellschaft. Bericht, ISeS-T, 1867-8. 8°. St. Petersburg. — Academie Iinperiale des Sciences. Catalogue des livres publics en langues etrangeres par FAcademie. 1867. 8°. SCHWEIZERISCHE Naturforschende Gesellschaft. Verhandlungen, 51, 52, Jahresver- sammlungen. 1867, 1868. 8°. Stockholm. — Kongliga Svenska Vetenskaps-Akadeniien. Handlingar, Ny Poljd, Bd. VII, i, 1867. 4°. Ofversigt af Forhandlingar. Bd. xxv, 1868. 8°. Me- teorologiska lagttagelser, Bd. viii. 1866. 4°. Lefnadsteckningar. Bd. i, i, 1869. 8°. Sundevall, C. J., Die Thierarten des Ari,stoteles, Stockholm, 1863. 8°. " " Conspectus Avium picinarum, Stockliohii, 18(>6. 8°. Nordenskiold, A. E., Sketch of the geology of Spitzbergen, Stockholm, 1867. 8°, pp. 55. Igelstrom, L. I., Rock of Nullaberg. 8°, pp. 11. Stuttgart. — Verein fiir vaterlandische Naturkundo in Wiirttemberg. Jahreshefte, Jahrg. xxiv, Heft. 3. 1868. 8°. Upsala. — Regia Societas Scientiarum. Nova Acta. Ser. in. Vol. vi. 1868. 4°. Vienna. — K. K. Geologische Reichsanstalt. Jalirbuch, 1867, Nos. 1-4, 1868, Nos. 1, 3, 4, 1869, Nos. 1-3. 8°. Verhandlungen, 1867, 1868, Nos. 1-6, 11-18, 1869, Nos. 6-13. 8°. Zurich. — Naturforschende Gesellschaft. Vierteljahrsschrift, 1867, 1868. 8°. Allen, J. A., Mammalia of Massachusetts. [Bulletin of the Museum of Comp. Zool., Cambridge, Mass., No. 8.]. From the Author. Barrande, J., Cephalopodes Siluriens de la Boheme, Introduction, Prague, 1867. 8°. pp. 48. From the Author. d'Elvert, C. Zur Geschichte der Pflege der Naturwissenchaften in Miihren und Schlesien, Briinn, 1868. 8°. From the Author. Galle, J. G , Ueber die Bahn des Pultusk Meteors. (From the Abh. d. Schles. Gesell.) Breslau, 1868. 8°. pp. 43. From the Author. Gore, G., On Hydrofluoric Acid. (From the Phil. Trans. 1869.) 4° pp. 27. From the Author. Quetelet, A., Physique Sociale, Tome i. Brux. 1869. 8°. Anuales Meteorol. de I'Observatoire de BruxeDes. 2e Annee, 1868. 4°. Notices [extraits des Bulletins de I'Academie Royale]. 9 pamphlets 8°. From M. A. Quetelet. Quetelet, E , Sur I'etat de I'atmosphere a Bruxelles, 1865. Brux., 1866. 8°. pp. 48. Memoire sur le Temperatur de Bruxelles. Brux., 1867. 4°. From M. E. Quetelet. Notice of the Crustacea collected by Prof. C. F. Hartt ON the coast of Brazil in 1867. By Sid^^ey I. Smith. Read, May 19th, 1869. Ix the first volume of these Transactions, Prof. Verrill has noticed the Radiata of the collection made by Prof Hartt upon the coast of Brazil during the summer of 1867, and the Crustacea of the same col- lection, having been submitted to me for examination, was found to contain so many species new to the Brazilian fauna that the publica- tion of the following list seemed desirable. The collection, although quite small in number of specimens and representing only the higher groups of the class, is interesting from the large proportion which it contains of species heretofore known only from the West Indies or Flordia. This is, perhaps, due chiefly to the fact that most of the collections brought from Brazil have been made at Rio de Janeiro where there are no coral reefs, while Prof Hartt's collection was made principally on the rocky and reef-bearing parts of the coast. BRACHYURA. Milnia bicornuta stimpson. Pisa bicornuta LatreiUe, Encyclopedie methodique, tome x, p. 141 {teste Edwards). Pericera bicorna Edwards, Histoire naturelle des Crustaces, tome i, p. 337, 1834. Pisa bicorna Gibbes, On the Carcinological Collections of the United States, Pro- ceedings American Association, 3d Meeting, p. 110, 1850. Pericera bicornis Saussure, Crustaces nouveaux des Antilles et du Mexique, p. 1 2, pi. 1, fig. 3, 1858. Milnia bicornuta Stimpson, Notes on North American Crustacea, Annals Lyceum Nat. Hist, New York, vol. vii, p. 1 80, 1 860. A single specimen collected at the Reefs of the Abrolhos does not differ from Bermuda, Florida and Aspinwall specimens. Mithraculus coronatus Stimpson. Cancer coronatus Herbst, Naturgeschichte der Krabben und Krebse, Band i, p. 184, Tab. 11, fig. 63, 1782, and Cancer Coryphe, Band iii, zweytes Heft, p. 8, 1801. Mithraculus coronatus (pars) White?, List of Crust, in the British Museum, p. 7, 1847. Teans. Connecticut Acad., Vol. XL 1 July, 1869. 2 S. I. Smith on Brazilian Crustacea. Mithraculm coronatm Stimpson, American Journal Sou, 2d series, vol. xxix, 1 860, p. 132 ; Annals Lye. Nat. Hist, New York, vol. vii, p. 186, 1860. Two females of this species were collected by Prof. Hartt at the Keefs of the Abrolhos. They do not differ perceptibly from A spin- wall specimens. The two specimens give the following measurement : — Length of carapax, 12-8"™ Breadth of carapax, 17-6™"! Ratio, 1 : 1-37 " " " 17-2 " '• " 23-4 " 1:1 -36 The differences pointed out by Stimpson at once distinguish this species from M. sculptus, but White cites the figures of both species under his Ifitkraoulus coronatus, so that it is not possible, without an examination of his specimens, to tell which species he had in view. Mithrax hispidus Edwards. Cancer hispidus Herbst, op. cit.. Band i, p. 247, Tab. 18, fig. 100, 1782. Mithrax hispidus Edwards, Magasin de Zoologie, 2^ annee, 1832 ; Historie natureUe des Crust., tome i, p. 322, 1834; DeKay, Zoology of New York, Crust., p. 4, 1844; Gibbes, loc. cit., p. 172; Stimpson, American Journal Sci., 2d series, vol. xxix, 1860, p. 132; Annals Lye. Nat. Hist., New York, vol. vii, p. 189, 1860. Several specimens collected at the Reefs of the Abrolhos agree well with Edwards' and Stimpson's descriptions of this species. The cara- pax is wholly naked above, the elevations anteriorly are smooth and polished, and there are no spines or prominent tubei'cles on the median regions. There are two small tubercles just at the base of the frontal teeth, and two more just behind these on the anterior lobes of the gastric region ; there are also traces of two tubercles on each of the antero-lateral gastric lobes, and several small tuberculiform elevations on the hepatic and branchial regions near the antero-lateral margin. The external angle of the orbit forms an obtuse tooth not projecting so far forward as the external lobe of the inferior margin ; the suc- ceeding tooth of the antero-lateral margin (the second normal) is quite small and obtuse, but the three remaining teetli are spiniform, slender and curved forward ; in addition, there is a very small tooth just be- hind the posterior spine of the antero-lateral margin. Several specimens give the following measurements: — Sex. Length of carapax. Breadth of carapax including spines. Ratio. Male. 15-5mm 18-omm 1: 1-16 " 18-9 22-7 1 : 1-20 Female. 13-4 15-4 1 : 115 u 15-4 18-0 1: 1-17 ^^>. I. Smith on Brazilian Crustacea. 3 Xantho denticulata White. X'lntho denticulata Whits, liist oi Crnat. in the British Museum, p. 17 (no descrip. tion), 1847 ; Annals and Mag. Nat. Hist., 2d series, vol. ii, p. 285 {X. denticulafus), 1848 {nnn Stimpson); Smith, Proc. Boston Soc. Nat. Hist., vol. xii, p. 274, 1869. A single specimen collected at the Reefs of the AbroUios does not differ from specimens from Bermuda and Aspinwall. It seemes to be an uncommon species as it is not mentioned by Dana, Gibbes, or Stimpson, and I have only seen a single one from each of the localities mentioned, Chlorodius Floridanns Gibbes. Chlorodius Floridanus Gihhes, \oc. cit., p. 175, 1850; Stimpson, Annals Lye. Nat. Hist., New York, vol. vii, p. 209. Several specimens, not differing perceptibly from those from Florida and x\spinwall, were collected at the Reefs of the Abrolhos. Three specimens give the following measurements : — Sex. Length of carapax. Breadth of carapax. Ratio. Male. 20-8mm 33.8mm 1 : 1-62 Female. 15-6 23-8 1 : 1-53 11 18-4 29-4 1 : 1-60 PanopeuS politUS Smith, loc. cit., p. 282, 1869. Plate I, figure 4. This species is allied to P. transversus Stimpson, and resembles somewhat the crenatus of Edwards and Lucas. The carapax is entirely naked above, broad, moderately convex in two directions, slightly granulous and uneven on the front and along the antero-lateral border, but smooth and highly polished on the median regions and posteriorly. The regions are slightly but distinctly indi. cated. The gastric region is surrounded by a well marked sulcus, but its lobes are not distinctly indicated except the anterior extremity of the median, which is slender and acutely pointed ; the frontal lobes are indicated by slight prominences. The hepatic region is not divided, but there are one or two slight plications on its anterior part parallel to the antero-lateral margin. The cervical suture is distinct in its outer portion but is not indicated near the gastric region. The median and posterior lobes of the bi-anchial region are separated by a distinct depression. The front is strongly deflexed, the edge somewhat bev- eled from above and four-lobed ; the median lobes are very broad, project prominently and are separated by a sharp notch ; the lateral lobes project as small narrow teeth. The antero-lateral margin is di- 4 S. I. Smith on Urazilicm Crustacea. vided by small notches into four lobes, the first of which is composed of the angle of the orbit coalesced with the second normal tooth ; the first lobe is broad, its edge slightly concave and projecting a little at the angle of the orbit ; the second and third lobes are broad and trun- cate ; the fourth lobe is small and obtuse and forms the lateral angle of the carapax. From each of the notches slight sulci extend a little way back upon the carapax. Beneath, the edge of the front is thin, projects obliquely downward and is not expanded in front of the antenuulre. The epistome is smooth, and its labial border has a prominent median lobe and a slight incision each side. The external maxillipeds are smooth ; the merus is quadri- lateral, its outer edge not projecting, and the antero-exterior angle rounded. The inferior margin of the orbit is divided into two lobes by a broad and shallow sinus ; the inner lobe forming a prominent tooth which projects as far forward as the lateral lobe of the front, and the outer lobe broad and slightly prominent. The external hiatus of the orbit is rather broad and shallow. The sub-orbital and sub-hepa- tic regions are quite granulous. The tubercle beneath the anterior lobe of the antero-lateral margin is depressed, forming only a slight granu- lous prominence. The sub-branchial region is somewhat hairy. The female abdomen is broadly ovate, the greatest breadth being at the fourth segment. The chelipeds are slightly unequal, the carpi and hands smooth and evenly rounded above and on the outside. The hands are stout, the fingers obscurely marked with longitudinal impressed lines, and irreg- ularly toothed within, and in the dactylus of the larger hand there is a prominent cylindrical tooth at the base. The ambulatory legs are smootli and nearly viaked except a close pubescence upon the dactyli, penultimate segments, and slightly on the carpi. In an alcoholic specimen the color is light bi-own above, tinged with bluish purple on the anterior part of the carapax and the upper side of the chelipeds. The fingers are black, lighter at the tips, and the black not spreading iipon the palm. Length of carapax in the single female specimen, 13 '8™™ ; breadth, 21*4 : ratio of length to breadth, 1 : 1*55. Collected at the Reefs of the Abrolhos. The P. transversus Stimpson (Annals Lye. Nat. Hist., New York, vol. vii, p. 210, 1860) of the west coast of Central America, differs from this species in having the carapax much less distinctly areolated, more regularly oval in outline and smoother and more evenly convex above. The front also projects much less prominently ; the antero lat- S. I. Sinith oil Brazilian Crustacea. 5 eral margin is smooth and even and tlie lobes separated by very slight incisions, and the edge of the first lobe is slightly convex and does not project at the angle of the orbit ; there is no noticeable depression be- tween the median and posterior lobes of the branchial region ; the in- ferior margin of the orbit is divided by a very slight sinus, and the inner lobe is not at all prominent ; and finally, the external maxilli- peds are slightly granulated. The color of alcoholic specimens is quite different, being dark slate-brown on the upper side of the carapax and chelipeds. The P. crenatus of Edwards and Lvicas is a much smoother species than the politKS, the regions being scarcely at all defined and the car- apax almost perfectly smooth along the front and antero-lateral bor- der. The front is not deflexed, its edge is nearly straight, and beneath it is expanded horizontally in front of the antennulse ; the sub-orbital and sub-hepatic regions are quite smooth, and there is no tubercle be- neath the first lobe of the antero-lateral margin ; and finally, the an- tero-exterior angle of the merus of the external maxillipeds projects laterally somewhat beyond the lateral margin and is broadly rounded.* Panopeus Harttii Smitii, loc. dt., p. 280, i869. Plate 1, figure 5. The carapax is clothed with scattered hairs along the borders, is broadest at the penultimate teeth of the antero-lateral margins, con vex anteriorly but flattened behind, and coarsely granulous on the front and along the lateral borders, but nearly smooth on the median and posterior regions. The gastric region is surrounded by a very deep sulcus, which is particularly marked posteriorly next the cardiac and the posterior part of the branchial region ; its median lobe is sep- arated from the antero-lateral lobes by a slight but distinct sulcus ; and the anterior lobes are prominent and marked anteriorly by a sharp plication. The hepatic region is pi-ominent, somewhat projecting and bears a transverse, granulous ridge. The cervical suture is very marked and extends as a broad depression quite to the gastric region. The median and posterior lobes of the branchial region are separated by a slight depression. The front is very much deflexed and the edge * The figure of the facial region of this species given in the Voyage dans VAmerique Meridionale (pi. 8, fig. 1 a) improperly represents the external maxillipeds with this an- gle truncate and not at all produced laterally. 6 aS'. I. Smith on Brazilian Crustacea. thin and four lobed ; the median lobes are very mucli the largest, are evenly rounded, and a little more prominent than the lateral, which project as small obtusely triangular teeth. The superior margin of the orbit is broken by two incisions leaving the margin between them projecting as a slight, rounded lobe. The post-orbital tooth is short and slender, and is separated from the second tootli of the antero- lateral margin by a broad sinus whicli breaks the margin completely. The remaining teeth of the antero-lateral margin are triangular in form, much thickened vertically, and separated by quite broad sinuses, and the posterior two on each side are very slender and of nearly equal prominence. Beneath, tlie edge of the front is thin and projects sharply down- ward. The epistome is smooth and its labial border has a small lobe in the middle, a slight notch each side and another at each angle of the buccal area. The external maxillipeds are smooth except the me- rus, which is slightly granulated and also has the antero-exterior angle very slightly produced laterally and not at all rounded. The inferior margin of the orbit is prominent and divided into two lobes by a deep and narrow sinus ; the inner lobe forms a stout tooth which projects as far forward as the inner angle of the superior margin ; the outer lobe is broad and its exterior angle projects slightly in advance of the post-orbital tooth. The external hiatus of the orbit is a deep trian- gular notch. In one specimen, however, it is wholly closed on one side, possibly from some accident. The 'sub-orbital and sub-hejjatic regions are quite coarsely granulous. The tubercle of the sub-hepatic region forms a slight granulous prominence just beneath the post-or- bital tooth. The sub-branchial region is pubescent and slightly gran- ulous. In the male, the sternum is smooth and the abdomen quite narrow, being narrowest at the penultimate segment, and the terminal segment is about five-sixths as long as broad, and its extremity evenly rounded. In the female the abdomen is broadly ovate, the greatest breadth being at the fourth segment. The chelipeds are a little unequal. The carpi are granular-rugose externally and have a deej? groove along the outer margin next the articulation with the hand. The hands are slightly rugose above, and the fingers are slender, deflexed, marked with slight, impressed longi- tudinal lines and slightly and obtusely toothed mthin, and the dacty- lus in the larger hand usually has a stout tooth at the base. The ambulatory legs are slender, and pubescent along the edges of all the segments and over the whole surface of the dactyli. *S'. 1. SiiiitJi on Brazilian Crustacea. *J Alcoholic specimens are light olive brown above and on the chelipeds. The fingers are black, lighter at the tips, and the black not spreading upon the palm. Several specimens give the following measurements : Sox. Length of carapax. Breadth of carapax. Ratio. Male. 15-Omni 22-5mm 1 : I'oO " 15-9 23-6 1 : 1-49 Female. 9-6 14-4 1 : 1-50 " 12-6 18-8 1: 1-49 Seven specimens were collected by Prof. Hartt at the Reefs of the Abrolhos, This species is very distinct from all other described species of the gemis. Its broad and deeply areolated carapax give it somewhat the aspect of a Chlorodms. Eriphia gonagra Edwards. Cancer gonagra'Eabvxcms, Supplemeutum Entomologiae systematicse, p. 337, 1798. iVj^/i/a grona^'ra Edwards, Histoire naturelle des Crust., tome i, p. 426, pi. 16, fig. 16, 17, 1834; Annales des Sciences naturelles, 3™e serie, tome xvi, 1851, pi. 8, fig. 10 ; White, List of Crust, in the British Museum, p. 22 ; Gibbes, loc. cit, p. 177 ; Dana, United States Exploring Expedition, Crust., p. 250; Stimpson, Annals Lye. Nat. Hist., New York, vol, vii, p. 217 ; Heller, Reise der osterreichischen Fregatte Novara um die Erde, p. 24, 1865. A large number of specimens are in the collection, all of them ob- tained at the Reefs of the Abrolhos. It seems to be a common species from southern Florida to Rio de Janeiro. A number of specimens give the following measurements : Sex. lale. Length ot carapax. Breadth of carapax Including spines. 24-8mm Katlo. 1 : 1-44 (1 24-0 34-5 1 : 1-44 i( 25-G 36-8 1 : 1-44 (( 26-8 37-8 1 : 1-41 (I 30-8 43-5 1 : 1-41 Y'male. 17-6 25-7 1 : 1-46 (1 19-6 28-2 1 : 1-44 " 23-0 33-2 1 : 1-44 11 28-2 41-3 1 : 1-46 Callinectes Danae Smith. Lupa diacantha Dana, United States Exploring Expedition, Crust., p. 272, pL 16, fig. 7, 1852. Callinectes diacanthus Ordway, Monograph of the genus Callinectes, Boston Journal Nat. Hist., vol. vii, p. 575, 1863. {Non Portunus diacanthus Latreille, nee Lupa diacantha Edwards, nee Callinectes diacanthus Stimpson.) Breadth of carapax including lateral spine. Ratio. 93.0mm 1 : 2-22 97-4 1 : 2-20 106-5 1 : 2-26 91-0 1 : 2-\n 94-8 ] : 2-12 T6-0 1 : 2-21 8 S. I. Smith on Brazilian Crustacea. A number of specimens which agree perfectly with the description of this species given by Ordway, were collected at Pernambuco by Prof. Hartt. A single female from Bahia does not differ from the Pernambuco specimens except in having the sub-median tooth of the front very short, scarcely projecting beyond the median teeth — probably an acci- dental character. Several specimens give the following measurements : — l^ength of carapax Sex. including sub-frontal spine. Pernambuco. Male. 41 -S"!™ " " 44-3 " 47-2 " Female. 41-8 44-8 Bahia. " 34-4 This species was known to Ordway only from Dana's original spe- cimen collected at Rio de Janeiro. Callinectes Ornatns Ordway, loc. cit., p. 571, 1863. A male specimen collected at Caravellas agrees perfectly Avith Ord- way's description and with a specimen from Bermuda. Length of carapax includmg sub-frontal spine, 36 "2™™; breadth of carapax including lateral spines, 80-5™'° ; i-atio of length to breadth, 1 : 2-22. A sterile female collected at the same locality may belong to this species. It differs from the male in being thicker and more convex, the areolation more strongly marked, and the granulations coarser ; the teeth of the antero-lateral border are less prominent and more ob- tuse ; and the chelipeds are quite short, the merus not reaching, by considerable, the tip of the lateral spine. Length of carapax, 34*6™"^ ; breadth of carapax, 75*0 ; ratio 1:214. In the deeply areolated carapax it approaches the larvatus, and it may possibly belong to that species. The description and figure of NeptxDius marginatKS A. Edwards* agrees very closely with this specimen, the figure of the abdomen and sternum representing it perfectly, and there can be little doubt that Edwards' species was based on a sterile female of some species of Callinectes. If the habitat. Cote du Gabon., given by Edwards be correct, it is safely inferred that the genus Callinectes is not confined to the American coasts. * Archives du Museum d'Histoire naturelle, tome x, p. 318, pi. 30, fig. 2, 1861. S. I. /Smith on Brazidan Crustacea. 9 The C. onialus was previously known from South Carolina, Tortu- gas, Ilayti, and Cumana. Callinectes larvatus Ordway, loc. cit., p. 573, i863. One specimen of this species, a male, was collected at Bahia. It is very much like tlie DancB and the oruatns in the carapax, etc., but diifers remarkably in the male abdominal appendages of the first pair (intromittcnt organs), which are very short, directed inward till they cross and then the extremities curved abruptly outward. Length of carapax including sub-frontal spine, 38-8™™ ; breadth in- cluding lateral spines, 8-2-4""" ; ratio of length to breadth, 1 : 2-11. Ordway's specimens were from Florida, Bahama, and Ilayti. Achelous spinimanus DeHaan. Portunus spinimanus Latreille, Encyc, t. x, p. 188 {teste Edwards). Lupa spinimana Leach, Desmarest, Considerations generales sur la classe des Crust., p. 98, 1825; Edwards, Histoire naturelle des Crust., tome i, p. 452, 1834; Gibbes, loc. cit., p. 178; Dana, United States Explorinpf Eexpedition, Crust., p. 273 ; Stimp- son. Annals Lye. Nat. Hist., New York, vol. vii, p. 57. Achelous spinimanus, DeHaan, Fauna Japonica, p. 8, 1833; White, List of Crust, in the British Museum, p. 28, 1847; Stimpson, Annals Lye. Nat. Hist, New York, vol. vii, p. 221, 1860; A. Edwards, Archives du Museum d'Histoire natureUe, tome X, p. 341, pi. 32, fig. 1, 1861 ; Heller, op. cit, p. 27. Three specimens, all females, collected at Bahia, give the following measurements : — Length of carapax Breadth of carapax Ratio of including frontal teeth. including lateral spines. length to breadth. 37-Omm 61-5""n 1 : 1 -66 44-4 77-4 1 : 1-74 56-0 95-0 1 : 1-70 All the S]Decimens have the lateral spine of the carapax nearly or quite twice as long as the one next in front of it. They appear to difter m no way from specimens from Florida. Achelous Ordwayi Stimpson. Achelous Ordwayi Stimpson, Annals Lye. Nat Hist, New York, vol. vii, p. 242, 1860. Xepfuntcs Ordwayi A. Edwards, op. cit, addenda, 186i. A male specimen of this fine species was collected, with the last, at Bahia. The carapax is narrower than in A. spinimanus, and the front more advanced. In areolation it resembles the spinimanus very much, the elevations however are not quite so thickly granulated. The teeth of the 10 S. I. Smith on Brazilian Crustacea. front are very long and slender, the length of the median ones exceed- ing slightly the distance between their tips. The teeth of the antero- lateral margin are much longer and slenderer than in spinimaims, the posterior one (lateral spine) being but slightly longer, in proportion to the other teeth, than in that species. The chelipeds are slender and fully as long as in sjnnimanus. The ambulatory legs are long and very slender, those of the first two pairs extending nearly to the mid- dle of the dactyli of tlie clielipeds. The sternum is convex in an antero-posterior direction, while in the spinimanns it is quite flat. In the male the terminal portion of the abdomen is narrowly triangular, the penultimate segment being quite narrow and its lateral margins straight or very slightly concave, while in the spinimanvs it is broad and the lateral margins of the penulti- mate segment quite convex. The male abdominal appendages of the first pair are very different in the two species. In both they are stout and separated by quite a broad space. In the sp'n^imanus they reach beyond the middle of the penultimate segment of the abdomen, the thick basal portion cui'v- ing strongly inward from the base, the slenderer portion at first di- rected nearly straight forward, then curved strongly outward, and the tips inward again. In the Ordwayi they are much shorter, reaching but slightly beyond the antipenultimate segment of the abdomen, and have but a single curve, curving inward from the base, then outward to the tip. Length of carapax in the single specimen, 37*0™'" ; breadth of car- apax, 61-8"""; ratio of length to breadtli, 1 : 1*67 ; breadth excluding lateral spines, 48'0"'"' ; ratio of length to this breadth, 1:1-29; greatest length of merus segments of chelipeds, 3 TO'""'; length of hand, right, 47*2, left, 47*0'""'. A male specimen of A. spinimanus from Florida gives the following : — length of carapax, 40-4'"™ ; breadth of carapax, 69-5""" ; ratio of length to breadth, 1 : 1-72 ; breadth excluding spines, 58-5'"™ ; r.atio of length to this breadth, 1 : 1-44. This species differs from the figure of Neptunus crucntatiis (A. Ed- wards, op. cit, p. 326, pi. 31, fig. 2) in having much longer chelipeds, the merus projecting much farther beyond the sides of the carapax, and the hands when folded in front lapping by each other considerably. The teeth of the front and of the anterolateral margin are very much more slender and prominent than in his figure. And m the descrip- tion of the crxtentatus no mention is made of the smooth and highly iridescent spaces on the supero-exterior surface of the hand, which is /S. I. Smith on Brazilian Crustacea. 11 mentioned by Stimpson in liis description of A. Ordvxnji, and is :i very conspicuous character in the species. I have retained this species in the genus Achelous of DeHaan in- stead of Xeptwms of the same aixthor, because the narrow carapax, prominent front, and the form of the external maxillipeds and of the male abdomen ally it very closel y to the spinimanus, and, together with tlie narrow dactyli of the first three pairs of ambulatory legs, separate it widely fi-om Nejytunus pelagicus^ the type of the genus Ncptunxis. The length of the lateral spine of the carapax, whicli appears to have been A. Milne Edwards' principal character for separating these genera, seems to be of slight importance, and in the present case, if used alone, is scarcely sufficient for a specific distinction. Stimpson's specimens of A. Ordwayi were from Florida and St. Thomas. Goniopsis cruentatns DeHaan. Cancer ruricola DeGeer, ]\Iemoires pour servir a I'histoire des Insectes, tome vii, p. 417, pi. 25, 1778 (non Cancer ruricola Linne). Grapsus cruentatus Latreille, Histoire des Crust, et Insects, tome vi, p. 70, 1803 ; Dcs- marest, op. cit, p. 132 ; Edwards, Histoire natureUe des Cnist, tome ii, p. 85 ; Gibbes, loc. cit., p. 181. Gonioims cruentatus DeHaan, op. cit, p. 33, 1835; Edwards, Annales des Sciences natureUes, 3™e serie, tome xx, 1853, p. 164, pi. 7, fig. 2; Stimpson, Proceedings Acad. Nat. Sci., Philadeiphia, 1858, p. 101 ; Heller, op. cit, p. 43. Grapsus longipes Randall, .Journal Acad. Nat. Sci., Philad., vol. viii, p. 125, 1839. Goniopsis ruricola ^hite^ List of Crust in the British Museum, p. 40, 1847 ; Saus- sure, op. cit, p. 30, pi. 2, fig. 18, 1858. Goniograpsus cruentatus Dana, American Journal Sci., 2d series, vol. xii, p. 285, 1851 ; United States Exploring Expedition, Crust., p. 342, pi. 21, fig. 7, 1852. A single male of this beautiful species was collected at the Reefs of the Abrolhos. Cryptograpsus cirripes, sp. nov. Plate I, figure 3. The carapax above is granulous and naked The front as seen from above is nearly straight witli only a sliglit median immargination. The orbits are broad, the margin sliglitly upturned and broken by a broad notch near the inner angle. The outer orbital teeth are long, acutely pointed, project straight forward, and the distance between their tips is nearly equal to two-thirds the breadth of the carapax. The succeeding teeth of tlie antero-lateral margin are prominent and acutely })ointcd, the third tooth much smaller than the others, and the 12 S. J. Smith on Brazilian Crustacea. fourth or last tooth with a sleuder spiuiform tip directed forward and upward and with a sharp granuhited ridge extending from its base inward upon the branchial region and nearly parallel to the postero- lateral margin. The areolation is well pronounced and agrees in the main with C. angulatu-t Dana. In the depression on each side just in front of the anterior lobes of the branchial region there is a trans- verse line of three obscure, oval, smooth spots. From the small tooth in the postero-lateral margin, a short ridge extends backward just above and pai'allel to the margin as far as the lateral angle of the carapax. The ehelipeds are stout and equal. The merus is triangular and the angles granulous. The carpus, and the hand nearly to the tips of the fingers, are sharply granulous. The fingers are slender and their inner edges nearly straight and armed with regular rounded tubercu- liforra teeth. In the ambulatory legs the meral segments are granulous above and on the angles. The dactyli of the first three pairs ai'e naked ex- cept a few hairs on the posterior edge at the base, slender, somewhat curved, smooth and deeply sulcate; those of the posterior pair are shorter, compressed, and their edges thickly clothed with soft hairs. In the first pair of legs the posterior edge of the propodus is clothed nearly its whole length with a brush of soft hair ; in the second pair there is a similar brush but only on the terminal half; in the third pair it is wholly wanting, or represented only by a few hairs near the articulation with the dactylus. In the jDosterior pair of legs the edges of the dactylus, propodus and carpus are densely clothed with soft hair. The male sternum is concave in a lateral direction, and the articula- tions between the segments of the abdomen are nearly straight instead of cur\^ed as in C. angulatus. Length of carapax in a male, 31'0'"™ ; breadth of carapax, o5"6'"'" ; ratio of length to breath, 1 : 1'15. Breadth between outer orbital teeth, 24-8"""; ratio of this breath to breath of carapax between lateral teeth, 1 : 1 -43. This species was not obtained by Prof Ilartt. The only specimens which I have seen are two males, in the collection of the Peabody Academy of Science, Salem, Mass., brought from Rio de Janeiro by Capt. Harrington. The C. clrripes differs from C. angulatus Dana (United States Ex- ploring Expedition, Crust., p. 352, pi. 22, fig. 6), from Rio Negro, Northern Patagonia, and heretofore the only known species of the S. I. Smith on Brazilian Crustacea. 13 genus, in having- tlio front as seen from above nearly straight instead of deeply hilobed, in the much greater breadth of the earapax between the outer orbital teeth — the ratio of this breadth to the breadth of the earapax between the lateral teeth being in C. an'julatus, J : '. -68, — and in the ciliated posterior legs, Uca cordata. Cancer cordatus Linne, Amoeuitatcs Academicoe, tome vi, p. 414, 1763 ; Systema Naturn?, eclitio xii, tome i, p. 1039; Herbst, o]). cit, Band i, p. 131, Tab. 6, fig. 38. Cancer uca Linne?, Systema ]!Taturie, edilio xii, tome i, p. 1041. Uca kevisf Dana?, United States Exploring Expedition, Crust, p. 375. (Nbn Uca una G-uerin, Iconographie du Regne animal, Crust., pi. 5, fig. 3, nee Ed- wards, Histoiro naturelle des Crust., tome ii, p. 22, et Regne animal de CuWer, Sme ^dit., pi. 19, fig. 1.) A single specimen of this species was obtained by Prof. Hartt at Bahia. There are also specimens from Pai'a in the collection of the Peabody Academy. All the specimens examined were males. The earapax is entirely naked and perfectly smooth above, very broad, the greatest breadth being much anterior to the middle, and very convex in an antero-posterior direction. The cervical suture is very distinctly indicated, especially in the middle of the earapax, where there is a broad depression on each side at the antero-lateral angle of the cardiac region. The gastric region is broad and flattened in the middle, the antero-lateral lobes are only indistinctly separated from the median, and the posterior portion is rounded and slightly protuberant but is still lower than the branchial region. The cardiac region is very large, scarcely divided, and the posterior portion ex- tends far back between the bases of the posterior pair of le'>s. The branchial regions are swollen, evenly rounded above and wholly nn- divided, and the lateral margins are very convex in the anterior por- tion and are indicated by a very slight denticulated ridge. The whole front is bordered by a sharply raised margin ; the median lobe pro- jects almost perpendicularly downward between the orbits, and its margin is regularly curved. The orbits are very lai-ge, and the mar- gin is broken by a broad and deep hiatus on the lower side at the outer extremity, just over which the outer angle of the superior mar- gin projects as a rounded lobe ; the inferior margin is nearly straight and is formed of two nearly parallel ridges, the inferior of which is armed with a line of small tubercles, and the superior is irreo-ularly granulous. The inferior obital regions are perfectly smooth and sep- arated from the buccal area by deep sulci. The inferior lateral re- gions are swollen and nearly smooth, there being only a few small 14 S. I. Smith on JBrazllian Crustacea. and scattered granules on the anterior portion near the inferior oi-bital region. On each side of the buccal area there is a liigh ridge which is armed with a few small tubercles. The external maxillipeds are smooth and naked on the outside, and the inner edge and the palpus thickly clothed with coarse hairs. The chelipeds are somewhat unequal and very large. The merns is stout, sharply triangular, both the inferior angles are armed with stout spines and the superior angle is coarsely granulous. The carpus is broad, smooth and evenly rounded on the outside, and spi- nous along the inner edge and on the anterior edge beneath. Tlie hand is broad, compressed, spinous on tlie superior margin and on the inside, the inferior margin granulous, and the outer side smooth; the fingers are high and compressed, their tips strongly incurved, and the inner edges slightly separated in the middle and armed with small irregular teeth except at the tips, which are slightly spoon-shaped with the edges horny, continuous and sharp. The ambulatory legs arc smooth and naked above, but all the segments in the first three pairs, except the basal ones, are thickly clothed beneath and on the anterior side with very long coarse hair. Those of the anterior pair are longer than the others, and those of the posterior pair are much shorter than the others and but slightly hairy. The dactyli of the first two pairs are very long and stout, slightly curved downward, their extremities compressed vertically and five-sided with the angles sharp ; those of the third pair are much shorter and curved backward as well as downwai'd ; those of the posterior pair are still shorter, strongly curved backward and six- sided, the superior side being much broader than the others. The sternum is narrow, very convex in an antero-posterior direction, and the depression for the lodgement of the abdomen is broad, very deep, and extends quite to the base of the maxillipeds. The male abdomen is broadest at the third segment ; the second segment is very small, and the two segments which precede it are completely coalesced. The appendages of the first segment are triquetral and very stout and extend to the extremity of the penultimate segment. The appen- dages of the second segment are very small, extending scarcely be- yond the third segment. Length of carapax, 54'0'""' ; breadth of carapax, 73-4™™; ratio, 1 : 1-36. Length of merus in right cheliped, 33-8'"'"; in left cheliped, 33-0. Length of right hand, 49-5 ; length of left hand, 49-0. One of the specimens in the collection of the Peabody Academy of Science has the chelipeds much more unequal than in the specimen described above but agrees with it in all other characters. S. J. Smith on Brazilian Crustacea. 15 There are at least three American sj^ecies of Uca : — tlie V. cor- data., described above and the IT. una (the species figured by Guerin and Edwards), from the east coast, and U. Imvis^ the species described and figured by Edwards in the Archives du Museum d'llistoire nat- urelle, tome vii, p. 185, pi. 16, from the west coast. The synonymy of these species appears to be in much confusion. The Cancer cordatus of Linne is described at length in the Amoenitates Academicse, and is evidently the species described above and the same as the one figured by Herbst. The description of C. uca in the Systema Naturje is very short and indefinite and no characters are given by which it could be distinguished from the C. cordatus. Milne Edwards in his Historie naturelle de Crust., 1837, quotes both these species under his Vca una Latreille ; he gives " I'Amerique meridionale " as the habitat of U. una, and describes a new species, U. Icevis, from " les Antilles." The slight descriptions of his Icevis here given would not distinguish it from the U. cordata. In his re- view of the Ocypodoidea in the Annales des Sciences naturelles, 3"'^ series, tome xx, 1853, these species are again briefly characterized and the same habitas given. In 1854, in the Archives du Museum, loe. cit., he describes U. Iceris at length and figures it, but says, " Je ne con- nais que des individus males de cette espece ; la plupart ont ete rap- portes des environs de Guayaquil, par M. Eydoux." The description and figure here given apply well to specimens in the Museum of Yale College collected at Guayaquil by Mr. Bradley, and distinguish it readily from the Atlantic species. To add to the confusion, Lucas in D'Orbigny's Voyage flans PAmerique meridionale, Crust., p. 23, 1843 gives, Avithout description, " Uca una Latr." as coming from " Envi- rons de Guayaquil : M. Eydoux," evidently having the same specimens before him that Edwards has described and figured in the Archives du Museum ! If Edwards' original specimens of Icevis were from the West Indies as stated, they are probably the U. cordata, but, even if this be the case, since the east coast species is evidently the Cancer cordatus of Linne, the name Icevis may be retained for the west coast species to which Edwards's last and fullest description and liis fioure apply. White, in the list of Crustacea in the British Museum, p. 31, 1847 has " Uca cordata " from the West Indies and Brazil, but quotes as synonyms, Cancer uca and C. cordatus of Linne, C. cordatus of Herbst, and Uca una of Guerin and Edwards, evidently confoundino- the two Atlantic species and intending to restore the older of the Lhmean names. 16 S. I. Smith on Brazilian Crustacea. Cardiosoma quadratum Saussure Girdisomi quairobtrt, Saussare, op. cit, p. 22, pi. 2, fig. 13, 1858. GirMsomi, diurwn G-lll, Aauils Lye. Nat. Hht., New York, vol. vii, p. 42, January, 1859. [Wrongly printed 1858 on the third signature.] A number of specimens were collected at Pernambuco. It is at ouce distinguished from the G. Guanhumi by the more quadrate form of the carapax, the branchial i-egions being much less swollen, by the latei'al margin being marked by a distinct carina in- stead of evenly rounded, and by the sharply triangular and spiny merus of the chelipeds. Some of the specimens collected by Prof. Hartt are nearly as large as ordinary specimens of G. Guanhumi and still retain the distinctive! characters, so that it seems scarcely possible that it can be the young of that species as suggested by Saussure. This species is in fact more nearly allied to the G. carnifex tlian to G. Guanhumi, and it resembles so closely a species in the collection of the Peahody Academy of Science from the west coast of Afi-ica — apparently the C. armatum of Herklots, — that it might readily be mistaken for it. The African species diifers however in having the carapax less convex and the carina of the lateral margin less promi- nent ; tlie front is broad and high, the anterior lobes of the gastric re- gion are protuberant and the depressed space between them and the frontal margin is coarsely granulous, while in the quadratum the an- terior gastric lobes are not protuberant and the depressed space be- tween them and the frontal margin is scai'cely granulous. The epis- tome and the nasal lobe are quite different in the two species ; in the quadratum the spistome is nearly straight and its anterior margin is not granulated, the nasal lobe is high, forming rather more than a semicircle, and the lobes of the front on each side of it do not reach down to the anterior margin of the epistome, while in the African species the epistome is higher, inore curved and the anterior margin granulated in the middle, and the nasal lobe is much lower, so that the lobes of the front on each side of it reach quite down to the an- terior mai'gin of the epistome. Finally the chelipeds and ambulatory legs in the African species are more spiny and granulous. Specimens of G. quadratum give the following measurements: — Male. Male. Female. Female. Length of carapax, 42-6mm 45 6mm 43-3mm 46-8mm Breadth of " Ratio of length to breadth, Length of merus in right cheliped, "" " hand " " " " merus in left " " hand " " 53-4 55-8 53-3 56 6 : 1-25 1:1-22 J L : 1-23 1: 1-21 21-7mm 28-4m!a 20- 8mm 24 ■4mm 29-0 51-8 30-2 37-0 26-8 23-2 23-4 46-0 31-8 35-5 S. I. Smith 071 Brazilian Crustacea. 17 ANOMOURA. Dromidia Antillensis stimpson. Bromidia Antillensis Stimpson, Proceedings Acad. Nat. Sci., Philadelphia, 1858, p. 225, 1859; Annals Lye. Nat. Hist, New York, vol. vii, p. 11, 1859. Several specimens of this species were obtained by Prof. Hartt at the Reefs of the Abrolhos. They give the follomng measurements and ratios : Sex. Length of carapax including ff ontal teeth. Breadth of carapax. Ratio. Male. 15.5mm IS-gmm 1: l-OI 11 18-2 18-5 1 : 102 Female. 16-0 16-0 1 : 1-00 " 18-0 18-2 1 : 1-01 All the specimens have a covering of tough, fleshy sponge, much broader than themselves, held closely upon the carapax. Stimpson's specimens were from Florida and St. Thomas. Petrochirus granulatus Stimpson. Pagurus granulatus Olivier, Encyclop., tome viii, p. 640 {teste Edwards) ; Edwards, Observations Zoologiques sur les Pagures, Annales des Sciences naturelles, 2de serie, tome vi, p. 275, 1836; Histoire natureUe des Crust., tome ii, p. 225; Dana, United States Exploring Expedition, Crust, p. 453. Petrochirus granulatus Stimpson, Proceedings Acad. Nat Sci., Philadelphia, 1858, p. 233, 1859; HeUer, op. cit, p. 85. A single specimen in a Scolymus was collected by Prof. Hartt at the Reefs of the Abrolhos. CalcinUS SUlcatUS Stimpson. Pagurus sulcatus Edwards, Annales des Sciences naturelles, 2cle serie, tome vi, p. 279, 1836 ; Histoire naturelle des Crust, tome ii, p. 230. Pagwrus tibicen White {variety), List of Crust, in the British Museum, p. 61. Cakinus sulcatus Stimpson, Proceedings Acad. Nat. Sci., Philadelphia, 1858, p. 234. A male of this species was collected at the Reefs of the Abrolhos. Length of body from front of carapax to tip of abdomen, 23 'S"^™ ; length of left hand, 7*6; breadth of left hand, 4-5. It is closely allied to C. tibicen Dana and C. ohscurus Stimpson, but differs remarkably from both of them in the deep and rugose sulcus on the outer side of the propodus of the left leg of the second ambu- latory pair. This sulcus is very marked, extends the whole length of the segment, and is limited on the upper side by a sharp carina. From the ohscurus it differs moreover in having the carapax broader in front, and the antero-lateral angle more prominent, and not rounded, as it is Trans. Connectictut Acad., Vol. II. 2 August, 1869. 18 S. I. Smith on Brazilian Crustacea. in that species. The larger hand is much narrower and more cylin- drical, and the dactyli of the ambulatory legs are not so strongly curved as in C. ohscurus. Clibanarius vittatus Stimpson. Pagarus vittatus Bosc, Histoire naturelle des Crust., tome ii, p. 78, pi. 12, fig. 1, 1802 ; Edwards, Histoire naturelle des Crust., ii, p. 237 ; Gibbes, loc. cit. p. 189. Clibanarius vittatus Stimpson, Proceedings Acad. Nat. Sci., Philad., 1858, p. 335, 1859 ; Annals Lye. Nat. Hist., New York, vol. vii, p. 84. Several specimens were collected at Caravellas, Province of Bahia. They do not differ perceptibly from Florida specimens, except that the hands are perhaps a little less tuberculose. Clibanarius sclopetarius stimpson. Cancer sclopetarius Herbst, op. cit, Band ii, p. 23, Tab. 23, fig. 3, 1796. Pagurus sclopetarius Bosc, Histoire naturelle des Crust., tome ii, p. 76, 1802; Ed- wards, Histoire naturelle des Crust., tome ii, p. 229. Clibanarius sclopetarius Stimpson, Proceedings Acad. Nat. Sci., Philadelphia, 1858, p. 235, 1859 ; Annals Lye. Nat. Hist., New York, vol. vii, p. 85. A single specimen was collected in shoal water at the mouth of the Caravellas River, Province of Bahia, Clibanarius Antillensis Stimpson. Clibanarius Antillensis Stimpson, Proceedings Acad. Nat. Sci., Philadelphia, 1858, p. 235, 1859; Annals Lye. Nat. Hist, New York, vol. vii, p. 85. I refer to this species a large number of specimens collected at the Reefs of the Abrolhos, It is certainly very closely allied to C. Braiiliensis Dana (United States Exploring Expedition, Crust., p, 467, pi, 29, fig, 7), but the opthalmic scales are somewhat larger than rej)resented in Dana's figure, and the right leg of the third pair convex upon the outside. In the alcoholic specimens the ground color of the hands and ambula- tory legs is reddish-yellow, instead of olive. MACROURA. Scyllarus aequinoxialis Fabricius. Scyllarus mquinoxialis Fabricius, Supplementum Entomologice systematic?e, p. 399, 1798; Bosc, op. cit, tome ii, p. 19; Edwards, Histoire naturelle des Crust., tome ii, p. 285, pi. 24, fig. 6. A single male specimen collected at Bahia appears to belong to this species. The carapax is broad, the breadth in front exceeding slightly the length of the lateral margin, evenly convex above, the regions scarce- /S. J. /Smith on Brazilian Cfustacea. 19 ly indicated, and covered, as is also the upper side of the abdomen, with small squamiform tubei'cles of uniform size, and each bearing several small fascicles of short setaceous hairs. The anterior margin, the margin of the orbits, and the lateral margin are armed with numerous, small, obtusely rounded, tuberculiform teeth. The antennulte extend slightly beyond the tips of the antennae ; the basal segments are clothed below with short seta^ ; the terminal seg- ments of the peduncle are smooth and cylindrical ; the inner flagella are nearly as long as the last segment of the peduncle, sparsely ciliate and tapering regularly to a slender point ; the outer flagella are stouter, and considerably shorter than the inner. In the antenna?, the basis is very short and broad, so that, on the outside, the base of the ischium nearly touches the anterior margin of the carapax ; the ischium is much broader than long, the middle portion rough and hairy, the outer and anterior margins smooth and naked, and the edges slightly and irregularly toothed, except the process on the inner side which has two strong teeth upon its inner edge and a smaller one on the anterior edge toward the articulation with the merus ; the car- pal, or last segment, is broader than long, the edge arcuate and cren- ulated, the middle portion above and below roughened mtli short, stiff hairs, but a broad space along the margin smooth. All the inferior surface of the thorax and the exposed parts of its appendages are rough with short, stiff hairs or setse. The thoracic legs have a carina upon the posterior edge of the merus and carpus, which is very high and thin on the merus in all except the posterior pair. The dactyli in the first and second pairs are smooth and unarmed, but in the second pair they are longer and much slenderer than in the first ; in the last three pairs they are armed with fascicles of stout horny setse. The lamellae of the appendages of the second segment of the abdo- men are lanceolate, and the inner and outer of about equal size. The appendages of the three succeeding segments are nidimentary and scarcely project below the edge of the segments. The lamellae of the appendages of the penultimate segment are broadly rounded at the extremities, and the inner ones project beyond the tip of the terminal segment. The terminal segment is broader than long, and the extremity truncate with the angles rounded. The following description of the colors was taken from the speci- men when recently preserved in alcohol, and when, according to Prof. Hartt, the colors were as in life. General color above reddish-brown ; antennae lighter, bordered with bright purple, and the teeth of the edge orange-red ; antennulae light 20 /S. J. SmitJi on Brazilian Crustacea. reddish ; carapax with the frontal and median tubercles, the tubercles of the orbits and of the anterior and lateral margins orange-red ; first segment of the abdomen bright orange, the median portion slightly mottled with purplish-red, and with two large circular reddish-purple spots ; the succeeding segments with the smooth anterior portion, orange mottled with purplish-red ; terminal segment and the lamelli- form appendages of the penultimate segment brownish-yellow, almost white at the extremities. Beneath, dirty yellowish; antennas with the colors of the upper side dimly repeated ; legs with slight purple annulations at the articulations. Length of body from tip of rostrum to extremity of abdomen, - lOO-O™^^ " of carapax from tip of rostrum to middle of jDOsterior margin, 86'0 Breadth of carapax, - - - - - - - 71-2 Lengtli of antennula3, below, - - - - - 55-0 " antennae, '• ..... 52-0 '' first thoracic legs, - - - - - 76-0 " second " - - - - - 92-0 " third " ..... 83-5 " fourth " ..... 72-0 fifth " 75-0 Panulirus echinatns, sp. nov. This species is closely allied to P. guttatus. The carapax is armed with numerous stout spines, those on the anterior part of the carapax larger than those behind; the surface between the spines is closely filled with small tubercles, which are beset with short, stifi^ hairs, and many of the tubercles in front of the cervical suture are tipped with spinules. The cervical suture is mark- ed by a deep depression. The antennulary segment is armed with two straight and slender spines which project forward and upward, their length twice as great as the distance between their tips. The superior orbital spines are stout and long, and extend slightly beyond the tips of the eyes. On the anterior border below the eye, there are two other spines project- ing over the base of the antennae ; from the inner of these there is a line of about eleven smaller spines, three of which are in front of the cervical suture, extending to the postero-lateral angle of the carapax ; below this line there are no spines on the branchial region. Just behind each of the superior orbital spines there is a stout spine as large as the spines on the anterior margin below the eye ; behind these spines, and in front of the cervical suture, there are four smaller spines, thus forming, with the orbital spines, two-subdorsal lines of four spines each, which are succeeded behind the cervical suture, by two S. I. Smith on Brazilian Crustacea. 21 lines of five small spines each. On the median line of the anterior part of the gastric region there are three small, sharp spines. The remaining spines of the carapax are disposed irregularly. The peduncle of the antennula extends slightly beyond the pedun- cle of the antenna ; the basal segments are armed with short setse. The inner flagellum is about as long as the carapax, quite slender and wholly naked ; the outer flagellum is shorter, much stouter, and the terminal portion ciliated beneath. The peduncle of the antenna is a little longer than the breadth of the carapax, and is armed with stout spines, three of which are on the anterior edge of the basis, and another on the inner side, below and near the outer of the three spiniform teeth of the anterior edge of the epistome. The flagellum is about three times as long as the carapax, tapers to a slender j^oint, and is armed with sharp spines. The external maxillipeds, when extended, reach nearly to the an- terior extremity of the basis of the antennae, and all the segments are thickly clothed on the inside, and the dactylus all round, with stiff hairs ; the exognath is rudimentary, about half as long as the dactylus of the endognath, quite slender, and is wholly without a flagellum. The thoracic legs are smooth and naked, except the dactyli and the outer portion of the under side of the propodi ; the meral segments are each armed with two sharp spines, one al)ove and another on the inside at the extremity next the articulation with the carpus. The legs of the first pair are shorter than the others, do not reach quite as far forward as those of the second pair, and the dactyli are stout and thick. Those of the second and third pairs are more slender than the others, especially the penultimate segments, the dactyli straight nearly to the tips, which are hooked abruptly down. The third pair reach slightly beyond the second. The fourth pair extend only to the middle of the propodi of the third pair ; the carpus is armed with a stout and sharp spine on the upper edge of the extremity next the propodus, where there is no spine in the other legs ; the dactylus is stout, the basal portion armed beneath with slender spines, which are articulated at the base and movable, and the terminal portion taper- ing to a slender point and curved evenly downward. The legs of the fifth pair reach to the middle of the propodi of the fourth ; the coxa is armed with a long, sharp spine on the posterior side and near the articulation with the basis ; the dactylus in the male is similar to that in the fourth pair, but shorter and more curved ; in the female the dactylus is somewhat shorter than in the male, and armed on the posterior side of the base with a stout process which closes against a 22 -iS'. I. Smith on Brazilian Crustacea. similar process from the extremity of the propodus, both processes being hairy upon the outside and having horny, spoon-shaped tips. The abdomen is nearly smooth, and all the segments, except the ter- minal, are crossed by a narrow and thickly ciliated sulcus, which is interrupted in the middle on the third, fourth and fifth segments. The first segment has a single, short lateral tooth. The remaining seg- ments, except the last, have this tooth spiniform and vei-y large, and a small additional one behind it ; the larger tooth is armed, except in the penultimate segment, with one or two small spines or denticles on the anterior edge, near the base. The posterior edge of the penulti- mate segment above is armed with close set, sharp teeth. The lanielliform appendages of the sixth segment of the abdomen are of about equal length, broad and truncate at the tips. The lamella of the last segment is slightly narrowed and truncate at the ti}), and does not extend beyond the lamellie of the sixth segment. In the male, the lamelliB of the second to the fifth segment are ovate and all of about the same size. In the female, these lamella? are very much larger ; in the second segment, the inner one is of the same form and nearly of tlie same size as the outer; in the three following segments the outer lamella? decrease in size successively, and the inner lamellae are each composed of two branches, the outer branch being narrow, triangular, its edges thickened, multi-articulate and clothed with long hairs ; the inner branch slender, not tapering, articulated at the base of the outer branch, not jointed like the outer branch, but composed of a single piece, and clothed beneath and at the tip with long hairs. Two specimens give the following measurements : — Length of body from base of antennulaj to extremity of ab- domen, ------- Length of carapax from base of antennulaj to middle of pos- terior margin, . - . . - Breadth of carapax, . - . . - Length of antennulie, . . . - - " inner flagellum of antennula>, " outer " " . - - " antennae, . - - - - " first thoracic legs, . . - - " second, " - " third, " . - - . " fourth, ''.-..- fifth, " ... - Several specimens were obtained at Pernambuco. This species appears to be closely allied to the P. guttatus of the West Indies, but that species, according to Edwards' description and figure Male. Female. 135-Omm 165-Omm 59-5 68-5 36-2 42-2 103-0 109-0 61-4 64-0 48-0 50-8 260-0 290-0 81-0 89-0 92-5 102-2 101-0 111-0 83-0 92-5 72-5 17-0 S. L Smith on Brazilian Crustacea. 23 (Ilistoire Naturelle des Crust., tome ii, p. 297, pi. 23, fig. 1 and 2,) has tlie thoracic legs of the second pair longer than those of the third ; he also states that the ti-ansverse sulci of the abdomen are not inter- rupted on the first three segments ; and raoreovei', in his figures no spines are indicated upon the bases of the antennae, or upon the coxae of the posterior thoracic legs, and the flagella of the antennaa and the antennulrc are much shorter than in our species. Heller (op. cit., p. 95) and DeHaan (op. cit., p. 159), both state that in the guttatus the spaces between the spines of the carapax are smooth, while in our species they are tubei'ctilose and hairy. Neither Edwards, De Ilaan nor Heller mention the sub-cheliform posterior thoracic legs as a character of the female of P. guttatus. Aipheus heterochelis Say. Alphsus heterochelis Say, Journal Acad. N'at. Sci., Philadelphia, vol. i, p. 243, 1818; Edwards, Histou'e naturelle des Crust, tome ii, p. 356; Gibbes, loc. cit, p. 196. Aipheus armiUaiics Edwards?, Histoire naturelle des Crust, tome ii, p. 354, 1837. Aipheus lutarius Saussure, op. cit, p. 45, pi. 3, fig. 24, 1858. A large number of specimens collected at the Reefs of the Abrolhos agree perfectly with specimens from Florida and Aspinwall. Palsemon Jamaicensis Olivier. Cancer {Astacus) Jamaicensis Herbst, op. cit, Band ii, p. 57, Tab. 2T, fig. 2, 1796. Paloemon Jamaicensis Olivier, Encyclop., tome viii, (teste Edwards,) ; Desmarest, op. cit., p. 237 ; Edwards, Histoire naturelle des Crust, tome ii, p. 398, Regne animal de Cuvier, 3^ edit, pi. 3, fig. 4 ; Saussure, op. cit, p. 49. Of this species there are in the collection two specimens, both males, from Penedo, Rio Sao Francisco. In both specimens the rostrum is stout, a little shorter than the antennal scale, and is armed above with twelve, and below with four teeth. The anterior legs are longer than the carapax, and nearly naked, except a few fascicles of hairs on the fingers ; the hands are slender, and about half as long as the carpus, which is sliglitly shorter than the merus. In the smaller specimen the second pair of legs are equal, stout, very long, and thickly beset with small spines; the hands are cylindrical, much longer than the carapax, and the fingers half as long as the palmary portion of the hand. In the larger speci- men the legs of the second pair are quite unequal, the left one being considerably longer and much stouter than the right, and the fingers only a third as long as the palmary portion ; the right hand is much as in the other specimen, but considerably smaller iu proportion. In both specimens the penultimate segment of the abdomen is broad, 24 S. I. Smith on Brazilian Crustacea. the lamellse of its appendages are broadly rounded at their extremities, and the outer ones slightly broader, but scarcely longer, than the inner. The terminal segment of the abdomen is stout, its extremity broad, rounded, ciliate, and has a small movable spine on each side. A single, small and somewhat imperfect specimen, also a male, from Caravellas, Province of Bahia, is apparently the young of this species, but presents some differences. The rostrum is armed with fifteen teeth above and three below, and the legs of the second pair are quite short, extending b»it little beyond the first pair, sparsely spinulose, and the hands quite slender. In other respects it agrees closely with the larger specimens. The three specimens give the following measurements : — Penedo, Sao Francisco. Caravellas. Length from tip of rostrum to extremity of abdomen, ISromm 126-Om'a 54 -4" Length of carapax from orbit to middle of posterior margin, . . . . - 48-0 41-2 18-0 Breadth of carapax, . . - - 27-2 23-5 9-8 Length of rostrum from its tip to base of eyes, 21-8 18-6 8-0 11 basal scale of antenna, 23-0 19-0 8-8 II first thoracic legs, 680 57-8 26-0 11 merus in first thoracic legs, 17-8 15-0 7-0 11 carpus, " '• . - 21-0 16-6 8-4 11 hand, " " - - 12-0 10-5 4-3 " dactylus, " " - 5-8 5-2 2-1 11 second thoracic legs, 114-0— 132-0 115-0 31-2 II merus in second thoracic legs, 20-0— 25-5 25-0 5-9 II carpus, " " . . 16-8— 24-0 17-2 6-0 " hand " "... 54-0— 58-0 59-0 10-8 11 dactylus, " " - - 27-2— 21-0 30-0 5-3 Palsemon forceps Edwards. Histoire naturelle des Crust, tome ii, p. 397, 1837; Saussure, op. cit., p. 51 ; White, List of Crust, in the British Museiun, p. 78. A large number of specimens of this sjjecies was obtained by Prof. Hartt at the mouth of the Para. The larger males agree with Edwards' description. The carapax is granulous, especially on the sides. The rostrum is stout, nearly straight, extends slightly beyond the antennal scale, and is armed above with nine or ten, and below with five to seven teeth. The antennal and hepatic spines are stout and of about equal size. The legs of the second pair are very long, cylindrical, the inner and the inferior sides of the merus, carpvis and the basal half of the hand are armed with about four longitudinal lines of slender spines, the upper and outer I aS*. I. Smith on Brazilian Crustacea. 25 sides tliickly set with short spinulcs and slightly hairy ; the fingers are slender, cylindrical and thickly covered witli a woolly pubescence. The lamelliforni appendages of the penultimate segment of the abdo- men are broadly rounded at their tips, and the outer ones are scarcely longer than the inner. The terminal segment of the abdomen is nar- rower than in P. Jauiaicensis, the sides are straight, and the tip has a strong median tooth and a slender spine each side. The young males are quite similar to the full-grown, but the car- apax is nearly smooth, the rostrum somewhat upturned at the ex- tremity, and the legs of the second pair are smaller in proportion, and the spines and spinviles less developed. The females differ remarkably from the males, all the specimens being considerably smaller, and resembling the young males. The carapax is much more gibbous and quite smooth, even in the largest specimens. The rostrum in front of the eyes curves upward con- siderably, and much more strongly in the small than in the large sj^ecimens. The legs of the second pair are quite slender, much shorter than in the male, only slightly spinulose in the large speci- mens, and almost wholly smooth and naked in the smallest. Of the ten specimens in the collection every one has large masses of eggs imder the alidomen. Five specimens given the following measurements : — Length of body from ti]? of ros- Male. Male. Male. Female. Female. trum to extremity of abdomen, 142 '0™™ 125-Onim 7 5 -Oram 106-Omm •76-Omm Length of carapax from orbit to middle of posterior margin, 36-4 33-5 19-6 27-4 18-0 Breadth of carapax. 23-8 20-4 11-8 18-4 11-2 Length ef rostrum from its tip to base of eyes. 31-0 29-0 17-2 22-6 20-0 Length of basal scale of antenna, 26-5 23-0 15-2 19-7 14-5 " first thoracic legs. 57-0 50-0 31-0 40-0 27-4 " merus in first thoracic legs. 15-2 13-0 7-6 10-4 7-4 Length of carpus. 19-2 17-4 10-5 13-4 9-4 " hand, - 8-0 7-6 4-8 6-0 4-0 " second thoracic legs. 171-0— 158-0 143-0 67-0 — 43-0 75-0 43-0 " merus in second tho- racic legs, - 35-0— 32-4 28-0 13-4_ 9-8 15-0 8-5 Length of carpus, 50-2— 44-0 40-0 20.0—10-0 20-2 14-0 " hand, 60-2— 56-0 50-0 22-6—14-0 22-5 10-8 " dactylus, 28-0— 25-0 240 11-0- 7-5 110 5-2 26 'S'. I. Smith on Brazilian Crustacea. Palsemon ensiculus, sp. noY. Plate I, figure 2. The carapax is somewhat gibbous, and the antennal and hepatic spines are slender, sharp and of about equal size. The rostrum is very long, strongly curved downward for the basal half of its length, the terminal half very slender, nearly straight, but strongly inclined upwards ; it is armed above with nine to twelve short teeth, which are ciliated along their edges, and of which seven or eight are on the basal portion, and the others near the tip, and below with eight to twelve teeth. The eyes are large and the peduncles rather long and slender. The flagella of the antennula are very long, the outer flagellum about as long as the whole body and the inner a little shorter. The peduncle of the antenna is armed with a small spine on the outside just below the articulation of the basal scale ; the basal scale is long but not reaching, by considerable, the tip of the rostrum, the extremity evenly rounded and extending considerably forward of the small, acutely pointed tooth at the anterior extremity of the outer margin ; the fla- gellum is very long, considerably exceeding in length the flegella of the antennuhx). The external maxillipeds are slender, reaching slightly beyond the base of the flagella of the antennae. The first pair of thoracic legs are very slender, reaching slightly beyond the basal scales of the antenna^, smooth and naked, except a few fiiscicles of hairs on the hands. The second pair of legs in the male are very long and quite slender, in full-grown specimens the merus reaching beyond the tip of the antennal scale and all the segments to the base of the fingers closely beset with short spinules ; the hands ai-e cylindrical, not swollen, the fingers slender and sparsely clothed with short, downy pubescence. In the females and young the second pair of legs are considerably smaller and much less spinulose. The third pair of legs reach to the tips of the basal scales of the an- tenna?. The fourth and fifth pairs are successively a little longer. The abdomen is rather slender. The penultimate segment is long and narrow, the length above being nearly or quite twice as great as the breadth ; the lamelliform appendages are rather narrow, the inner ones rather acutely rounded at the tips and reaching a little beyond the terminal segment of the abdomen, the outer ones evenly rounded at the tips and considerably longer than the inner ones. The terminal segment is narrow and tapers regularly to a very slender and acute point. Male. Male. Female. Female. 108 0™m 910mni 89 0mm 65-Omm 25-0 19-3 21-0 14-4 15-5 12 13-G 90 29-0 260 210 20-6 lf)-0 16-0 16-0 12-8 36-4 27-0 28-5 20-0 9-6 7-5 8-0 5-7 11-8 9-0 8-8 6-6 4-8 4-2 40 3-0 103-0 54-0 55-7 32-tf 21-0 11-4 11-2 7-2 30-0 16-7 17-0 10-4 32-5 14-4 15-5 7 14-8 6-7 6-5 2-8 S. I. Smith on Brazilian Crustacea. 27 Several specimens give the following measurements : — Length of body from tip of rostrum to ex- tremity of abdomen, Lengtii of carapax from orbit to middle of posterior margin, . - - - Breadth of carapax, Length of rostrum from its tip to base of eyes, " basal scale of antenna, " first thoracic legs, " merus in first thoracic legs, " carpus " " " hand " " " second thoracic legs, - " merus in second thoracic legs, " carpus " " hand " dactylus " " A large nmnber of specimens of this fine species were obtained by Prof Hartt at Para. Peneus Brasiliensis Latreiiie. Peneus Brasiliensis Latreiiie, Nouveau Dictionnarie d'Histoire naturelle, tome xxv, p. 154 {teste Edwards); Edwards, Histoire naturelle des Crust, tome ii, p. 414; White, List of Crust, in the British Museum, p. 80; Gibbes, loc. cit., p. 198. I refer to this species a large number of small specimens obtained by Prof. Hartt at Bahia. They agree perfectly with a specimen from the west coast of Florida, which is undoubtedly the same as the species described by Gibbes from South Carolina. Xiphopsneus, gen. nov. The cai-apax is much as in Peneus, but the rostrum is very long, its extremity very slender, and the gastro-hepatic sulcus is scarcely per- ceptible, while the cervical and branchio-cardiac sulci are distinct. The antennuLo?, are long and slender, and the peduncle has only a very small lamelliform appendage on the inside, which is not foliaceous and expanded over the eye as in Peneus / the flagella are very long and slender, the upjier ones being much stouter and longer than the lower. The antenna?, maxillipeds and the three anterior pairs of thoracic legs are nearly as in Peneus. The fourth and fifth pairs of thoracic legs are very long, and the terminal segments very slender and flagelliform. The abdomen is quite similar to Peneus, but the lamella3 of the appendages of the first five segments are much longer than is usual in that s:enus. 28 S. J. Smith on lirazllian, Crustacea. This genus has much tlie aspect of Peneus, and is closely allied to it in the antennas, maxillipeds, anterior tlioracic legs and abdomen, but differs from it remarkably in the carapax, antennute and posterior thoracic legs. Xiphopeneus Harttii, sp. nov. Plate I, figure 1. The carapax is not at all swollen ; a very slight, rounded dorsal carina extends from tlie base of the rostrum to the posterior border ; the cervical and branchio-cardiac sulci are very distinct, and together form a nearly straight groove from near the base of the antennas al- most to the posterior border; the inferior margin of the carapax is nearly straight, jjrojecting slightly along the branchial region ; the antennal spine is prominent and rather stout, and the hepatic spine slender and acute. The rostrum is very long and slender, in length nearly equalling or considerably exceeding the carapax, wholly un- armed below, but the basal portion armed above with a thin and liigli carina, w^hich extends back upon the carapax a short distance, and for- ward as far as the eyes, and is armed with five sharp and prominent teeth, and at its posterior extremity with another tooth which is smaller, much below the level of the others, and separated from them by a considerable space ; tlie j^ortion in front of the eyes is nearly straight or a little upturned, sub-cylindrical, slightly flattened laterally, unarmed, perfectly smooth and tapers to a very slender point far in front of the antennal scales. The eyes are of moderate size, and the peduncles much shorter than in most species of Peneus. The appendages upon the inside of the peduncle of the antennulae are siirmounted by a tuft of hairs which fills a little depression in the ocular peduncle. The first antennulary segment in advance of the eye is sub-cylindrical, flattened on the under side, and nearly as long as the peduncle of the eye ; the next anterior segment is cylindrical and one-half as long as the last. The upper flagellum of the antennula is slender, about three times as long as the carapax, and has a short portion at the base slightly thicker than the rest ; the lower flagellum is very slender and about half as long as the upper. The basis of the antenna is armed with a small, sharp spine just be- low the articulation of the antennal scale. The antennal scale reaches to the base of the flagella of the antennula, is much narrowed toward the tip, the outer margin is straight and armed with a sharp tooth at the anterior extremity, and the inner margin is nearly straight and S. I. Smith on Brazilian Crustacea. 29 thickly ciliated. The three anterior segments of the peduncle are cylindrical, and the last (carpal) is much longer than in most species of JPeneus, so that it reaches to the middle of the antennal scale. The flagellum is very much longer than the Avhole length of the body. The second pair of maxillipeds, when extended, reach nearly to the base of the antennal scale ; the merus is nearly three times as long as broad, and thickly hairy on the inner edge ; the exognatli is very slender, clothed along the edges with long cilia, and scarcely reaches the tip of the extended dactylus. The external maxillipeds reach slightly beyond the middle of the antennal scale and are thickly setose along the inner edges ; the exognath is slender, extends slightly beyond the merus of the endognath, and is ciliated as in the maxilli- peds of the second pair. The thoracic legs of the first pair reach about to the middle of the propodus of the external maxillipeds, are slender and beset with stiff hairs along the edges, and the basis is armed with a short spine on the inner side near the articulation with the ischium. The second and third pairs of legs are successively a little longer, perfectly smooth, and the basal segments unarmed. The legs of the fourth and fifth pairs are smooth and unarmed, and all the segments, except the coxal and basal, are very slender and very much prolonged, the terminal segments being fully as slender as the terminal portions of the flagella of the antennulae. The abdomen is compressed, and upon the fourth, fifth and sixth segments there is a dorsal carina which is high and sharp upon the sixth, and terminates posteriorly in a slight tooth upon the fifth and sixth. The terminal portion of the appendages of the first segment is long, slender and ciliated along the edges ; in the appendages of the four succeeding segments the outer of the terminal branches are like the tenninal portion of the appendages of the first segment, and of about the same length, while the inner branches are but half as long. The penultimate segment is strongly compressed, and its lamel- liform appendages are rather long and naiTOw, the inner ones project- ing considerably beyond the terminal segment, ciliated alono- both edges and narrowly triangular at tip, the outer ones ciliated along the inner edges and rounded at the tip. The terminal segment tapers regularly to a very slender and acute point, the edges of the terminal half are ciliated, and there is a deep median groove upon the dorsal surface. In the male, the appendages of the first abdominal segment (plate I, fig. 1=^), are connected together near their bases by a peculiar sexual 30 S. I. Smith on Brazilian Crustacea. organ which depends between them, and consists of a central tubular portion articulated with the bases of the abdominal appendages by a short process on each side and furnished at the lower extremity with two stiif, horn-like, tubular processes. The central portion is open on the postei'ior side for its whole length, and the membrane of which it is composed is folded into deep longitudinal grooves, except on the anterior side which is smooth and flattened, and traversed longitudi- nally by a median suture. The horn-like, terminal processes curve slightly backward and downward, and have an opening on the lower side at the tips. The inner of the terminal branches of the append- ages of the second abdominal segment are furnished at the base on the anterior side with a small, ovoid, flattened, cushion-like organ which is wanting in the appendages of the other abdominal segments, and in all of those of the female. Three specimens give the following measurements : — Length of body from tip of rostrum to extremity of ab- domen, ...... Length of carapax from orbit to middle of posterior margin, ...... Breadth of carapax, .... Length of rostrum from tip to base of eyes, " basal scale of antenna, " first thoracic legs, .... " hand in first thoracic legs, " second thoracic legs, " hand in second thoracic legs, " third thoracic legs, . . - - " hand in third thoracic legs, " merus in fourth thoracic legs, " carpus, " " - - " fifth thoracic legs, . . - - " merus in fiftli thoracic legs, " carpus " " - " propodus " " - - " dactylus " "... " first pair of abdominal appendages, " second " " - - Several specimens of this remarkable species — all of them some- what broken and in rather bad condition — were obtained by Prof. Hartt at Caravellas, Province of Bahia. Male. Female. Female. 87 -Oram 133 -Omm 112-Omm 18-0 31-8 25-5 8-5 15-0 12-5 22-0 31-5 26-0 13-4 20-8 18-4 17-5 29-0 25-4 4-3 7-7 6-1 22-2 41-5 35-0 5-4 10-0 8-2 31-5 58-0 46-0 6-2 12-8 9-8 14-2 32-2 20-0 14-6 85 + n-5 27-0 23-5 21-0 27-5 29-4 23-0 16 + 21-6 32-0 29-0 22-0 32-5 29-4 S. I, Smith on Brazilian Crustacea. 31 SQUILLOIDEA. Gonodactylus chiragra LatreiOe(?). Squilla chiragra Fabricius, Supplementum Entomol. systematicse {teste Edwards). Gonodactylus cMragrus Latreille, Encyclopedie methodique, tome x, p. 413, plate 325, fig. 2 (fesfe Edwards); Edwards, Histoire naturelle des Crust., tome ii, p. 528, Gibbes, loc. cit., p. 201. A species of Gonodactylus was collected by Prof. Ilartt at the Reefs of the Abrolhos and at Caravellas, Province of Bahia, which does not diifer from the common West Indian and Florida species. The American species is, however, very likely distinct from the true G. chiragra of the old world. In the foregoing list 32 species are mentioned, of which 21 appear to be new to the fauna of Brazil; and of these 21 species, 6 are des- cribed as new to science, and the remaining 15 are all species pre- viously known from the West Indies or Florida. In order to give a better idea of the crustacean fauna of the whole Brazilian coast, I append the following list. List of the described species of Brazilian Podopthalmia. Previous to Milne Edwards' general work,* scarcely anything was known of the Crustacea of South America, and even in this work Edwards records Brazil as the habitat of very few sj)ecies. Some additional species, however, are recorded in his later papers on the Ocypodoidea,f and Alphonse Milne Edwards has added a single species in his monograph of the Portunids.J A few other species are men- tioned in short papers by Bell,§ Weigman,|| and Bate,^ and quite a * Histoire natureUe des Crustaces. Paris; tome i, 1834; ii, 1837; iii, 1840. •j- Observations sur la Classification des Crustaces. Annales des Sciences naturelles, 3me serie; De la famille des Ocypodides, tome xviii, 1852, pp. 128-166, pi. 3-4; Suite (1), tome XX, 1853, pp. 163-228, pi, 6-11. — Notes sur quelques Crustaces nouveaux ou peu connus. Archives du Museum d'Histoire natureUe, Paris, tome vii, pp. 145-192, pi. 9-16, 1854. X Etudes zoologiques sur les Crustaces recents de la famile des Portuniens. Archives du Museum d'Histoire naturelle, Paris, tome x, pp. 309-428, pi. 28-38, 1861. § Some Account of the Crustacea of the coasts of South America. Transactions Zoological Society, London, vol. ii, pp. 39-66, pi. 8-13, 1841, and Proceedings Zoological Society, 1835, pp. 169-173. II Beschreibung einiger neuen Crustaceen des Berhner Museums aus Mexiko und Brasilien. Archiv fiir Naturgeschichte, 1836, Band i, pp. 145-151. T[ Carcinological Gleanings, No. III. Annals and Magazme of Natural History, 4tB series, vol. i, June, 1868, p. 447. t 32 S. I. Smith on Brazilian Crustacea. number of species are indicated by White in the list of Crustacea in the British Museum,* but unfortunately descriptions of many of the new species have not yet appeared. But by far the largest accessions to our knowledge of the Crustacea of this coast were made by Prof. Dana in his work on the Crustacea of the United States Exploring Expedition.! Although the expedition touched on the Brazilian coast only at Rio de Janeiro, over forty species of Podophthalmia alone were collected and described. More recently Heller has enume- rated the species taken by the naturalists accompanying the Austrian Expedition round the world during the years 1857-1859, J Unfortu- nately, however, this expedition also touched only at Rio de Janeiro, and consequently but few species were obtained which were not observed by Dana. From the works of these authors, Prof. Harrt's collection, and a few species in the collection of the Peabody Academy of Science, the following list has been compiled. A few species, of which the localities are questionable or suspected are preceded by a mark of doubt, thus (?), but all queries which are not inclosed in parenthesis are quoted directly from the author whose name they precede. When I have personally examined specimens from the localities mentioned, they are followed by an ! . In all other cases the authority on which it is inserted follows the locality. BRACHYURA. MAIOIDEA. Maiid^. Lihinia spinosa Edwards. "Les cotes du Bresel" (Edwards, Hist. nat. des Crust., tome i, p. 301). Libidoclea Brasiliensis Heller. Rio de Janeiro (Heller, op. cit., p. 1). MiTIIEACIDyE. Mithrax hispidus Edwards. Abrolhos ! (Hartt). — Antilles (Edwards). Tortngas, Key Biscayne (Stirapson). South Carolina (Gibbes). Mithraculus coronatus Stimpson. Abrolhos ! (Hartt). — AspinwaU ! (F. H. Bradley). Tortugas (Stimpson). * List of the specimens of Crustacea in the collection of the British Museum. Lon- don. 1847. f United States Exploring Expedition, during the years 1838-42, under command of Charles Wilkes, U. S. N., vol xii. Crustacea. Philadelphia, 1852. Plates, 1855. X Reise der osterreichischen Fregatte Novara um die Erde. Zool. Theil, zweiter Band, dritte Abtheilung, Crustaceen. Wien, 1865. S. I. Smith on Brazilian Crustacea. 33 EURYPODID^. (V) Eurypodius Latreillii Guerin. Rio de Janeiro (Bell, Transactions Zoological Society, London, vol. ii, p. 40). — Chili (Edwards and Lucas, Bell, White, Dana).— '-Les lies Malouines " (Edwards, Hist. nat. des Crust., tome i, p. 284). There is probably some confusion of localities here. Bell alone mentions the spe- cies as coming from Brnzil, and as he had it also from Chili, some interchange of specimens may have taken place. The Chilian species is very likely distinct from the East Indian one. Pericekid^. Milnia hicornuta Stimpson. Abrolhos! (Hartt).— Aspinwall ! (P. H. Bradley). Antilles (Edwards, Saussure). Jamaica (White). EloridaKeys! (E. B. Hunt). Bermudas 1 (J.M.Jones). Peltinia scutiformis Dana. Rio de Janeiro (Dana). Acanthonyx Petiverii Edwards. " Coast of Brazil " (Bell).— Antilles (Edwards).— (?) Valparaiso (Dana). (?) Gala- pagos Islands (Bell). Epialtu8 Srasiliensis Dana. Rio de Janeiro (Dana). Epialtus marginatxs Bell. "Ad oras Brasilia}" (Bell, Proceedings Zool. Soc, London, part iii, 1835, and Trans- actions Zool. Soc, London, vol. ii, p. 62). — '-Ad Insulas Galapagos" (Bell, Trans- actions Zoo). Soc, loc cit.). The specimens from the two coasts are probably distinct species, and if so the name marginatus should be retnined for tlie Brazilian one. as in the first description Bell mentions only the Brazilian specimen. There is some confusion in regard to the locality from which the west coast specimen came, the habitats being given as quoted above, but in the remarks following the description in the Transactions, it is stated that the male specimen canje from Valparaiso, where it was found in company with E. dentatus by Mr. Cuming. Lucippa levis Dana. Rio de Janeiro (Dana). CANCROIDEA. Xantuid^. Xantho parv.ula Edwards. Brazil (Edwards). — Antilles (Edwards). Cape de Verdes (Stimpson). Xantho dlspar Dana. Rio de Janeiro ? (Dana). ' Xantho denticidata White. Abrolhos! (Hartt).— West Indies (White). Aspinwall! (P.H.Bradley). Bermudas! (J. M. Jones). Trans. Connecticut Acad., Vol. II. 3 August, 1869. 34 S. I. Smith on Brazilian Crustacea. (?) Menippe Rumphii DeHaan. Rio de Janeiro? (Dana). Pernimbuco (Wliite). — Jamaica (White). — East Indies (Herbst, Edwards, etc.). TUe American species is probably distinct from the true EumpMi ot the East Indies, Panojieus poUttis Smith. AbrolhosI (Hartt). Panojyeus Harttii Smith. Abrolhos! (Hartt). Panopeus Herhstli Edwards. Rio de Janeiro (Heller, op. cit., p. 16). — Aspinwall ! East and west coast of Flor- ida! Bahamas 1 South Carolina! Ghlorodius Floridanus Gibbes. Abrolhos ! (Hartt) —Key West 1 (Gibbes). Aspinwall ! (F. H. Bradley). Pilumnus Quoyi Edwards. Rio de Janeiro (Edwards). ERirillDuE. Erlphia gonagra Edwards. Rio de Janeiro (Dana, Heller). Abrolhos! (Hartt).*— Aspinwall! (F.H.Bradley). Tortugas (Stimpson). Florida Keys! (E. B. Hunt). Bahamas! (Coll. Bost. Soc. Nat. Hist.). — (?) Panama (Stimpson). PORTUNID^, Gallinectes ornatus Ordway. Caravellas! (Hartt).f — Cumana; Hayti; Tortugas; Bahamas: South Carolina (Ord- way). BermndHs! (.1. M. Jones). Gallinectes larvatus Ordway. Bahia ! (Hartt). — Hayti ; Turtuo-as ; Key West ; Bahamas (Ordway). Gallinectes Dan(ff Smith. Pernambuco ! (Hartt). Rio de Janeiro (Dana). Acheloils spinimaniis DeHaan. Rio de Janeiro (Dana, Heller). Bahia! (Hirtt).:]: — South Carolina (Stimpson. A. Ed- wards). West Fl')rida! (E. Jewett). Martinique (A. Edwards). Acheloils Ordwagi Stimpson. Bahia 1 (Hartt). — St. Tiioraas ; Tortugas ; Bay Biscayne (Stimpson). Acheloils Sebce. [JSfeptunus Sebce A. Edwards). '•Les cotes du Bresil" (A. Edwards). — Martinique (A. Edwards). Gronius ruber Stimpson. Brazil (Edwards, White, .I.Edwards). Rio de Janeiro (Heller). — St. Thomas (Stimp- son). Gulf of Mexico ; Vera Cruz (A. Edwards). Key West (Gibbes). — Panama (Stimpson). * This species was collected from the whole coast. It is very lively, running over the rocks and hiding in holes at low water. — c. f. h. f Taken in nets in shallow water on the borders of the bay. — c. f. h, \ Taken in shallow water and sold in the market for food. — c. r. H. /S. J. Smith on Brazilian Crustacea. 35 Arenmus c; ibrarius Daua. Rio cl3 Janeiro (Dana).— Guadaloiipe; Gulf of Ifexic >; Vera Craz (A.Edwards). Key West; South Carolina (Gibhes). New Jersey (Leidy). Platyonychid^. (?) C<(rchius Jlcenas Leach. Eio de Janeiro (Heller, oi). cit., p. 30) — European coast. OCYPODOIDEA. GONOPLACID^. Eucratopsis crassimanus. {Eucrete crasshnanus Dana).* Rio de Janeiro ? (Dana). OCYPODID^. Gelasiinus maracoani Latreille. Rio de Janeiro (Dana). Pernambuco (White). Porto Seguro ; St. Cruz (Hartt). — Cayenne (Edwards). West Indies (White). Gelasimus palustris^diw^vdi^. [G. vocans Dana). Rio de Janeiro (Dana, Stimpson). — Aspinwall; Hayti; Texas; South Carolina; Old Point Comfort (Stimpson), Gelasimus mordax, sp. no v. Pariii (Caleb Cooke, Coll. Peabody Acad. Sci.). (?) Gelasimus stenodactylus Lucas. "Bresil" (Edwards, Annales des Sci. nat, 3™? serie, tome xviii, 1852, p. 149). — Chili (Lucas, Edwards). Ocypoda rhomhea Fabricius. Rio de Janeiro (Dana, Heller).— Jamaica (White). Gecarcinidje. Gecarcinus sp. White (List of Crust, in British Museum, p. ;12). Pernambuco (White). * Stimpson, from an examination of alcoholic specimens of Ericrate crenaius De Haan, has shown (Boston Journal Nat Hist., vol. vii, p. 588, 1863) that DeHaan's genus Eucrate ia distinct from the .E'Mcrafe as described by Dana, DeHaan's genus hav- ing the male organs, or verges, arising i'vnn ihe coxas of the posterior legs, and there- fore belonging to the Carcmop^aci'tice i if Edwards, while Dana's species has sternal verges, and must therefore form the type of a new genus, for which I propose the name Eucratopsis. The genus thus constituted appears to be nearest allied to Speocarcinus Stimpson (Annals Lye. Nat. Hist., New York, vol vii, p. 59, 1859), from which it is distinguished by the larger orbits, by the approximation of the inner margin of the maxillipeds, and by the much greater narrowness of the posterior part of the sternum. 36 S. I. Smith on Brazilian Crustacea. Felocarcimis Lalandei Edwards. ( Gecareoidea Lalandei Edwards). Brazil (Edwards). Cardiosoma Guanhumi* Latreille. Brazil (White).— Am illes (Edwards, Saussure). Florida Keys! (Gibbes). Cape de Verdes (Stimpson). Cardiosoma quadratum Saussure. Pernambuco! (Hartt).f — AspiuwalU (F. H. Bradley). Hayti (Saussure). Barba- does ; St. Thomas (Gill). Uca cordata. Bahia! (Hartt). Para! (Coll. Peabody Acad. Sci.).- Surinam (Linne). (?) Uca una Latreille, Edwards.J •'Araerique meridionale" (Edwards). Rio de Janeiro (Von Martens, Zo51. Record, vol. iv, 1867, p. 613). Trichodactylid^. Trichodactylus quadratus Edwards. {T. fluviatilis Latreille ?), Brazil (Edwards). Rio de Janeiro (HeUer). (?) Trichodactylus punctatus Eydoux et Souleyet ?, Dana. Rio de J.meiro (Dana). Trichodactylus (?) Cunninghami. ( Uca Cunninghami Bate).§ Tijuca, Province of Rio de Janeiro (Bate). Syluiocarcinus Deoillei Edwards (Archives du Museum d'Hist. nat., tome viii, p. 176). " Dans la riviere de I'Araguya, a Salinas, province de Goyas " (Edwards). Dilocarcimis emarginatus Edwards (Archives du Museum d'PIist, nat., tome viii, p. 181). '•Loretto, sur la Haute- Amazone " (Edwards). Dilocarcinus pictus Edwards (Archives du Museum d'llist. nat., tome viii, p. 181). '' Loretto ( Haute- Amazone) " (Edwards). Dilocarcinus Castelaaui Edwards (Archives du Museum d'Hist. nat. tome viii, p. 182). " Salinas (province de Goyaz) " (Edwards). * Prof. Hartt informs me that this species, which lives in the mangrove swamps, is called Guayama, and that it is mentioned under that name by Fonseca, so the specific name Guanhumi is probably a mistake for Guayamu. \ Taken in swamps. — c. F. H. \ According to Prof. Hartt a species of Uca is still called in Brazil V^a-^tna. A tracing of tlie original figure of Marcgrave, however, indicates that his Vra-i'ina was not the Uca una of Latreille and Edwards, but more likely the U. cordata. § Annals and Mag. Nat. Hist., 4th series, vol. i, June, 1868, p. 447, pi. 21, fig. 3. S. I. Smith on Bra'dlian Crustacea. 37 GeAPSIDuE. Goniopsis cruentatus DeHaan. Rio de Janeiro (Dana, Heller). Abrolhos! (Hartt).*— Surinam (Randall). Cuba (Saussure). Florida Ke3's ! (Coll. Best. Soc. Nat. Hist.). Pachygrapsus simplex Stimpson. ( Goniograpsus simplex Dana). Rio de Janeiro? (Dana). — Madeira (Stimpson). Pachygrapsus intermedius Heller {o\\ eit., p. 44). Rio de Janeiro (Heller). (?) Pachygrapsus innotatus Stimpson ( Goniograpsus innotatus Dana). "Locality uncertain; probably from the South American coast" (Dana). — Madeira (Stimpson). If Dana's specimens came from South America, as supposed, they were undoubtedly from Brazil, since Stiaipsoii's discovery of it at Madeira shows it to be an Atlantic species and the Wilkes Exploring Expedition touched, on the east coast of South America, only at Rio de Janeiro and on the coast of Patagonia. Pachygrapsus rugulosus. {Leptograpsus rugulosus Edwards). " Bresil " (Edwards). This species is very likely tlie same as P. innotatm, which, according to Stimpson. is scarcely to be distinguished from P. transversus Gibbes. Edwards' descrip- tion, three lines in length, is, however, too imperfect to determine anything in regard to the affinity of the species. Pachygrapsus maurus Heller (Lucas). Rio de Janeiro (Heller). — Mediterranean (Lucas, Edwards, Heller). (?) Pachygrapsus marmoratus Stimpson. [Goniograpsus varius Dana ?). Rio de Janeiro? (Dana). — Madeira (Stimpson. Heller). Gibraltar (Heller). Medi- terranean (Edwards, Heller). Cryptograpsus cirripes Smith. Rio de Janeiro ! (Coll. Peabody Acad. Sci.). JVautilograpsus sp. (" Planes " White). Brazil (Wh te. List of Crust, in British Museum, p. 42). Cyclograpsus integer Edwards. Brazil (Edwards). — Florida (Stimpson). Helice granulata HeWer (op. cit., p. 61). (Chasmagnathus granula- tus Dana). Rio de Janeiro (Dana, Heller). Rio Grande! (Capt. Harrington, Peabody Acad. Sci.). (?) Sesarma angustipes Dana. South America (Dana). — Aspinwall; on the east coast of Central America, neae Graytown; Florida (Stimpson). Since this has proved to be an east coast and tropical species, there can bo littl doubt that Dana's specimens were from Rio de Janeiro. * Found running about over the rocks at low tide on the fringing reef. It did not appear to be common. — c. f. h. 38 S. I. Smith o)i JBrazUian Cnistacea. Aratus Pisonii Edwards. (Sesarma Pisonii Edwards). Rio de Janeiro (Heller). — Antilles (Edwards). Jamaica (While). Florida (Gibbes, Stimpson). CALAPPOIDEA. Hepatic^. Jlepatus angiistatus White. [H. fascidtus Latreille, Edwards). Rio de Janeiro (Dana, Heller). — Aspinwall (Stimpson). ANOMOURA. Dromidia Antillensis Stimpson. Abrolhos! (Hartt). — St. Thomas!; Tortugas; Key Biscayne (Stimpson). PORCELLANID^. Petrolisthes leporimis. {Porcellana leporina Heller). Rio de Janeiro (Heller). The figure and description given by Heller would scarcely distinguish this species from the /'. armatus Stimpson (Gibbes sp,). Petrolisthes Prasiliensis, sp. nov. {Porcellana Poscii? Dana, p. 421, pi. 26, tig. 11, non Savigny, Crust. Egypt, pi. 7, fig. 2). Rio de Janeiro (Dana). Pachycheles ^noniUferus Stimpson (Dana). Rio de Janeiro (Dana). Porcellana frontalis Ileller. Rio de Janeiro (Heller). Minyocern^ angustus Stimpson (Dana). Rio de Janeiro (Dana). Hippa emerita Fabrieius. Rio de Janeiro (Dana, Heller). HlPPID^. Cexobitid^. Cenohifa Diogenes Latreille. Brazil (White, List of Crust, in British Museum, p. 61). Pagurid^. Petrochirus granulatus Stimpson (Olivier). Rio de Janeiro (Dana, Heller). Abrolhos! (Hartt).— Antilles (Edwards) Key West (Gibbes). "West coast of Florida! (E. Jewett). S. I. Smith on Brazilian, Crustacea. 39 Calcinus sulcatus Stimpson (Edwards). Abrolhos ! (Hartt). — Antilles (Edwards). White reports C. Ubicm Dana from Brazil and the West Indies, but as he included C sulcatus as a synonym, his specimens were perhaps all of this species. Clibanarius Brasiliejisis Dana. Rio de Janeiro (Dana). Clibanarius A?itillensis Stimpson. - Abrolhos ! (Hartt). — Barbadoes (Stimpson). Clibanarius vittatus Stimpson (Bosc). Abrolhos! (Hartt). — Key West; Charleston (Gibbes). West coast of Florida! (E. Jewott). Clibanarius sclopetarius Stimpson (Herbst). Caravellas River, in the Province of Bahia ! (Hartt). — Trinidad (Stimpson). Aspinwall ! (F. H. Bradley, Stimpson). Tortugas (Stimpson). Eupagurus criniticornis Stimpson (Dana), Rio de Janeiro (Dana). (?) Eupagurus scabrieulus Stimj^son (Dana). Brazil ? (Dana). (?) GALATEIDiE. Under tlie name of Galathea amplectens, Fabricius, in his supple- mentmn Entomologise systematicae, p. 415 [teste Edwards), has des- cribed a crustacean from Brazil which seems to be unknown to subsequent writers. It is probably not a true Galathea. MACROURA. SCYLLARID^K. Scyllarus mquinoxialis Fabricius. Brazil (White). Bahia! (Hartt).*— Antilles (P^dwards). Key West (Gibbes). Palinurid.e. Panulirus argus White. {Palinurus argus Latreille, Edwards). Bahia (White).— Antilles (Edwards, White). Panulirus eelnnatus Smith. Para! (Hartt) f Pal.emonid.'E. Alpheus heterochelis Say. Abrolhos! (Hartt). — Aspinwall! (P. H. Bradley.) Cuba (Saussure). Key West (Gibbesj. West coast of Florida! (E. Jewett). South Carohna (Gibbes. Say). * T.iken in shallow water on the borders of the bay and used for food. — c. F. ii. f Used for food and sold in the market. I have seen it from much farther south. — c. F. H. 40 S. I. Smith on Brazilian Crustacea, Alpheiis tridentulatus Dana. Rio de Janeiro ? (Dana). Alpheus nialleator Dana. Rio de Janeiro ? (Dana). Hippolyte exlliro stratus Dana. Rio de Janeiro (Dana). Hipi^olyte ohliquimanus Dana. Rio de Janeiro (Dana). Palmmon Jamaicensis Edwards. Penedo, Rio Sao Francisco! (Hartt).* Pernambuco (White). — Antilles (Edwards). Antilles and Gulf of Mexico (Saussurej. Palmmon spinimanus Edwards. Brazil (Edwards, White). — Antilles (Edwards). Cuba (Gibbes). Palcemon OZ/ersi* Weigman (Arcliiv fiir Naturges. 1836, p. 150). "An der Kiiste Braziliens " (Wiegman). Palcemon forceps Edwards. Pernambuco (White). Rio de Janeiro (Edwards). Mouth of the Para ! (Hartt). — Antilles, Gulf of Mexico (Saussure). PalcBmon acanthtirus AYiegman (loc. cit., p. 150). " Das Vaterland ist die Kiiste Braziliens " (Wiegman). Palmmon ensieulus Smith. Para! (Hartt). (?) " Palmmon Lamarrei Edwards ? " (White). Pernambuco (White). — Cotes du Bengale (Edwards). Penkid^. tSicyonia carinata Edwards. Rio de Janeiro (Edwards, Dana). Peneus Brasiliensis Latreille. Brazil (Latreille, White). Bahia! (Hartt). — West coast of Florida! (E. Jewett). South Carolina (Gibbes). Peneus setiferus Edwards. Rio de Janeiro (Heller). — Florida (Edwards). South Carolina (Gibbes). Xiphopeneus Harttii Smitli. Caravellas, Province of Bahia! (Hartt). * This species, called pitii, is quite common in the river Sao Francisco and the larger streams flowing into it. — c. f. h. S. I. Smith on JBrazUlmi Crustacea, 41 SQUILLOIDEA. Squillid.e. Lysiosquilla inornata Dana. Rio de Janeiro (Dana). Squilla ruhro-lineata Dana. Rio de Janeiro ( Dana). Squilla prasino-Uneata Dana. Rio de Janeiro (Dana). Squilla scabricazida Latreille. Brazil (White). Gonodactylus chiragra Latreille. (?) Abrolhos! (Hartt). Caravellas, Province of Bahia! (Ilartt). — Aspinwall ! (F. H. Bradley). Florida Keys! (Gibbes). Bermudas! (J. M. Jones). — Mediternmean Red Sea ; Pacific Ocean (Authors). Erichtiiid^e. ErichtJnis vestitus Dana. South Atlantic, lat. 25° south, long. 44° west (Dana). Erichthus spiniger Dana. South Atlantic, between Rio Juneiro and Rio Negro (Dana.) MYSIDEA. MySIDyE. Macromysis gracilis Dana. Rio de Janeiro (Dana). Jiachitia sjnwflis Dana. Atlantic, off the harbor of Rio de Janeiro (Dana). LUCIFERID.E. Lucifer acicularis Dana. Harbor of Rio de Janeiro (Dana). Zocc(. ruheJla Dana. South Atlantic, lat. 24° 45' south, long. 44° 20' west (Dana Zoea echinus Dana. Atlantic, lat 23° south, long. 41° 5' west (Dana). EXPLANATION OF PLATE I. Figure 1. — X'phopeneus IfurUii, male, cephalothorax ; a, b, c, d, e, thoracic legs, tho-e of the fourth and fifth pairs incomplete, la, appendages of the first segment of the abdomen in the same specimen, lb, rostrum of a larger, female specimen ; Ic mandible enlarged two diameters, Fi""nre 2. — P^dcemon ensiculus, male, carapax ; 2a. leg of the second pair ; 2b, extremity of abdomen, seen from above ; 2c, rostrum of a small female. Figure 3. — Cri/ptognqisus cirripes, male; 3a. sternum and abdomen of the same spe- cimen. Figure 4.— Pa opens jmIUics, female, carapax enlarged two diameters. Figure .5. — Panoneus IliritH, male, carapax enlarged two diameters. All the figures are natural size, except Ic, 4 and 5, and all are copied from photo- graphs, except 1 a and Ic. Topographical Map of the New Haven region. Explanations. — A, AUingtown village. B, Beacon Hill. Bh, Beaver Hills. Ch, Cher- ry Hill. E, East Rock range, consisting of ?]ast Rock proper to the northwest, Indian Head, and then Snake Rock. Ed, Edgewood, the estate of Donald G. Mitchell, Esq. F, Fort Hale F, Ferry Point, or Red Rock, on the Quinuipiac near its mouth. J, Judges' Cave, on the "West Rock ridge. L, Light House. M, Mill Rock. M P, Maltby Park, only three of the proposed lakes of which are constructed. 0, Oyster Point. P, Pine Rock. Rd, Round Hill. Rt, Rabbit or Peter's Rock. Sm, Sachem's ridge. T, Turnpike; also Tomlinson's bridge, across the head of New Haven bay. V, the village of Whitneyville. "\Y, West Rock, the south end of West Rock ridge. WC, West Cape, or West Haven Point. Wh, Whitney Peak. WL, Wintergreen Lake, just north of Wintergreen Falls. Wn, Warner's Rock. 6??2, Beaver Pond Meadows ; m, Mineral Spring, southeast of North Haven ; «l, rfi, rfi, w-i, different notches in the West Rock ridge ; «l, rfi, the upper and lower Bethany Notches; m^, the Hamden Notch; m^, Wintergreen Notch. The names of the towns ORANGE, WOODBRIDGE, BETHANY show the course of the Woodbridge plateau; and from W in the word Westvillo to Savin Rock is the course of the Edgewood series of hills, the eastern border of the plateau. Scale 4-lOths of an inch to the mile. II. Ox THE Geology of the New Haven Region, with special REFERENCE TO THE OrIGIN OF SOME OF ITS TOPOGRAPHICAL FEA- TURES. By James D. Dana. with a map. 1. The 'New Haven reg;on. Either side of New Haven bay, — an indentation of the coast about four miles in depth, — there is a north-and-south range of hills, the trap and sandstone ridges of East Haven and North Haven on the east, and the eastern portion of the Woodbridge plateau on the west ; and these make the eastern and western boundaries of the New Haven region. Their height, which is greatest to the north, probably nowhere exceeds 600 feet. The width of the region varies from about four miles on the south to seven on the north, and the whole length from the Sound to Mt. Carmel — its true northern topographical limit — is twelve miles. The northern half of the resfion is divided lonsfitudi- nally by two lines of ridges : (1) the long West Rock trap ridge near the western side, four hundred feet and upward in height ; and (2), nearly midway in the area east of West Rock, the short isolated East Rock (E) range of trap and sandstone, and the continuation of this range northward to Mt. Carmel in the low Quinnipiac sandstone ridge which divides the waters of Mill River and the Quinnipiac. The New Haven region hence consists in its northei'n half of three subordinate north-and-south regions; (1) a narrow valley west of West Rock, drained by West River; (2) a broad central plain (the Hamden plain), continuous with the New Haven plain, rising into hills to the north- ward, and drained along the east side by Mill river ; and (3) a wide eastern portion occupied by the river-course and the extensive meadow lands of the Quinnipiac, in other words, the wide valley of the Quin- nipiac. South of East Rock, the central New Haven plain blends with that of the Quinnipiac. The West Rock ridge to the north throws off a branch on the east which curves around to Mt. Carmel and forms the northern boundary of the central of the three subordinate regions. This central region* is partly subdivided across, on a line, nearly, with West and East Rocks, by two short trap ridges ; Pine Rock, (P) a third of a mile from West Rock, and Mill Rock, (M) which adjoins EastRock ; the width of the interval between the two is nearly a mile. Mill River passes through a deep cut in the Mill Rock ridge, at the vil- lage of Whitneyville. A clear idea of the topography of the reoion 46 J. D. Dana on the origin of some of the is necessary in order to an appreciation of tlie observations that folloAV. 2. General course op Geological events before the Post-tertiary era. One of the last events of the Paleozoic ages was the formation of the Connecticut River valley, by the bending of tlie earth's crust ; and this took place as a sequel to, or in connection with, the crystalli- zation of the granite, gneiss, crystalline schists, and other similar rocks, which make the bottom of the valley. The first fact of the succeeding age, the Reptilian, of which there is record, is the existence of a Connecticut valley estuary, twenty miles or more wide, stretching from New Haven to northern Massachusetts, (New Haven being the proper southern termination of the valley and estuary), and the commencing deposition in this estuary of the Red Sandstone formation. The production of this formation is believed to have taken the whole of the Triassic period, the first period of the Reptilian age, and also part of the next or Jurassic period. After, if not before, the close of the Sandstone era there were erup- tions of trap — a rock that came up melted through wide fissures in the sandstone and subjacent rocks. East and West Rock, Pine Rock, Mill Rock, Mt. Carmel, tlie Meriden Hills, are ridges of trap along with what remains of the old sandstone walls. Tlie sandstone in the vicinity of the dikes, or near any fissures, tlirough which heat and vapor escaped, was more or less hardened by the heat, and rendered comparatively durable ; while other portions were left unhardened or but little so, and therefore in a state admitting of easy erosion and removal. Cotemporaneously with the ejections of trap, veins of cop- per were made, as those of Bristol, Simsbury, Cheshire, etc. ; and veins of barytes, as those of Cheshire. The thickness of the sandstone formation in the New Haven region is not yet ascertained ; in Massachusetts, it is according to the lowest estimate three or four thousand feet. There is abundant evidence that its beds once covered the top of East Rock, now 360 feet in alti- tude, and if so it reached upward to a level which is now at least 400 feet above the sea. Many of the trap ridges to the north in the Connecticut valley were also once topped with sandstone, although much higher than East Rock. West Rock has a height of 400 feet, and the West Rock ridge, between Hamden and Woodbridge, over 500 feet; Mount Carmel about 800 feet; Middletown mountain is 899 feet high ; West Peak, the western summit of the Meriden Hang- ing Hills, 995 feet; Mount Holyoke 985 feet, biit the highest point of the Holyoke ridge, a little farther to the east, 1126 feet; and Mount Topographical Features of the JVew Haven region. 47 Tom, 1211 feet. (The last four altitudes are from Prof. Guyot's measurements.) Although the precise original elevation of the sand- stone about these heights is not certain, there is no doubt of the great increase of height to the north.* This however was not one of the original conditions of the rock, for the beds were made in one com- mon estuary aid nearly to a common level. It has resulted from an uplift which affected the interior of New England more than its south- ern borders ; and the trap also owes much of its greater height to the north to the same uplift. The sandstone mass intersected by dikes of trap constituted the block out of which the future ISTew Haven region was to be carved by various denuding forces. The hard dikes of trap, and the distribution of the hardened sandstone among those feebly hardened, had great influence in guiding the modeling agencies and determining the future features of the country. At the time of the eruptions, or soon after, the land before sul)mer- ged rose above the level of the waters ; rivers took size and direction according to the slopes ; the estuary dwindled into the Connecticut ; and the Connecticut, finding in its way the trap dikes of Weathers- field, Berlm and Meriden, and also elevations of sandstone, took a route, in the latitude of these hills, to the eastward. So the river was lost to New Haven.* Other changes in the old hydrographic ba- sin of the Connecticut valley have taken place since the throwing up of the trap dikes, and part of the following may date from that event, Farmington river, which in Triassic times flowed into the estuary from the western heights of Massachusetts and northern Connecticut, still enters the Farmington region ; but near Farmington it turns abruptly north, flows in that direction sixteen miles, at the foot of Talcott moun- tain and other trap hills of the range, then makes a cut through the range into the Connecticut river valley and joins that river. The Quinnipiac, which starts in the Farmington valley just below the northward bend of the Farmington river, on approaching the region of the trap hills of Cheshire bends eastward out of the valley in front of the Hanging Hills of Meriden, into the valley where the Connec- ticut river might have had its course but for the trap eruptions and disturbances ; and finally, the Farmington valley being thus deserted by the Quinnipiac, Mill river at this point commences its flow, taking its rise in the adjoining hills, • and becomes the principal stream for the rest of the valley southward to New Haven bay. During the Cretaceous period closing the Reptilian age, and the * This view was brought out by the writer la Ward's Life of Percival, p. 420. 48 J. D. Dana on the origin of some of the Tertiary period which opened the JMammalian age, no marine foima- tions were here made ; and there is hence no proof that in the long interval between the origin of tlie trap dikes and the Glacial epoch, the land of the region, or of any part of central New England, was at any time under the sea. Whatever the fact, there must have been, during the time that elapsed, a large amount of denudation over the region ; so that West Rock, Pine Rock, Mill Rock and East Rock finally be- came j^rominent above the plain, altliough much less so than now. 3. General cnARACTER and results op the Post-tertiary period. Next came the Post-tertiary period, the last in Geological history. In order to understand the following remarks it is necessary to bear in mind that the Post-tertiary in America, as the writer has elsewhere shown,* included three eras, corresponding to three great changes of level over the northern portion of the Continent. 1. The Glacial epoch ; when the land stood at a higher level than now, and a universal glacier and a frigid climate covered the continent north of the parallel of 40°, (not a sea with icebergs, as facts about New Haven demonstrate.) 2. The Champlain epoch, an era of subsi- dence ; when there was a sinking of the land below its present level, resulting in a mild climate and a melting of the great glacier ; sub- merging beneath the sea the land along the coast, and giving great extent to lakes and rivers. 3. An epoch of elevation; bringing the land up to its present level, and raising the submerged sea-shore and river flats to a habitable and cultivable height, thus making them available for man. The movements were up — down — up; up for the Glacial era, down for the era following, and up again for the third or finishing era. The origin of the features of the New Haven region cannot be understood without keeping constantly in view these three great movements of the land. In the first of these eras this region stood probably one or two hundred feet above the level of the sea; in the second sixty-five feet or more, and afterward forty and less, below the present level ; and in the third it passed gradually to its present condition. With reference to the question whether icebergs may not have been the agent in the glacial era instead of glaciers, a single argument only need here be brouglit forAvard. Icebergs, as is well known, are frag- ments of glaciers broken off in the sea into which they descend ; and the freight of stones and gravel they bear was received mainly when they were in the glacier condition. The boulders of the Connecticut * Am. Jour. Sci., II, xxii, 325, 346, 1856, and Manual of Geology. Topographical Features of the JSfeio Hnven region, 49 valley if brought by icebergs, should hence have come from the White ^lountains, or perhaps from some Green Mountain peak, for these would have been the only summits above the v^^ater in a sea covering the valley to a depth of four thousand or more feet (the depth that the eyville (V.) Another view with regard to it we regard as much more probable. * Owing to the dam at Wliitneyville, the water of Mill River is not only set back for two miles and more up the valley, but also flows back into Pine Marsh Creek val- ley for njpre than a njile, to within a short distance of Mill Rock (See Map.) Topographical Features of the New Haven region. 55 At the mouth of Pine-Marsh Creek, Mill River takes a bend a little to the eastward of south, while the creek has a course as much to the westward of south, and Mill Rock stands between the extremities of the V thus made by the two channels. In this position of Mill Rock, we find the explanation of the facts referred to. The great glacier having had its ploughing under-surface shaped by the gap west of Mt. Carmel, through which Mill River passes, moved southward, excavating the valley of Mill River, while, at the same time, abrading the soft strata over the hills and plains. The Mill Rock dike, making now a ridge 200 feet in height, stood in its path, the brittle ice confronting the unyielding trap mountain. Under such circumstances, it would have been a natural consequence that at some point north, the brittle ploughshare should have divided, the smaller part to pass toward the Whitney ville opening, by the east end of Mill Rock, and make a shallow furrow because of the hard trap rock under foot at the gap ; the larger part, encountering only the soft sandstone, to plough out the deep broad valley of Pine-Marsh Creek, leading by the west end of Mill Rock and almost directly toward the Beaver Pond region. The question arises whether the excavation was continued into the Beaver Pond basin and thence southward to tlie bay, or whether there was a lifting of the ploughing portion of the glacier through the elevating action of Mill Rock and merely a transfer of the exca- vating pressure to a line more to the westward — the j:)rocess of trans- fer producing the six or eight bays characterizing the eastern side of the Beaver Pond depression and the broad southwesterly surface channels which lead into them. In the former case. Mill River would have run through the Beaver Pond excavation and West Creek ; in the latter, the waters of Pine-Marsh Creek would always have been trib- utary to Mill River in its present position ; for in the Glacial era they would have been those of a sub-glacier stream, and these would have become far moi'e abundant in flow during the melting of the glacier, and thus have made a stream commensurate with the Pine-Marsh Creek valley. There are three objections to the view that Mill River once dis- charged itself through the Beaver Pond Meadows. (].) The Beaver Pond depression is prolonged half a mile north of the point where the Pine-Marsh valley makes its nearest approach to it, and this northern • extremity does not bend toward the valley or show any inclination that way. There is here evidence that the Beaver Pond excavation had its own independent beginning. 56 J. D. Dana on the origin of some of the (2.) If Mill River ouce flowed tlirougli the Beaver Ponds and thence through West Creek to the bay, the force of its waters would have continued to keep this channel open, and West Creek would not have been disjoined from the part above. (3.) If, during the Glacial era. Mill River had had no channel through the Whitneyville gap, it could hardly have afterward gained a foot- hold there where the alluvium has a height of 60 feet or more above mean tide level. There is hence not only no proof of a former connection between Pine-Marsh valley and the Beaver Pond depression, but strong rea- son against it in the condition and character of Mill river and Its present channel. Secondly — The Elevations, or Hills and Ridges made by the Glacier. Besides extensive excavations, there are also elevations which were due to the glacier. They were a consequence mainly of the interrupt- ed series of trap ridges in its way. The hard trap-rock dikes. Mill Rock and East Rock, were fenders both to the sandstone lying on their north- ern side, and also that on the southern, and especially to the latter. The glacier, moving from the north and approaching Pine Rock, would have had its under surface forced up into an arch by the resisting mass, and the ice thus shaped would have been made firm and solid by the pressure ; and as such an arching of the ice below is an arching of the abrading surface of the glacier, an elevation of sandstone correspond- ing to it should have been left by the glacier on its southward mai'ch. An elevation vms tJms left south of Pine Rock — that of the Beaver Hills {Bh.) The Hills are now disjoined from the Rock because of erosion (a) by the waters and ice that descended the slope during the declining Glacial era ; {h) by the waves and marine currents of the subsequent period of submergence in the sea ; (c) by streamlets down the declivities due to the rains and melting snows of later time when the land was elevated to its present level — an era of greater elevation or emergence. It was the eastern abutment of this great Pine-Rock arch that scooped out the Beaver Pond basin. In the same manner the narrow north-and-south Sachem's ridge {Sm^ a mile and a half in length, was evidently made through the lifting action of Mill Rock. Similarly also, the small Cedar Hill, south of East Rock, owes its existence, apparently, to the arch made by the East Rock range ; it is sntall because the East Rock range has a north- and-south direction, or lies with its end toward the moving glacier; and also because the ice of the wide Quinnipiac valley would have pressed westward as it escaped the limits of the valley and passed Topographical Features of the Neio Haven region. 57 the southern extremity of the Rock, and so have swept away the sand- stone there remaining. The great ghieier did not succeed in ploughing out the Mill Rock dike at the Whitney ville notch helow the level of the bottom of the present dam, for the dam is built on the solid trap dike. The ice must therefore have plunged down the front of it (the land having been higher than now), and with it the sub-glacial stream descended. South of this it appears to have made a deep Mill River channel. The glacier acted like the moulding tool in the plough of the car- penter.. But the convexities and concavities on the cutting or abrading edge of the tool were not needed in the pliant material ; for by the fenders placed in its front, in Pine Rock, Mill Rock, and East Rock, the edge was made in these parts to rise or arch upward, and by this means long ridges of various heights were made beween the furrows. The correspondence between the channeling of the plain and the position of the trap ridges is so close (especially if it is considered to what an extent subsequent river and marine action must have tended to modify the features of the surface and obliterate the tracks of the glacier) that there seems to be here visible demonstration of glacier action, and of the insufficiency of the iceberg theory of the drift. If Sachem's ridge, the Beaver Hills and Pine Hill were the only examples of north-and-south sandstone elevations due to hard-rock fenders, the correctness of the explanation offered might be reasona- bly questioned. But they are the least remarkable instances. Over Hamden there are three north-and-south ranges three to four miles long, as exhibited on the map, and they may be distinctly followed northward to elevations in the transverse range of heights west of Mt. Carmel. Cherry Hill (Ch) is the termination of one of these lines. A still more striking example is the Quinnipiac ridge, the dividing ridge be- tween Mill River valley and the Quinnipiac. It stretches from the south side of Mt. Carmel to Whitney Peak, a distance of six miles, and while broad and broken into hills to the north, is to the south an evenly rounded elevation, looking from the summit of Mt. Carmel like a splendid example of landscape grading. According to the theory jiresented, this long ridge of sandstone owes its ex- istence to the arching upward of the ice by the high east-and-west Mt. Carmel range, the ridge being apart of the great sandstone formation left thus elevated in consequence of this arching. The arch, although narrowing somewhat, did not flatten out before reaching Wliitney Peak, as the continuation of the ridge shows ; and here it was raised 58 J. D. Dana on the origin of some of the into a new arch by tins dike, losing in the encounter the red sandstone from the back (or north side) of its head, down nearly one-third way to its base. Either side of this dividing ridge the glaciei*, besides abrading the general surface of the sandstone formation and thereby preparing the rocky basement for the alluvial plains, was ploughing out the river channels adjoining — that of the small Mill River on the west, and that of the broad Quinnipiac on the east. It is a strong confirmation of the view brought forward that the direction of the Quinnipiac ridge, (as well as that of Sachem's ridge,) is S. 12° W. (true course), thus coinciding with the average direction of the Connecti- cut valley, and therefore with that of the movement in the glacier. The largest of the valleys in the Hamden portion of the New Ha- ven region lies along side of the West Rock ridge, where the erosion of the glacier, and of the waters flowing from them would have been greatest in consequence of the height of the rock and its slopes, and where, moreover, erosion from running waters has been going on ever since from the streamlets that the rains and melting snows have made over the long declivities. In this valley lie Wintergreen Lake (due to a recent damming of one of the streams), and farther north the sites of other " contemplated" lakes. This western part of Hamden is drained by Wilmot brook with its tributaries, which flows through the gap between Pine Rock and West Rock and soon after enters West River. The northern poi'tion of the brook, which lies among the sandstone ridges, points southward nearly toward the northern extremity of the Beaver Pond depression, and approaches it within two-thirds of a mile. It might therefore be queried whether Pine Rock had any effect toward dividing the excava- ting action of the glacier on the north, like that from Mill Rock above described. But there is this great difference in the two cases, that the gap between Pine Rock and West Rock is very much broader than the Whitney ville gap, being about a quarter of a mile across, and be- sides there is no continuous pavement of trap at bottom. Moreover Pine Rock has an oblique position with reference to West Rock, its direction being E. 20"" N. true course, (about E. 12° N., compass course,) and owing to the convergence of these two ridges and the broad opening intermediate, and also to the S. 12° W. direction of the glacier movement, the principal part of the excavating portion of the glacier would naturally have passed between them, where Wilmot brook has its actual course. Looking beyond the limits of the New Haven region, still other examples of this north-and-south ridging of the soft sandstone occur. Topograpliical Features of the JSFew Haveii region. 59 South of the Hanging Hills of Meriden .an elevation commences which stretches southward to Mt. Carmel, showing that the ice was arch- ed up by the Meriden mountains, and the arch continued to Mt. Car- mel. And here, as just observed, it was thrown anew into a high arch for the ridging and ploughing southward, in the course of which the Quinnipiac ridge was formed ; then it was raised by Whitney Peak again, and its continuation East Rock; and finally it died out as it left the region of Cedar Hill south of the East Rock range. Besides the large ridges and excavations made by the glacier, the ledges over the hills are often approximately north-and-south in course, and were probably a result of glacier ploughing. The chlorite schist of the Woodbridge plateau is easily torn up in consequence of its slaty structure and its joints or lines of fracture, and also readily reduced to fragments by the freezing of water or growing of veg- etation in the crevices. A large trap dike, intersecting this rock on the Woodbridge heights west of Westville, often stands up above the schist, as a prominent ridge, which sometimes has on one side or the other a bare precipice of forty feet. But much of this wear is undoubtedly the work of subsequent centuries. Without adducing other cases, it appears safe to conclude that over the region of the Connecticut valley the principal part of the coarse gouging out of the plains, and shaping of the mountains and valleys, were performed by glaciers and by the streams that were in action during the progressing and declining Glacial era. The same agents also carried southward the earth, sand and gravel that were afterward to be deposited by the ice, and worked over by the rivers, or, near the sea-shore by the rivers, tidal currents and waves, into ter- raced " alluvial " plains, or stratified drift foimations. Scratches having the course S. 33° W. — A wide variation from the usual course of the glacier scratches (South, to S. 12° W.) occurs over the chlorite rock along the Milford turnpike half a mile to a mUe west of Allingtown. The place is about two and a half miles south of West Rock, and one and a half miles south of the line of East Rock. The course (true) of the scratches is quite uniformly S. 33° W., or full 20° west of the usual direction ; and they are so deep and numerous and so completely free from crossings by scratches in any other direction, that S. 33° W. must be viewed as the course of the under surface of the glacier over this part of the western margin of the New Haven region. The scratches are seen at the top of the first ascent on the turnpike, about 1 30 feet above the sea, (or 90 above the level of the New Haven plain), and at many other points 60 J. D. Dana on the origin of some of the where the rock has been recently exposed, for half a mile west. The ledges that have been long bare have lost their scratches by weather- ing ; on this account, and owing also to the covering of soil over other parts, observations have not yet been extended farther west. The following is offered in explanation of this southwestern throw of the under portion of the glacier. It has been stated on page 45 that the New Haven region, be- tween the summits of the ridges confining it on the east and west, has a width of seven miles to the north, and narrows to four at the south. While the mass of the glacier was continuing its southward movement, the portion below filling this depression would have had to accommodate itself some way to the narrowing limits. This ac- commodation might have taken place, through an increasing depth of the depression southward. But if this was insufficient to meet the whole, there would have been a tendency to a thickening upward of the glacier and relief would have been obtained from the accumula- ting pressure by a lateral escape of the ice. There was evidently no yielding or escape on the east or Quinni- piac side, the side of the broadest and deepest valley, and therefore of deepest or of thickest ice ; for the ploughings of the glacier which are exhibited along that side on a grand scale over the East Haven sandstone, have the usual southward (S. 13° W.) course. Hence the escape, if any where, must have been on the west side; and here it is that we find these S. 33° W. scratches. The place is southwest of where the Quinuipiac valley opens on the New Haven plain, and con- sequently it is situated just where such an effect from the expansion and pushing action of this part of the glacier would be produced. Now to the vjest of the region of these scratches within three-fourths of a mile, there is the rather broad valley of Cove river, which ex- tends southward and reaches the Sound two and a half miles below ; it is pai'allel nearly with the New Haven region, but has a much steeper slope, the descent to the salt water flats being at the average rate of about 125 feet in a mile. This slope of the valley would have given the ice that tille 1 it (the under portion of the glacier, if not the whole above) relatively a rapid movement. The overflow from the New Haven depression caused by the conditions stated would there- fore have naturally taken a course into this valley. The direction of the scratches, S. 33° W. accords well with this view. Making of Lake-basins. — The lifting of the lower or abrading sur- face of the glacier by hard rocks, which has been shown to have re- sulted in the production of the norfch-and-south ridges, and which ap- Topograpliical Features of the New Haven region. 61 pears to have terminaterl southward the basin of Pine-Marsh Creek, might under other circumstances have made basins for lakes. Lake Sal- tonstall, four miles east of New Haven, probably owes its existence to this action. The lake is 3.^ miles long and has an average breadth of a third of a mile. The basin is scooped out of a very soft, crumbl ing shaly sandstone, and lies between two bow-shaped trap dikes, three- fourths of a mile apart, whose average trend is north-northeast. Its depth is stated at 112 feet ; and since its surface is only half a dozen feet above high-tide level, the bottom is more than 100 feet below that level. At the present outlet the waters flow over solid trap at a low cut in the western trap ridge, so that the basin is here rock-bound on the south. The stream from the lake (called Stoney River, but properly the lower part of Farm River), flows for its last mile be- tween granite shores and has in some places a rocky bottom. Thus there is a granite as well as a trap barrier between the lake and the sea, and the depression it occupies is a true basin. We may believe therefore that the long narrow basin occupied by the lake is an excava- tion made in the soft sandstone by the ploughing glacier, and that it was not continued to the sea because the ploughshare was lifted out of its trench by the hard unyielding rock before it. Height of the Land in the Glacial era. — With regard to the height of this portion of Connecticut above the sea in the Glacial era we have as yet few facts for definite conclusions. a. In sinking an artesian well on Green st., 120 yards from the harbor, a bed of fine clay 14 feet thick was struck at a depth of 140 feet, or 126 feet below mean tide level. Above this clay there were the ordinary sand or gravel deposits of the New Haven plain. The clay bed was evidently a mud deposit made in the harbor as it existed immediately before the deposition of the sand ; and as the sand beds of the New Haven plain date from the era following the Glacial, the harbor very probably was that of the Glacial era. If the land then stood 125 feet above the present level, the mud bed would have lain just at the water's surface, like those of the present day. The evidence as to the level of the land in the Glacial era is uncer- tain; still it affords a presum2:)tion that it was at least 125 feet higher than now. No clay has hitherto been found in any other part of the New Haven plain. h. Near Stoney Creek, eleven miles east of New Haven, on Smith's Island, one of the " Thimbles," there are two pot holes in the hard gneiss rock; oiie of them is 1\ feet deep, and 3 in diameter, and the other 3 feet deep and 10 inches across. They are situated within a 62 J. D. Dana on the origin of some of the few yards of one another upon the coast, but above lugh tide level. The large one contained, when recently opened by Mr. Frank Smith, its discoverer, many large rounded stones. Another pot hole of less depth exists upon Pot Island, about a mile to the southeast of Smith's Island. It is like a bread-trough in shape, and is 4 and 2 feet in its diameters, and 1^ feet deep. Still another, as I am informed, occurs on Rogers' Island, one of the westernmost of the same group. It is within reach of the tides and is 4 feet deep and 2 in width. These pot holes must have been made by torrents from the land. For the existence of such torrents the land should have been above its pre- sent elevation. We cannot fix positively the era of this higher level, but it may have been that of the great glacier, and the torrents, sub-glacier streams then existing. c. The valleys of the streams of Connecticut and even those of the north side of Long Island are in general continued over the bot- tom of the Sound beneath its waters, apparently excavated for the most part out of the sand and mud deposits which constitute it; and this fact appears to indicate that the Sound was once dry land — a great east-and-west depression of the surface — into which the streams of the adjoining country flowed, and there concentrated their waters in a grand central river which received the existing Connecticut a few miles before enteiing the Atlantic. The admirable chart of the Sound by the TJ. S. Coast Survey, which is covered with figures in- dicating the soundings, enables any one interested in the subject to draw the lines of equal depth, and verify this statement.* There is nothing in the depth of the Sound to render the above supposition incredible. An elevation of 100 feet would now lay bare all but a fifth of its bottom across from New Haven, and one of 140 feet the whole breadth ; and one of 200, would dry it up all the way to the line of New London, 50 miles east of New Haven. Further, a rise of even 50 feet would wholly separate the narrow western portion of the Sound from the more eastern by a bare area in the meridian of Marm- aroneck and Rye, or 50 miles west of New Haven. Only the broader depressions corresponding to the courses of streams are to be looked for over the bottom, even with the fullest possible series of sound- * It is best, in order to exhibit well on the map the curve of the deeper and shal- lower parts of the Sound, to draw the hnes for each fathom of depth up to 8 fath- oms, and then for every two fothoms, that is for 10, 12, 14 and so on; and in addition, to make the lines for 7, 18 and 24 fathoms much heavier than the others; and to use differently colored inks for the lines 4 to 8 fathoms, 10 to 22, and 24 and beyond; or else give the areas 3 to 8 fathoms, 8 to 24, and over 24, different shades of color. Topographical Features of the New Haven region. 63 iugs. For like all New England, the Sound received vast deposits of gravel and sand in the Champlain era from the depositions of the great glacier ; and ever since these depositions were made, the riv- ers have been carrying in detritus, each year making its large con- tributions ; the estimate, therefore, that the original surface, as it was before the Glacial era, had been covered by all these deposits to an average depth of 50 feet, cannot be excessive. After such a process tending to obliterate all depressions, especially over the north- em half of the Sound which has received the most of the detritus, it is certainly obvious that better defined river channels than exist are not to be expected. But the conclusion from the existing channels above suggested has at least three sources of doubt — one arising from the present action of tidal currents ; a second, from outflowing under currents which oc- cur at times in connection with large bays ; and a third, from the con- figuration of the rocky basement beneath the mud and sand of the bottom of the Sound, (1.) Jutting capes, especially if prolonged far out beneath the wa- ter, as well as obstructing shoals or reefs, inasmuch as they narrow the Sound, give increased velocity to the tidal currents passing by them. This cause is sufficient to account for the large deep holes — 30 to 33 fathoms — opposite Norwalk, where " Eaton's Neck" on the Long Island side makes a long projection into the Sound beneath its wa- ters, which projection at its extremity, three miles out (and hence nearly half across this part of the Sound), close along side of the deep holes, is within 6 fathoms of the surface. Again, near the " Middle Ground," south of the mouth of the Housatonic, or of Stratford, a large shoal but 2 feet deep in one part, there are deep holes both off its northern and southern extremities, the former of 20 to 21|^ fathoms and the latter of 20 to 27^ fathoms; and they are in part at least an obvious consequence of the tidal currents sweeping by. Ten miles west of the mouth of the Connecticut, the Sound com- mences to narrow toward its eastern termination, its southern side here bending up to the northeastward ; moreover shoals made from Connecticut river detritus, contract the breadth on the north. Conse- quently, here begin two depressions, and half a dozen miles east, a third on the north, which three unite in one broad range of deeper water, 18 to 32 fathoms in depth, that continues eastward, and finally increases to 50 fathoms as the waters approach the channel, called " The Race," by which they leave the Sound and enter the Atlantic. 64 J. D. Dana on the origin of some of the (2.) The outfloxoing iinder-currents of bays are produced, especially when the broad opening has a comparatively narrow principal chan- nel with other passages, among or over reefs ; and they are strongest when the waves and currents occasioned by a storm drive heavily to- ward and into the bay ; and still more so if a river add its floods to the waters which the storm waves and cm-rents pile up within the bay. I do not know of any observations about the bays on the Sound tending to show where such under-currents exist, or what in any par- ticular bay is their force or direction ; and we are at a loss as to the effects to be attributed to this cause, (3.) The actual configuration of the rocky substratum of the great basin in which the waters of the Sound rest is also little understood. Long Island has no rocks at surface, or about its points ; and the Sound east of Hurl Gate, except quite near its shores, is also without any projecting rocks. Some of the prominent sand-spits of the shores, as those of New Haven and Stratford Point, may be traced far southward by means of the soundings. But it is not always easy to decide whether they have resulted solely from the detritus of the riv- ers to the west of whose mouths they lie, or whether a rocky base- ment has determined the form of the projecting spits. On the sand bed ofi" the west point of New Haven harbor there are sur- faces of bare rock, giving evidence of a rocky basement. Off Stratford Point, west of the mouth of the Housatonic, soundings have discovered no such rocks ; and yet it is probable that the form of the bottom is here determined by the rocks underneath. On Eaton's Point the map says " rocky " at one spot ; and the existence of this spit may also have been determined by the rocky basement below. But even when the spits or projecting sand-bars are proved to cover a ridge of rocks, it is not certain that this ridge may not have been a result of the excavations of the glacier, and of sub-glacier streams. The shoals and deep holes in the vicinity of " Eaton's Neck " are directly south of the mouth of Norwalk river, and those about " Mid- dle Ground " are south of the mouth of the Housatonic ; and the question arises : Were they partly made by the rivers when the land was more elevated, or may they have been determined solely by the rocky configuration beneath and existing currents ? It is apparent that without some direct investigations our conclusions can only be uncertain probabilities. Yet notwithstanding all the doubts from the above mentioned sour- ces, there are so many examples of depressions leading from the bays at the mouths of rivers over the bottom of the Sound, so many in Topograijhical Features of the JVeic Haven region, 65 which the outflowing under-currents of bays appeal' to be insiiificient to account for the facts, either because the bay is not of the shape to produce appreciably such an eflfect, or there is not in the currents the proper accordance with the ebb in direction, that we think the facts afford strong evidence in favor of a former elevation of the region — an elevation probably not less than 1 50 feet. In such a case Long Island would have been literally the southern border of New Eng- land, and the universal glacier would have had no great basin of salt water to span in order to reach what is now the Island, and deposit there the boulders of Connecticut rocks, some of which, according to Prof. Mather, are from 500 to 1000 tons in weight.* * It is difficult to explain the facts in detail with regard to the Sound without a map at hand. The following observations on the subject are however here added. The main course of deep water through the Sound west of the meridian of Guilford commences near the northern shore of the Sound, off Coscob harbor and Greenwich Cove, (near the boundary between Connecticut and New Tork), and just here enter Bjram, Mianus and Turn rivers. From this region it stretches eastward, passes the north point of the Eaton Neck spit, leaves " Middle Ground " to the north (and consequently in this part is south of the middle of the Sound), and then continues directly eastward till it almost touches the north coast of Long Island (being less than a mile off) in the line of Guilford. At the very end of the deep water channel the depth is 18f fathoms; just east of it, the depth is only 11^, then 10 and 9 fathoms. But about 6^ miles a httle to the north of east, about two from the shore of Long Island there is an oblong deep hole 18 to 19 fathoms in depth; and 2^ miles beyond, in the same direction, commences the southern arm of the great central range of deep water which continues eastward out of the Sound. The great range of deep water, seventy miles long, that commences m the west near Greenwich, must, as already observed, owe something of its depth, in its eastern portion at least, to its distance from the northern shore of the Sound or the region of rivers and detritus ; and, again, it may have had its course determined origi- nally by an east-and-west depression in the configuration of the basement rocks of the Sound. Still it affords some reason for believing that it once contained the channel of a great river. It begins against the north shore near Greenwich, just where three streams enter the Sound, as if a continuation of their united channels. Its depth at its eastern extremity, and its abrupt termination there, are reasons for inferring that it once continued stiU farther east, and was probably kept open by a flow of water through it. If the land were formerly higher by 150 feet, as has been supposed, the re- quired conditions would have existed for making it a river course. But the query comes up, where in that case would have been the discharge? Its abrupt eastern ter- mination takes place right opposite the large and broad Peconic bay which divides the eastern end of Long Island for a distance of nearly 20 miles, making the Island in form like the profile of an alligator, with its long mouth (Peconic bay) wide open ; and the interval of dry land between the Sound and this bay is hardly three miles wide. Moreover, directly in the line of the depression, the land is low, and is intersected by Matituck lake, and also by various channels on the Peconic side. These facts lead to the supposition that this Sound stream of the Glacial era, whose tributaries included Trans. Connecticut Acad., Vol. II. 5 Sept., 1869. 66 J. D. Dana on the origin of some of the 5. Events and Results of the Champlain era. The Glacial era closed in a subsidence of the land over a large part of the continent, the initiatory event of the next or Champlain era. 1. Amount of Subsidence. — The amount of the subsidence about New Haven is uncertain, because the actual height of the land in the Glacial era is not yet satisfactorily determined. It was so great as to carry the land considerably below its present level, as evinced by the height of the New Haven plain, this plain having been made and lev- eled off in the waters of the era. Taking the level of this plain as marking the water level, we learn that about the College square and for some distance to the north, and either side of this region on the same east-and-west line, the depression was near 40 feet. Farther to the north it increased gradually to 70 feet and more in Hamden ; w^hile the Housatonic and other rivers to the west, may have discharged through an open- ing into Peconic bay. and that this opening was filled up by sands during the following era of submergence (the Champlain era), and cotemporaneously the adjoining southern portion of the Sound was made shallow by the same means. The form of the bottom in this part of the Sound favors the idea that the sands for fiUing it came from the di- rection of the Peconic bay. But the existence of the oblong deep hole in the course of a direct line to the south- ern arm of the great eastern deep water region of the Sound hardly nine miles distant, brings up the enquiry whether the river channel may not after all have been over this route within the Sound. The submergence of the Champlain era would have afforded the same means as stated above for filling up with sands this part of the Sound and for stopping off abruptly not only the channel of the Sound river, but the great depres- sion in which the channel lay ; for the waves of that era must have swept across the land in one or more places from the Peconic bay into the Sound. If this latter view is the right one, the great Sound river, commencing in the rivers of the vicinity of Greenwich and taking into itself the waters of other rivers eastward to the Housatonic, and still others from Long Island, would, after receiving the Housa- tonic, have derived little else du'ectly from the north until reaching what is now the eastern deep-water region ; and this it would enter by the southern arm of that region. The rivers of the New Haven coast and other small streams between it and Sachem's Head, would have taken an intermediate course over the Sound to the same meridian, and then entered the middle arm. The rivers from Guilford to Killingworth harbor would have flowed eastward to the commencement of the northern of the three arras. And then a few miles beyond this, the northern arm would have received the Connec- ticut river, the great tributary, and from this point all the fresh waters of the various rivers would have been combined in one grand flow on their way to the ocean. From the depth of water and the character of the deep holes over the deep-water region south of the Connecticut, it may be inferred that here was actually the great bay of the Sound river into which the ocean waves set as they do now into the mouth of the present Connecticut. The latter has its deep holes inside of its bar; for the depth within the channel of the present river at low tide is 6 to 7 fathoms, while there are but 10 feet of water over the bar. Topographical features of the Kew Haven region. 67 to the south it diminished in height, being but 30 feet in tlie latitude of Halleck's place on the bay. The facts on this point are given be- yond (p. 88). North of Connecticut, over New England, the amount of depression below the present level was still greater, and increas- ingly so with increase of latitude, it having beeri 200 to 250 feet at least in central New Hampshire, 400 about Lake Champlain, and 500 feet on the St. Lawrence. 2. General consequences of the Subsidence. — As the writer has re- marked upon elsewhere, an immediate consequence of a subsidence of the land, and especially of one which was greatest as a general thing to the north, would have been the bringing on of a warmer climate, and thence, the commencement of melting in the glacier. As another result we note that the slope of the great valley of the Connecticut would have become less than it is now. Consequently the motion of the Connecticut valley glacier would have been greatly retarded, if not rendered altogether null. Moreover the rivers would have had a diminished rate of flow, and would therefore have spread in wider floods than ever before, becoming in some parts a series of lakes ; and the lakes also Avould have had an unwonted expansion. The great flow of waters from the melting ice would have immensely augmented the floods in all directions. Such an extended change of climate over the glacier area was equivalent in effect to a transfer of the great glacier from a cold icy region to that of a temperate climate and melting sun. The melting would therefore have gone forward over vast surfaces at once, wide in latitude as well as longitude, and not merely along a southern edge with slow creeping progress northward. Hence, as another result, the depositions of sand, gravel and stones from the glacier, would have taken place almost simultaneously over regions scores of miles wide in latitude, and in general without special accumulations along a southern border like what is called the terminal moraines in the Alps. They would have descended alike over the hills, plains, and valleys, lake regions, flooded rivers and sea-shoi-e bays ; but not with like results over these various regions, for wherever there was water in motion beneath, the water would have worked over the pebbles and sand and produced some stratification of the material, or at least have leveled all off at top. Thus unstratified and stratified drift (the latter including the so-called modified drift, as well as a large part of the "alluvium" of river valleys) were formed simultaneously, and both in the Champlain era. The depositions made directly from the glacier as a consequence 68 J. D. Dana on the origin of some of the of its melting, and which belong to the oj^ening part of the Champlain era may be first considered ; and afterward the secondary and later results. 1. Events and results of the Opening Champlain era. 1. Depositions over the hills. — The deposits over the hills in the New Haven region, consist, like those of the rest of New England, of sand, stones, and large boulders, mingled pell-mell, or without strati- fication, except wliere they fell into lakes and rivers. This unstratified drift is found wherever the land rises above the level of the stratified " alluvium " of the New Haven plain, except along the upland valleys. In some places it appears to be more or less stratified, as near the Seymour turnpike (running west from Westville) after passing the first hill (that of the Edgewood line) ; but in this and other similar cases the stratification is owing to the fact that the region in the time of the melting glacier was the course of a flooded stream. The boulders and stones are not to be looked upon as lying just where they were dropped in all cases, nor as being formerly in the same large numbers as now over given areas ; for the sands and smaller stones that fell with the larger masses have to a great extent been washed away to lower levels, and carried off by streamlets to rivers, and by rivers seaward, and thus the large stones that cx-owd the surface in some regions may when first dropped, have been many feet apart, or even scores of feet away from the spot where they now lie. The character of the stones and the size of the boulders over the hills show what is the nature of much of the material which fell into the waters, and which now lies over what was the bottom of the bay in the Glacial era. The larger boulders of the New Haven region consist mostly of trap and sandstone ; and next to these in size and numbers are those of gneiss and quartz. Those of trap, sandstone and gneiss are quite numerous over the western border of the region, especially along the eastern margin of the Woodbridge plateau ; those of quartz rock have a very wide distribution. Only a few of gneiss have been observed as far east as Sachem's ridge. Some of the largest of the trap boulders are as follows : One 2^ m. north of Westville, on Boulder Hill, measuring along its diameters 29, 14 and 12 ft. and weighing at least 400 tons. The boulder in pieces making the Judges' cave (the place of con- cealment of the regicide judges for a while in 1661), If m. east of south of the preceding, on the top of the West Rock ridge, the masses when together having weighed at least 1000 tons. Topographical Features of the JVew Haven region. 69 One on the Woodbridge heights, 1^ m. southwest, about 10 feet in its diameters, but now in halves. One in the northern part of the Edgewood grounds, a mile southeast of the last, and 2 m. a little west of south from the Judges' cave, about 8 feet cube. Three others, half a mile south of the last, in the same grounds, measuring 25, 18^ and 8^ feet, 14, 8^ and 7 feet, 8, 5 and 4 feet. One near the Derby turnpike, -^m. E. of S. of the last, of 14, 6 and 5 feet in its diameters. One in the woods north of the Stoeckel farm, \ m. S.W. of the last, and in the same line nearly with the Judges' cave and the Edgewood boulders. On the Milford turnpike nearly a mile east of south of the latter, \ m. west from Allingtown, measuring IJ), 8 and 5 feet. One at Savin Rock, farther sovith, 8, 6 and 4 feet. These masses are all on the western border of the New Haven re- gion. The height given in each case is the height above ground, the depth to which the boulder extends below the surface being uncertain. Many of those that formerly lay over these heights have been bi-oken up for use in house-building. Over the same region sandstone boulders are numerous, but they are seldom very large, owing to the nature of the rock. One of tabular form on Boulder Hill measures 21, 15 and 5 feet. There ai-e also large trap boulders more to the eastward. One on Sachem's ridge measures 16 feet in length and 8^ in gi'eatest breadth; and one in East Haven, back of Mr. Woodward's, of 11, 9 and 6 feet. There are also occasional masses of native copper derived from the copper mines of the Connecticut trap and sandstone region. A mass from the vicinity of East Rock, given to the Yale Cabinet by Mr. Eli W. Blake is probably of this kind. Another weighing 90 lbs. was found early in the century on the Hamden Hills. 2. Depositions over the loaters. — The New Haven bay in the Cham- plain era covered the whole breadth of the New Haven region, from the Woodbridge range on the west to the sandstone ridges of East Haven and North Haven on the east, and spread northward into Ham- den. East and West Rocks, Pine Rock and Mill Rock were cliffs within its area, or on its borders. Sachem's ridge was a long north- and-south peninsula south of Mill rock : and the Beaver Hills, another south of Pine Rock. The Beaver Pond region was, for a while at least, the deep central portion of the New Haven bay ; it lay in the interval between Mill Rock and Sachem's Ridge on one side, and Pine 70 J. D. Dana on the origin of some of the Rock and the Beaver Hills on the other, close alongside of the latter. Mill River entered a narrow arm of the bay between East Rock and Sachem's ridge, and the waves widened its head and battered Mill Rock for some distance west of Whitneyville. This Mill River arm was encumbei'ed by two or three low sandstone islands, the northern- most of which is now the site of the residence of Stephen Whitney. West River opened into another ai'm which lay between the eastern of the Woodbridge heights (or the Edgewood range of hills) and the Beaver Hills, and West Rock Cliff and Pine Rock overlooked it on the north. Up the Quinnipiac valley, beyond East Rock, stretched a long and broad arm of the bay, which was the great inner harbor. We come now to the consideration of the action of the waters of the bay in arranging the material dropped into them by the melting glacier. The large boulders were evidently the first to fall ; for none were found on the plain when it was first taken possession of by the colonists, although such masses were then very numerous over the low Beaver Hills and wSachem's ridge, and are somewhat so still notwith- standing man's free use of them. Furtlier, in no excavations into the alluvium of the plain for cellars, wells, or other purposes, (as we are informed by Messrs. Perkins & Chatfield, Mr. Isaac Thomson and Mr. D. W. Buckingham, who have superintended such work for years past) have boulders anywhere been found, with only two exceptions ; and these are really no exceptions, since the boulders in each case lay on the foot slopes of sandstone ridges. One occurred at a depth of 10 feet beneath the gravel of the alluvium, and was found while making a pit in Trumbull St., near the house of Prof. Fisher; it was of trap and about two feet across. In the other case a number of large stones were met with in digging a well on Whalley Avenue near Blake Street ; Mr. Buckingham, who reported tlie fact to me, attributes their occur- rence there to the nearness of the place to the Beaver Hills. As the melting went forward, the sand, pebbles and cobble stones were thrown down together; but they underwent as they fell an aiTange" raent which varied according to the movements in the waters beneath. The bay had its tidal currents, as now ; its areas of comparatively still waters ; and besides, certain channels along which the flow of the rivers increased greatly the force of the ebbing tide. The strati- fication of the deposits varied accordingly. Where the cun-ents were strong, they washed away the sand from the stones, or if very strong, the sand and smaller pebbles, and thus layers of coarse gravel were made — gravel beds being always deposits from which the sand has been sifted out by moving or flowing water. Along the main river courses there ought to be found, consequently, long gravel courses^ Topographical Features of the New Haven region. 71 marking the direction of the strongest currents, and these gravel courses should be not far below the surface unless the depth of water in which they were deposited were too great for this. Accordingly, we find that the valley of West River, near West Rock is a pebbly and stony region. Another more remarkable gravel course extends from the head of the harbor between Meadow and Franklin streets over State and Orange streets, toward and beyond Whitneyville, and this was evi- dently the course of the Mill Miner channel. It follows (see map) the west side of Mill River from Whitneyville down to Grand street, then diverges a little westward, the I'egion between Mill River and Franklin. Street, as I am informed by Mr. Chatfield, being less stoney than that to the west. Franklin street is about 500 yards from the river. At N'eck Bridge, below the East Rock range, the " alluvium " on the west side of 3Iill River is four-fifths stones ; and on the east it is very pebbly, but the proportion of stones to sand is not more than 1^ to 5 ; and farther east the proportion of pebbles becomes quite small. The gi'avel is in all parts exceedingly coarse, and consists largely of cobble stones. This gravel course extends far up Mill Riv- er, and is as coarse in its stones n6ar Ives' Station 4^ miles to the north, as it is over the New Haven region. Just south of the Mt. Carmel gap, the stoney character is still more remarkable. Another gravel course, but coalescent with the preceding as it approaches the bay, passes northward along the Canal railroad to the west of Sachem's ridge [Sni]^ instead of to the east of it. It has the course of the East Greek valley. The pebbley deposits or gravel underlie the surface from the head of the bay, northward across State street; its western border follows approximately (as I learn from Messrs. Perkins & Chatfield) a line along State street to Crown, across from this point to the corner of Chapel and Church ; along Church from the corner of Church and Wall to the corner of Grove and Temple ; and thence along the east side of the cemetery. The extent of the region shows that the flow producing it had the breadth and character of a tidal flow. This East Creek tidal channel was connected directly with the central interior basin of the harbor, the Beaver Pond depression, as the channels in the surface along Webster and Munson Streets demonstrate (p. 53, 54). The Mill River and East Creek tidal courses were branches of the great central tidal flow up the bay. The gravel-course of the Quinnipiac is not in sight. This inner harbor of the bay was deep, and swallowed a vast amount of great stones, gravel and sand, without being filled to the sui-face. 72 J. D. Dana on the origin of some of the The courses of the tidal currents of the bay are also apparent in the less height of the drift formation wherever they swept along. Thus, within the range of the Mill River tidal course, at Neck Bridge, the heiglit of the terrace on the loest side of the river is but 32 feet, while it is 42 feet on the east side. This Mill River tidal current, although strongest, perhaps, to the west of Franklin street, had a wide spread to the eastward. For the area over which the terrace formation is below its normal height includes not only a region west of Franklin street, but all east of it to the river, and even a large part of Grape Vine Point (the wide point of land between Mill River and the Quinnipiac). Near the bridge at the foot of Chapel street on this Point the height of the drift or terrace formation is only 12 feet, and between this and the southern extremity of the Point, it is still less ; half across the Point in the same line, it is only 21 feet: and half a mile to the north, near the Barnesville or Grand street bridge (the sec- ond bridge over Mill River), the height is only 29 feet. On the Quin- nipiac side of Grape Vine Point, on the contrary, the plain has its full height, being 34 feet in the same east-and-west line with the Chapel street bridge, and 40 feet in that with the Grand street bridge. It is evident therefore that the central part of the great tidal wave up the bay in tbe Champlain era swept northward between Meadow street on the west and Ferry street in Fair Haven (on Grape Vine Point) on the east, an area over 1^ miles wide ; that it continued to be felt on the east side of the river to the north of Barnesville bridge ; but at Neck bridge, approaching the south point of the East Rock range, it was pushed more to the westtcard, the terrace on the east bank at this point having a height of 42 feet, or the full normal elevation. An eastern branch of the tidal wave entered the Quinnipiac basin through the broad channel which forms the lower part of this river. Owing to the bend to the westward in the lower part of this channel, the wave was thrown against the eastern shore, so that the terrace forma- tion on that side is mostly wanting while built up neai'ly to its full height apparently on the western side of the channel even quite to its mouth. By closely studying the nature of the stratification of these deposits beneath the New Haven plain, the particular character of the action of the waters may generally be made out, even, in some cases, to dis- tinguishing the effects of individual waves and changes in the action of tidal or river currents. A good example of this is afforded in the region south of the East Rock range (or of Snake Rock, its southern termination) between jNlill River and the Quinnipiac, where sections of the deposits have been made in grading for the Hartford and Air-Line Topographical Features of the New Haven region. n railroads (see the course of these raUroads between INIill River and the Qninnipiac on tlie map). The whole height of the alluvium above mean tide is in this region from 42 to 45 feet. The cnt through it for the rail- roads extends nearly southwest and northeast, and is about two-thirds of a mile, or 1200 yards, long. After the first 700 yards, the railroads pass under a bridge, and just beyond, the separate cut for the Air- Line railroad commences. The depth of the section is about 16 feet at its Mill River end, 20 at the bridge, and 26 toward the upper or Qninnipiac end. A number of interesting facts are to be observed in the sections : a. The diminution in the proportion of pebbles on passing east from the ]Mill River valley is well seen. Along the Air-Line road they con- stitute hardly a fifteenth of the whole mass, although in an occasional small layer they are of large size, even like cobble stones. Toward the more northern or Qninnipiac end of the cut, the layers are not only less pebbly but the lower part of the section contains two to four irregular layers of exceedingly fine clayey sand (M, fig, 2). The material adheres rather firmly, holds water well, and is so damp at all times that the exposed surface has in part become green from a covering of moss. The clayey layers are separated by others of sand, and an occasional one of pebbles. 1. h. The alluvium is iu nearly horizontal layers, just as it was origi- nally laid down. But these layers are quite irregular, often of small lateral extent, and where composed of sand are very commonly made up of wave-like parts, from two to many yards long, as in the annexed figure — which represents a part of the surface six feet in height, about half way from top to bottom in the Air-line railroad cut. c. A marked variation from horizontality occurs at the northern or Quinnipiac end of the cut, where the layers, as shown very distinctly in the firmer beds of clayey sand (fig. 2), dip downward four feet in a length of 30 feet, or from ^\ to ?<\ feet above the railroad. This dip is toward the Quhmipiac river, or toward the old harbor, and may have some relation to the orioinal bottom of the basin. 74 J. D. Dana on the origin of some of the d. The sand of the sandy layers is obliquely laminated (not an uncommon fact in such deposits), as shown in the above figure. This division into tliin oblique layers is not apparent when a cut is first made through it with the spade, but appears often on a surface of nat- ural fracture, and very distinctly after exposure for a while to the vdnds. It shows that the sands were deposited in these delicate oblique layers while they were being accummulated in the long or horizontal beds which consist of these. e. Throughout the upper part of the section, (above the line N S fig. 1), the inclination in the oblique lamination is mostly to the north- ward: while in the lower part (below NS) it is as generally to the southward. Layers containing both slopes may be observed in each, but the above are the prevalent courses. The formation is thus made up of an upper and lower division ; and in many parts the two are separated by a thin band of large and small pebbles. Near the junc- tion of the Air-line and Hartford tracks, the dividing plain is but two feet above the level of the railroad ; to the eastward it retains nearly the same level (about 22 feet above meantide), but it is higher above the track, there being here a descending grade in the road. Below this bridge, or toward Mill river, the upper division is the only one in sight ; the level of the dividing plain passes beneath the surface of the railroad excavation, and for this reason cannot be traced. For the first two hundred yards on the way toward Mill river, the slope of the oblique lamination rises quite uniformly to the southward as in the upper division above the bridge ; through the next one hundred yards, this is still the prevalent direction ; but farther toward the Mill river end of the cut both slopes occur, and that of the lower division finally becomes the most common. f In the upper part, the sands, through the cut for the Air-line road, have the ordinary dirt-brown color ; in the lower part they are hrownish-red. Thus there is a marked distinction between the two divisions in color, as well as in lamination. This color is of course not observable in the pebbly layers. It is owing mostly to the fact that the grains of quartz are tinged outside by red oxyd of iron, like those of the red sandstone. The following conclusions flow from the facts here noted. (1.) It has already been observed tliat Mill river valley, especially its west side, was the course of a powerful tidal current which set in and out over what is the head of the j^resent harbor, whose ebb was increased by the flow of the river. From the diminution in the amount of pebbles to the eastward of the river (§ a), it appears that the tidal Topographical Features of the New Saren region. 75 flow, as it spread in that direction around what was Snake Rock head- land, rapidly lost its force ; and finally, when fairly in the Quinnipiac basin, as the beds of fine clayey sand show, there were intervals of comparative quiet or of only gentle movements. The fact of these gen- tle movements is proved not merely by the fineness of these beds, but also by a very delicate contorted lamination in tliem, which in some places looks as if due to the smallest of eddyings in the water at the time of deposition ; and also by successions of obliquely-laminated layers of sand only one or two inches thick, constituting here and there an overlying bed. Where layers of stones, or thick obliquely-lamin. ated sand-beds, exist between these clayey beds, they indicate that a time of rougher movements intervened. (2.) Since the slope in the oblique lamination throughout the lower division of the alluvium dips to the southward, or rises to the north- ward (§ e), the deposition of these beds took place under the action of a tidal current flowing northward, that is, into the old Quinnipiac har- bor ; and the reverse direction of the lamination in the upper division implies a current during its formation to the southward, <:«crt?/ /rom the old harbor, or toward the present hay. Such a change of current (A) would have attended the flow and ebb of each tide. But this cause of the transition in the beds would make the whole deposition a twelve-hour operation ; which, even with a melting glacier above to supply material, would have been incredi- bly quick work. It might (B) have proceeded from a change in the place of discharge of the Quinnipiac waters, such as would have added the river current to the ebbing tide. But there is no evidence in favor of this in the existence of an old channel, and much against it, in the character of the layers along the present channel north of Fair Haven. It might (C) have resulted from the setting in of an ex- traordinary river flood, giving great force and volume to the out- flowing tide, and not only along the proper channel of the stream, but far and wide over the low lands adjoining. Through such means the action of the incoming tide would have been as much weakened as that of the ebb enhanced ; and, as a consequence, the oblique lam. ination of the sands would have been produced by the outflowing flood The special influence of the Quinnipiac flood would have dimin- ished westward, where finally it would have encountered a similar though smaller. Mill river flood; and hence it is natural that the alluvium should here lose its Quinnipiac characteristic and take that of the other stream, as stated in the closing part of (§ e). It might (D), if a flood were in progress, have been due to the fact that the 76 J. D. Dana on the origin of some of the depositions had reached such a level as to impede the inflowing tide, and thereby give the ascendancy to the river current ; this being favored by the height of the land at the time. Of the causes suggested above for the diiference between the lower and upper divisions in the sections, C and D are the only ones that can be entertained; and such a flood would have been sooner or later a natui-al consequence of the melting of the glacier in progress. It would have been a flood enormous in extent and vast in effects ; it may have been not merely an overflow of a few months, but of a period of years. (3.) The subdivisions of the layers into subordinate wave-like parts may have resulted from the plunge of the waves that accompanied the tidal or current movements of the waters. Each of these subordi- nate parts is not the Avhole that was formed by the plunge and flow of a wave, but this minus what it lost by the succeeding plunge or plunges — as a little study of figure 1 will make apparent. In these wave-like parts of a bed, the oblique layers usually diverge as they rise upward, as shown in figure 1. The wave struck at the end from which the lines diverge, and as it pushed forward with slackening force, it dropped more and more of the the sands taken up, and so the little layers formed l)y it were made gradually thicker. So much material deposited with one fling of a wave would seem to indicate rapid work in the deposition of the beds. The reasons for regarding these and other like beds as depositions directly from the glacier are the following. (1.) Stratified deposits were thus made by the glacier and the waters beneath somewhere about the New Haven region ; and no others exist that can be such. (2.) These beds, consisting largely of interstratified sand and gravel, and in part of layers of coI)ble stones, have characters according pre- cisely with the sup])osed mode of origin. It will be noticed that the layers of cobble stones have not required for their formation, on this view, streams of tremendous magnitude and violence, beyond all physical probability, in order to transport the stones from their place of origin, 50 or 100 miles or more to the north ; the work of transpor- tation was done quietly by the glacier, and they were simply dropped to their places ; much more moderate streams served to sift out the finer material so as to leave the larger stones alone. (3.) These sand and gravel beds could not have been formed like ordinary sand-banks on a sea-shore or in a bay. For the waves and currents, which are the means of piling up such banks, could not have introduced the lay- ers of gravel or cobble stones except they had been furnished by some other agency ; and the amount of sand that the waves move in Topographical Features of the New Haven region. 77 a stroke is but little, and this is spread widely, (4.) Again, they could not have been formed as sea-beach deposits. They have not the structure of such deposits. Moreover, if the beds of the New Haven plain had been produced by the gradual growth of a beach seaward, the harbor would have also existed somewhere, in narrow areas at least, among the encroaching beaches, and remains of the co- temporaneous mud-flats of the harbor should occur to mark its posi- tion. But no such clay or mud deposits have in fact any where been found, except in the Quinnipiac valley (upon which we remai'k be- yond) ; none along the courses of Mill and West rivers, where we should naturally look for clayey interpolations among the sands, if not thick beds. The work of filling up the bay was evidently too rapidly done for the accumulation of mud or clay from the contribu- tions of rivers. 3. Filling of depressions with the drift. — The depositions from the glacier filled up the greater part of the New Haven bay nearly or quite to the sea limit, as is shown by the even surface of the plain, the whole having been leveled off by the waters. The rivers, whei'e not too deep to be filled, had currents to sweep out the sand and keep them clear. The Beaver Pond depression, the great central basin of the bay, was one of the unfilled channels ; and unfilled, in all probability, be- cause of its depth. The drift was dropped over it as over the rest of the bay ; but its depth saved it from obliteration ; it still remained the open central basin of the bay. Its original communication with the wider outer portion of the bay was probably, as has been shown (p. 53), through the West Creek channel, whose extent, north-and- south course, and approximate conformity in direction and line with the Beaver Pond basin, accord with this view. During the melt- ing of the glacier there would have been an abundant flow of waters from the northward through it ; and these currents, together with the inward setting of the tidal flow, would have made the steep terrace- slopes that form its boundary, and those also of West Creek valley, which resemble in all respects the terrace-slopes along the rivers. But while not filled by the depositing drift, the Beaver Pond de- pression appears to have lost mucli in breadth ; for in the surface of the adjoining plain, especially along Crescent street, there are several large isolated basin-like depressions — deep holes, as they are often called, although sometimes 100 yards or more across — which must have been cut off by the depositions made by the glacier. The east and west Goffe street ponds occupy such exscinded depressions. The 78 J. D. Dana on the origin of some of the' valley of West Creek appears to have been dissevered from the Bea- ver Pond basin by the same means ; having no river (p. 52) to per- form the office of sweeper, it would have been unable to resist the encroaching sands. But while the Beaver Pond depression was thus closed in the direc- tion of West Creek, a tidal communication appears to have been kept oj)en between it and the deep parts of the bay, through the wide val- ley-like depressions near Webster and Munson streets, and thence through East Creek. The gently slojjing sides of the East Creek valley along the course of Chapel and Elm streets below Temple, as well as near Webster and Munson streets, and other facts ah'eady mentioned (p. 54) correspond with the view, as just stated, that this channel was originally a depression in the sandy bottom made by the sweep of the tides. Accepting these views, the channels of East and West Creeks, which diverge at the bay, make together the circuit of the original New Haven square, and converge toward the south- ern extremity of the Beaver Pond depression, were both, though at different times, outlets of this great central basin. The valley of Pine-Marsh Creek was another of the deeper glacier ex- cavations, as already explained ; and one too deep to be filled with the droppings of the glacier. This is proved by the remarkable breadth of the valley, and the fact that it is bordered by a steep terrace-slope, although no large stream but that made by the melting of the glacier ever flowed through it. There are deep holes or basins in the plain along its borders which may be explained in the same way as those adjoin- ing the Beaver Pond depression ; that is, they are spots that were un- filled by the sand and gravel of the glacier, because of their dejjth. The Quinnipiac valley was far the largest and deepest of the deep basins of the New Haven bay ; for while in one part a mile in width, the terraces on its eastern and western sides are very narrow. More- over they are mostly below the usual height ; and in some places so poorly defined as to be apparently altogether wanting. But to the south, between the basin and the bay, there is a great development of the drift or terrace formation, indicating that over this wide area the material was dropped by the glacier in shallow water. Red sand- stone, the basement rock, outcrops from beneath the sands of the for- mation south of East Rock and in Fair Haven, opposite borders of the plain. The fact that the tidal flow in the bay during the Cham- plain era was not over this area but either side of it (along Mill Riv- er and the Quinnipiac channel), is other proof that the region was originally shallow; for the course of the tidal wave is along the deep- er parts of a bay. Topographical Features of the Xew Haven region. V9 In contrast with the basin, the Quinnipiac valley near the village of North Haven and north of it has its lower flats exceedingly narrow and the uj^per plain of great extent ; and here, concordantly, the red sandstone is but a little way beneath the surface, for it outcrops along the river, and, as I am informed by Mr. D. H. Pierpont, is the bottom of all wells in the village. But the poor condition of the terraces in the Quinnipiac basin cannot be attributed solely to its extent and depth ; it must be owing partly to the currents that swept through the basin in the era of the melting glacier ; for the upper plain or ter- race, evidently for the same cause, has in general been left remarka- bly low, often not half its normal height, about North Haven and to the north. It is to be noted also that the drift formation or plain south of the basin may owe something of its extent and height to the diminished velocity which the waters would have had after passing East Rock, as they there escaped the bounds of the Quinnipiac val- ley, and were free to spread widely to the westward. 4. Origin of the material of the drift. — The sand, gravel and stones of the drift-deposit of the plain came largely from the Red sandstone formation; (1) the pulverized sandstone aflTording sand; (2) the associated conglomerate yielding pebbles and stones ; (3) the wear of fragments from the harder varieties of sandstone and con- glomerate making other stones or pebbles. There are some pebbles of trap, but they are very few in comparison with what the prece- ding source supplied. The rest came from the region of crystalline rocks to the northwest and northeast. The great trap boulders may have been derived from any of the trap mountains to the north. Those of the western border of the New Haven region, which are often tabular in form and sometimes thick- laminated in structure, were probably carried off from the heights be- tween the western of the Hanging Hills of Meriden and Mt. Tom, though possibly in part from the West Rock ridge more to the south. The great fallen masses in some of the valleys of the Meriden moun- tains resemble many of these boulders in form, in fine-grained text- ure, and in laminated or jointed structure. The masses of the Judges' Cave are probably from these more northern trap ridges which, as already mentioned, are the highest of the valley. This view of their origin accords with the fact that the gneiss boulders so common along with them are probably from the adjoining region of the town of Granby, or from Massachusetts, farther north, as stated by Percival after a comparison of the rocks. The quartz and quartzite boulders may be from the adjoining region in Massachusetts. But they are 80 J. D. Dana on the origin of some of the widely spread over the New Haven region, and tliey may have come from Vermont or New Hampshire, where such rocks occur,* 5. Rapidity of deposition. — The wasting of the glacier, beginning as the warm Champlain era opened, (p. 06) would at first have been slow, and mainly above. But after a while, the glacier would have been reduced to a comparatively thin sheet of ice, and then, through the heat conveyed into it in all directions by means of waters from above, and that received through flowing waters and air below, the rotting of the mass would have become general, and the unload- ing of the glacier would have gone rapidly forward. The period of years occupied by the deposition of the sand and boulders may there- fore have been short. It may be queried, considering how much ap- pears to have been done by a single wave, whether one year, or even less, would not have sufficed for the upper division, or the upper twenty feet, in the part of the formation rej^resented in figure 1, on page 73. With so quick a way of dumping the load of the great glacier it is nothing incredible that the channel of West Creek should have been cut ofi^" from its northern continuation, the Beaver Pond basin ; nor is it impossible that, by like means. Mill River should have had its course through the same basin and channel intercepted by half a mile of sand and gravel, and have been forced to open a new way for it- self by Whitneyville, although deemed improbable for the reasons stated on page 55. Even the floods of Niagara were thus stopped short; the old gorge, as long since made known, was filled to the brim for miles by the drift, and the river was turned ofi" to work out another passage through the rocks.f The accumulations of a " ter- * Besides the boulders described on page 68, there are the following in more remote localities. One, of trap, 6 miles in an air-line north of the city, 1-J- m. west of Ives' Station, fifty rods west of W. Fenn's, south and east of a bend in the road, is 88 feet in girt and 17 feet high, and must weigh over 600 tons. Less than half a mile south of tliis spot, near, and east of, the " West "Woods " road, a little south of R. Warner's house, there are four great trap boulders, nearly in a north-and-soutli line, the largest 50 feet in circuit. Half a mile north of the Mt. Carmel gap, a short distance west of the raOroad track, (and in full sight from the cars when passing), there is a boulder of trap, somewhat house-like in shape, which is 25 feet long by 14 wide and 16 high, with a girt of 68 feet; and along side lies a slice from its broad face, averaging 4 feet in thickness, which when a part of the mass, would have-made its diameters 25, 18 and 16 feet, and its original weight at least 450 tons. It shows traces of vertical lamination, like a trap dike, and was probably taken off from some trap-mountain before it had fall- en from its place. f Dr. Newberry in his memoir on Surface Geology, already referred to (p. 49), men- tions the Ohio river as another that was diverted by the filling up of the old channel Topographical Features of the JVeio Hacen region. 81 minal moraine " in tlie ordinary slow way would never have stopped the course of a Niagara. But before a sudden down-throw of sand and gravel from a freighted glacier, no stream is too large or rapid to hold its jDlace. Although the accumulation of freight by the old glacier must have required a very long period, even that of the whole Glacial era, the deposition of a large part of the older " alluvium," if the above view is right, was a rapid work — much more rapid than has hitherto been suspected. Any attempt to measure the interval of time between the depositing of the top and bottom layers by comparing the thickness of the formation at New Haven with the accumulations now going for- ward along the shores would lead only to great error. This conclu- sion holds not merely with reference to all similar formations made by direct deposition from the glacier, but also to others accumulated by the action of moving or running 'waters immediately afterward, inas- . . t in the Champlain era, and states that its present course along the falls or rapids near Loiiisville was thus determined. Other cases also are referred to. It is possible that in Mill River we find an example of such a change of course, as I have stated above. But the facts with regard to the Mill River gravel-course (p. 71) are another argument against it. It wiU be understood that this gravel is not that of the bed of the stream, but the material of the terrace or drift formation standing high along the border of the river ; and that it is similar in character above and below the WhitneyvUle dam. It seems to be good evidence that the river occupied its present channel throughout the period of the deposition of the drift. A change of course in the Quinnipiac through the cause alluded to is quite probable. The river at the bridge flows in a sandstone trough, the rock rising above the river 10 feet on the western side and over 20 on the eastern. Along the road running thence eastward to the depot (50 to 60 rods distant) which rises from 15 to 25 feet in level, there is no sandstone, and instead a deep bed of the sand of the stratified drift. The wells at three of the houses west of the depot go down 2 to 4 feet below high- water mark in the river, without reaching the " red rock." Moreover the low flats of the river north of the bridge spread eastward and sweep around to within 40 yards of the depot ; and ia consequence, a brick house recently buUt opposite the depot (across the street), while it stands in front on the firm sands, rests its northern or back walls on piles which were driven down in the meadows 20 feet without finding for all of them a firm footmg. That the river's bed was once here is the supposition of those on the spot who know the facts. But we may suspect further, that the river from this point flowed southward to join the present channel half a mile below, the low level of the bottom of sandstone over this region determining it; and that the sands and gravel derived directly from the glacier or indirectly through the river floods, during the sub- mergence (45 to 50 feet in amovmt, as shown beyond) of the Champlain era, filled up the earlier channel, so that the stream, when the land was afterward elevated was forced to open a new channel, in doing which it took a course over the rocks because compelled to it by the existing slope of the surface. Trans. Connecticut Acad. Vol. II. 6 January, 1869. 82 J. D. Dana on the origin of some of the much as the hills and valleys were everywhere left by the glacier loaded with sand and gravel ready and convenient for transportation. The evidences of rapid deposition are so many and obvious that they appear to set aside any theory of the glacial cold which demands a slow decline of the era. 2. Later events and results of the Champlaiu era. 1. Continuation of the Drift formation. — It has been stated that, during the progress of the depositions by the melting glacier in the bay, the lighter or finer portions would have been largely sifted out by the moving waters ; and while part of the sands would have been eddied off to one side, a much larger part would have gone with the current and the ebbing tide down the bay to be distributed by the tides chiefly at their influx along its borders. Over the whole of the wide western portion of the New Haven pltein, and especially the southwestern, the terrace formation consists of sands. To the north, toward Westville, at the entrance to West River valley, there are pebbly layers ; but, on passing southward, these rapidly lose most of their pebbly character, and increase in fine- ness ; and between Congress Avenue and Oyster Point, the beds are almost solely sand. The detritus which is now borne by the rivers to the bay is distributed largely along its western side, and there, consequently, are the great sand flats ; and this is so because the di- rection of the tidal current in the Sound on its influx is to the west, and as it entei'S the bay to the northwest ; and the depositions of de- tritus take place mainly during the inflowing tide. The same would have been the action of the currents and tides in the Champlain era ; and hence this western part of the New Haven region would have been, from the beginning of the depositions, an area of accumulating sa7id beds. The part of these sand beds that were made during the progress of the melting, should be marked off, if they could be distinguished, as belonging to the first section of the Champlain era, and only the sub- sequent additions, as " later results ;" but the progress of the beds through the two intervals was continuous, and it is probably impossi- ble to ascertain the limit between them. The hills and valleys, after the melting was completed, would in many places have been left thick- ly covered with sand and gravel ready for transportation by every lit- tle rill the rains might make, and the rivers would for a considerable time have continued to transport an unwonted amount of sand. The depositions along the borders of the bay for a while would, therefore, have gone forward with a rapidity almost equalling that of the melt- Topographical Features of the Neto Haven region. 83 ing period itself; and the decrease of rate would have been quite gradual. On the west side of the bay near Halleck's place (where the present railroad grounds abut against it), a section of the terrace for- mation 25 feet in height (the upper twenty-five) is exposed to view, and throughout it, the beds have pi'ecisely the structure exhibited in fig. 1 (p. 73), and difier only in the paucity of pebbles ; they evince the same free supply of material and rapid deposition under the action of the waves. Moreover, the slope of the oblique lamination is toward the south (as in the lower part of fig. 1), showing that the deposition was accomplished mainly during the inflowing tides. The result of all this transportation and deposition was an exten- sion southward of the sand beds, as well as an increase in their height ; and the terrace formation was thus completed to its outer limit. The plain stretching south to Oyster Point and over West Haven gives us some idea of its extension in that dii'ection ; but not necessarily its original extent, since the sea may have washed away much from its borders as well as from its upper surface. Over the region toward Oyster Point, the beds are sandy through- out, and free from any upper layer of fine river or bay detritus, such as is deposited about existing mud-flats and sand-banks. On Grape Vine Point, between the mouths of Mill River and the Quinnipiac, there is the same absence of any thing like a layer of harbor mud over the sandy beds of the drift formation. The proof appears hence to be quite positive that these sandy beds did not lie for a long period beneath the water, after the material was deposited. 2. Sand-formations on the borders of the Quinnipiac valley, — The Quinnipiac valley was the site of the inner harbor of the bay, during the Champlain era — and a harbor of great extent and depth, as already stated. While the sand-formation was in progress down the bay, changes should have been going forward within its area. On its bor- ders we naturally look for sand beds distinguishable from those that were made during the hurrying time of the melting by unconforma- bility, and also by freedom from layers of coarse pebbles and cobble stones. One locality of such sand beds of considerable extent occurs on what was the southwestern border of the old harbor, at the north- eastern end of the cut made through the terrace formation for the Air- line railroad, south of the East Rock range. The character of the terrace formation along this cut has been described on pages 73, 74. The position and general character of these whitish sand beds are shown in the following cut. The part A B C is the terrace formation, con- sisting of beds of sand, some quite coarse pebbly, and below including 84 J. D. Dana on the origin of some of the three layers (M) which are of clayey sand. A B is the plane of sepa- ration between this part and the layers of whitish sand. The latter differ not only in their white color, but also in the absence of all peb- bles, and in the much greater fineness of the sands. Through the washing of the waters against the shores, they were not only ground up, but they also lost almost entirely the oxyd of iron that tinges the quartz grains of the proper terrace formation. At the foot of the slope A B there is a collection of pebbles or stones, and for a short distance east of B, reddish sands ; the pebbles and sand evidently fell down the bank from the layers above, when it existed as an exposed slope before the beds of whitish sand were deposited. These sands, moreover, were laid down in even layers, free from the oblique lami- nation that occurs in the terrace formation. 3. Mudformations in the Quinnipiac harbor. — Besides these shore formations, the old harbor had its mud beds. They are the clay beds situated along the borders of the present river flats or meadows imder 3 or 4 feet or less of sand : in these later times they have become the sites of numerous brick-yards. The clay beds vary in depth from 6 or 8 feet to over 25, the bottom in some places not having been reach- ed at the latter level. Where penetrated they are found to rest on sand. The clay is very thinly and evenly laminated. The beds have been opened at several points near the outer borders of the meadows, on both their eastern and western sides, through a length on each of about three miles. The width of the border of clay is reported to be from 100 yards to a third of a mile. The two ranges converge toward North Haven, where the harbor had its head, and where, moreovei', the terrace formation becomes wider and crowds upon the river. The clay continues north, between layers of sand, under the lower part of the village of North Haven. I leai'n from Mr. D. H. Pierpont, that in digging a well in the village of North Haven, after passing through 7^ feet of sand, a bed of clay 4 feet thick was met with, the bottom of which was ^\ feet above the level Topographical Features of the New Haven region. 85 of the Quinnipiac river. This 8^ feet was made up of fine quicksand. The clay was a sandy clay, or what the brick-makers call " weak clay." This well is about 80 rods east of the depot. At two others, between the depot and the river, clay was found, and in one, there was at top 4 feet of sand ; then 5 feet of " weak clay ;" and below quick-sand, 3 feet of it above the level of high-water in the river. The clay beds, according to Mr. I. L. Stiles, do not extend beneath the deep muck of the great meadows ; on reaching the muck, instead of keeping at the same level, it dips dowyiward with a rather large angle beneath the muck. What lies beneath the muck, whether clay or sand, has not been ascertained. In making the track for the Air- line railroad, which runs for nearly a mile and a half obliquely (north- eastwardly) across the flats, piles were driven to various depths, down to forty feet ; solid bottom was reached, but the nature of its material is unknown. Over the region north of North Haven village, the upper plain or terrace is very wide and the lower relatively narrow, the reverse of what is true to the south. Moreover, the country is remarkably sandy, large fields of loose moving sands making part of the surface. These sands are the present top of the upper plain or terrace. When this region, in the Champlain era, lay at the head of the great Quinnipiac harbor near high tide level, it was in a condition to be washed over by the running waters, and it is probable that the grinding and sifting then went on that robbed the sands of their feldsj^athic and other softer grains ; and that what the sands lost the harbor received as a contribution to the mud of the liarbor, now the clay beds. The description of these beds of clay is here inserted under the head of the "Later events of the Champlain era." But it is not at present possible to decide whether part, or even all, of the deposition may not belong to the early part of the era. We need to know some- thing more definitely with regard to the relative positions of these beds and the others of the drift-formation before a positive conclusion can be arrived at. The layers of sandy clay in the section at the cut for the Air-line railroad, represented in fig. 2 (p. 83), although 20 feet above the level of the meadows, may have some relation to the clay beds farther north. The fact that they have a dip toward the Quin- nipiac basin is a significant one, as intimated on page 72. What depositions were going forward at this time in the Beaver Pond basin, the central basin of the New Haven harbor, cannot be ascertained without an artesian boring. Such a boring would de- velop several facts of interest ; for the depth to the sandstone bot- 86 J. D. Dana on the origin of some of the torn would give the depth of the original excavation ; that of the beds of sand over it, the thickness of the drift derived from the gla- cier ; that of any clay hed, or infusorial bed, or shell deposits, and of the peat, other important points in its history. The de]3th of the basin was small compared with that of the Quinnipiac harbor, as is evident from the present level of its meadows. 4. Denudation. — In this era of submergence, the sea breaking against the foot of East Rock and the other cliffs of the bay, must have worn away the sandstone along the base, and thus carried for- ward the degradation of the trap dikes and sandstone hills which had been begun by the glaciers. The waves acted upon the region in front of Pine Rock both from the direction of the Beaver Pond basin, and that of the broad West River channel. The part of the Beaver Hills occupying this position being thus attacked on both sides woidd have been soon swept away and a free passage made across for the waters. This spot is now occupied by a portion of the New Haven plain, directly proving that waters communicated across from the Beaver Pond basin to the West River channel, or the reverse, as just stated; and the degraded condition of the front of Pine Rock is further proof of the action of the sea here supposed. The sweep of the tides across this region, would have some where made a tidal chan- nel ; and this channel, as the high terraces either side show, was that which after a while became, and now is, the outlet to the Beaver Pond, along the north side of the Beaver Hills (see map). In like manner, a depression was made in front of the larger part of Mill Rock, by encroachments upon Sachem's ridge. The disjunction was not so com- plete as in the case of the Beaver Hills, because the central basin of the bay, the Beaver Pond, gave no aid through its currents and wa^'es, since it was remote from Sachem's ridge, while close along side of the Beaver Hills. As already observed, the streamlets descending the front of the Rocks after rains would have aided in the process of denu- dation, and with much gi'eater effect after the elevation of the land which closed the Charaplain era. 3. Life of the Champlain Era. More than a score of years since, according to Mr. I. Lorenzo Stiles, the antlers of a buck were dug up at a dejjth of 10 or 15 feet at the Stiles clay-bed near North Haven village. Mr. Stiles informs us that they were those of the common species of deer. The specimen was deposited in the New Haven Museum, an institution which years since came to its end, and it has been lost sight of, so that the fact Topographical Features of the Neio Haven region. 87 with regard to its species cannot be verified. It is also stated by Mr. Stiles that impressions of leaves liave been found in the clay. The muck at a depth of 6 to 12 feet has been found to contain at places great logs and stumps, nuts and leaves, accredited popularly (and probably rightly) to trees of existing species. But these are subse-- quent in age ; for the muck beds of the interior of the basin could not have been begun until the salt-water harbor had been mostly ob- literated by an elevation of the land. The above is all we have yet gathered from the deposits of the New Haven region with regard to the life of this era. It is certain that there is much more to be learned ; for there is good evidence of the existence of the ^lastodon formerly in this part of Connecticut. While digging for the Farmingtou Canal in Cheshire, 13 miles north of New Haven, three or four teeth of a Mastodon were found, (Am. J. Sci., xiv, 187, 1828); and long before, remains of the same animal were obtained near Sharon. xVIso later, a vertebra of a Mastodon was brought to light in digging a canal for a manufactory in Berlin, the bone occurring in "a tufaceous lacustrine formation, containing bleached fresh-water sliells of Planorhis, Lymncea, Cyclas, etc., sim- ilar to those of the waters in the vicinity." (Am. J. Sci., xxvii, 165, 1835). This Berlin Mastodon existed as late as the Champlain era ; for if of earlier time the lacustrine deposit would have been buried beneath drift, either the stratified or unstratified. 6. Terrace or Recent Era. The work of the waves, tides and rivers went forward until the great drift formation of the bay and river valleys was completed. An elevation of the land then commenced which affected cotempora- neously all New England, and, it is believed, a large part of the con- tinent, and bordered the rivers and lakes with terraces. This eleva- tion marks the transition to the Terrace or Recent era. 1. Amount of Elevation. In determining the amount of elevation of the land about the New Haven region, we have to take it for granted, not only that the plain was leveled off by the waters, but further, that a considerable part of its surface at the time nearly coincided with that of the water. The even character of the plain shows that this is not an improbable as- sumption. The following are the results of the observations upon its level thus far made. The heights along the river valleys, the Beaver Pond ba- 88 J. D. Dana on the origin of some of the sin, the valley of Pine -Marsh Creek and the borders of the bay are from approximate measurements, by means of a hand-level, by the author. The rest are from the large map of the city, published in 1858, from surveys and drawings by Mr. S. W. Searl. The distances from Oyster Point given are differences of latitude, or northings, in statute miles, and are derived from published maps. In reckoning the heights mean-tide level is taken as the base. The heights are not given of such parts of the terrace or plain as are obviously below the true or normal level (owing to river or tidal currents, or other causes), a fact generally made manifest by neighboring portions being at their full elevation. I. Height of the surface along a nearly north-and-south course through the middle portion of the New Haven plain, from Oyster Point, by the College Square, to the Beaver Pond Meadows, and thence, half a mile to the eastward along the valley of Pine-Marsh Creek, (or as it is sometimes called Pine-Swamp brook). Northings from Oyster Point. Height of Terrace. Oyster Point miles In line with id., w. of West R. " N. of Oyster Point 0-50 " Halleck's Place, S. side* 075 " " " N. side 0-87 « College St., front of S. College 1-85 " York Street, corner of Broadway ^'00 " Beaver Pond basin, S. end 270 " Id., E. end of Munson street Creek 2-80 " Id., W. end of Munson street 2-80 " Id., outlet, W. side 3-40 " Id., opposite outlet, on E. side 3*40 " Id., farther north, E. side 3-65 " Id., farther north, at road crossing 3*80 " P. M. Creek valley, at southwest point 3-80 " Id. at road crossing, N. W. of VV. end Mill Rock 4-20 " Id., farther north 4-30 " Id., S. E. of Hamden Church 4-55 " Id., at mouth of Creek 5-15 " * The reader is advised to put a letter H on the map (p. 44) three-sLxteenths of an inch north of the letter 0, west of the harbor, and the letters Gr P on the point of land between the mouths of Mill River and the Quinnipiac. 2H feet 24i 27 30 30 38 4H 43 43 44 53i 53i 55^ 56 56 62 63 66 72 Topographical Features of the New Haven region, 89 II. Up West River valley. Crossing of N. Y. railroad End of Washington street Crossing of Milford turnpike, W. side North of Id. Above crossing of Oak st., W. side Crossing of Derby turnpike, W. side " " " E. side Crossing at Westville, E. side Crossing at Westville, W. side Near Congregational Church, Westville III. Up Mill River valley. Near Neck Bridge, east side, to highest point Suydam Grounds, Whitneyville, W. side Above dam, below 1st bridg-e, W. side Whitneyville Church N. of the Church Moutli of Pine-Marsh Creek Above mouth of Id. IV. Up the Quinnipiac valley. Foot of Third St., Fair Haven, near mouth of river Crossing of Shore-line R. R., W. side of river North of Id. Crossing of Air-line R. R., W. side of river At North Haven Sorth. from Oyster Pt. Height of T. iu feet. 0-20 23 0-50 27 0-62 29 1-25 36 1-40 37i 1-85 38i 2-25 41 215 40 3- J 5 46-47 315 45-46 3-45 56-57 2-00 42 3-30 53 3-90 63 4-40 66 4-55 69 515 72 5-60 76 1-33 34i 1-80 40 1-90 41 2-50 45 700 46 V. South of the latitude of Oyster Point. On the west coast of the bay, 0*33 m. south of Oyster Point, the terrace is 24 feet high ; 0-43 m., 21 to 23 feet; 1-30 m., at the Savin Rock beach, only 8 feet, but about 300 yards north, 14 feet. In the following table the results in the preceding four tables are brought together for comparison. The Roman numerals I, II, III, IV, indicate the table from which the numbers below are taken. 90 J. D. Dana oti the origin of some of the North, from I. II. III. IV, OyBter Point. Middle of plain. "West River. Mill River. Quinnipiac 0- 21i-24i 2U 0-20 23 0-50 27 27 0-62 29 0-75 30 0-87 30 1-25 36 1-33 Sii 1-40 37* 1-80 40 1-85 38 38. V 1-90 41 2 00 4H 42 2-25 41 2-50 45 2-70 43 2-80 43 2-88 44 315 46-47 3-30 53 3-40 53^ 56-57 3-45 365 55 i 3-80 56 3-90 62 4-20 63 4-30 64 4-40 66 4-55 66 69 515 72 72 The Beaver Pond Meadows and the valley of Pine-Marsh Creek are natural levels, the former over a mile long, the latter three-fourths of a mile, each con- taining a range of nearly still water along the bottom through this distance; and hence the height of the terraces on either side is ascertained with great facility. It has already been stated that in the latter this water level is determined by the Whitneyville dam, so that the height of the dam gives, after an allowance for the back-water rise, the height of the water above mean-tide level, even for that of the upper part of the valley west of Mill Rock. The edge of the dam over which the water falls is 34 feet 8 inches above the base of the dam, according to Mr. Eli Whitney ; and the surface of the water a few yards back is 4 inches hig-her, making in all 35 feet for the whole height of the fall. The base of the dam is very near mean-tide level. The back water above the dam extends about 2^ miles, to within 300 feet of the dam at Augerville ; the inciease in the height of the surface along this distance has been estimated to be about 6 inches a mile ; or 15 inches for the whole distance, and 8^ inches to the mouth of Pine-Marsh Creek. The elevation above mean-tide level of the water-surface in the Beaver Pond Meadows near its outlet, is taken at 22 feet, in accordance with information re- ceived from Mr. Eli W. Blake as to the heights of the dams between the Beaver Pond meadows and West River. A few hundred feet above the outlet of the Beaver Pond basin, the meadows commence a rising grade northward, as is obvi- Topographical Features of the New Haven region. 91 ous in the rippled surface of the little streamlet which flows along' it; the increase of height thereby at the crossing- of the road to Pine Rock (3-SO from O. P., in Table I) is at least 3 feet; and beyond this to the north the slope of the meadows runs parallel closely with that of the terrace plain either side, the height of the plain, even to its northern extremity, above the meadows being quite uniformly 31 to 32 feet. The observations show that the plain rises gradually to the north- ward. The average increase of eleviition from Oyster Point to the mouth of Pine-Marsh Creek, a distance of five miles, is 10 feet per mile. From Oyster Point to York street, two miles, it is 8|feet; and to College street nearly 1^ feet ; from College street to the Avest end of the Munson street crossing of the Beaver Pond meadows, one mile, it is only 6 feet ; along the Beavei- Pond basin, from its south- ern end to the I'oad which crosses it a little south of the line of Pine Kock, a mile in distance, the rise is 1 3-|- feet ; along the valley of Pine- Marsh Creek, the average per mile is about 12 feet. The slope for a mile north of College street, that is, between 1-80 and 2-80 miles in latitude from Oyster Point, is more gradual than either to the north or south ; and the same is true for a surface of like latitude near West river, on which the increase in elevation from Oak street (1"85) to Westville (3-15), 1-30 miles in distance, is only 8^ feet. The fact that the increase of elevation northward is by a gradual slope, a}id not througJi a succession of two or more ahrupt terraces, is manifest along Dixwell Avenue (the road to Ilamden Plains). The Avenue lies to the east of the Beaver Pond Meadows, and to the west of Pine-Marsh Creek, and extends northward in a nearly straight line, bej^ond the mouth of Pine-Marsh Creek; and hence any terrace would be apparent along its cou.rse, or in the fields either side, if such existed. The observations prove the fact of an elevation of the land along this part of Connecticut after the Champlain era, the era in which the drift formation was made. They also appear to prove that this eleva- tion was greatest, by a nearly regular rate, to the north. But before arriving at any conclusions as to the amount of elevation, or its rate of increase northward, it is necessary to consider: First, Whether part of the slope above pointed out did not exist in the surface before the elevation began. Secondly, Whether part of the slope was not formed by the retreat- ing waters during the ])rogress of the elevation. Tlnrdly, Whether part is not a result of a sinking of the more southern portion of the plain since the elevation. 92 J. D. Dana on the origin of some of the (1.) Slope antedating the elevation. — There can be no doubt that part of the slope antedates the elevation. This may be true of each end of the slope, that is (A) the southern part adjoining the bay, and {B) the northern part. A. The Southern Part. — A slope in the southern part may have arisen («) from the tidal curi*ents with or without the waves, aided by the river floods ; [h) from the waves alone ; or (c) from increasing depth in the bay outward, and decreasing supply of sand. (a.) From tidal currents. — Oyster Point projects toward the Sound between West River and the bay, and in this exposed condi- tion would in all probability have been swept by the tides in such a manner that it would have failed to be built up to the water's surface or to mean tide level. The eastern part of this Point for the last half mile is actually half lower than the western or that bordering West riv- er, owing undoubtedly to the action of the cause here mentioned ; and the western sufiered also ; ibr, on the other side of West River the ter- race is 24 to 24^ feet high. Grape Vine Point, a mile and a quarter farther north in the bay, is another example of this effect, as observed on page 72, where the facts as to its little height over its middle, and the western or Mill River side, and the full height of the terrace on the Quinnipiac side, are stated. Had the Point been a little narrower, it might have been low all the way across, so that it would have re- mained doubtful whether this low level was due to tidal currents or not. But the heights on the Quinnipiac side are as great as those of the middle of the New Haven plain in the same east-and-west lines, so that they have nearly the normal elevation. They show, therefore, that the lower part of the Point is over 20 feet below the normal level, owing to the action of the great central tidal flow up the bay. Again, at the corner of State and Chapel streets, along side of the channel of the old East Creek, the present height is 15 feet, or about 22 feet be- low the full height for the latitude ; and this influence of the sweep of the tides is felt all the way nearly to an east-and-west line through the corner of College and Chapel streets. It is quite certain, in view of these facts, that Oyster Point was in no part built up to the water level. How much to allow for the defi- ciency, we have not facts to determine. An allowance of 10 feet could not be too great; and this would give 31 or 32 feet as the height which the Point would have had, if no such cause had ope- rated. If the surface of the plain at Oyster Point, corresponding to the original water level, is to be reckoned at 30 feet above the present lopograpMcal Features of the IVew Haven region. 93 mean tide, then the slope normal for the whole of the plain to the southern part of the Beaver PonS. J. Smith on American Cnistacea. A. — Species in which all the sejments of the abdomen are separated by distinct articula- tions, and in ivhich the front is very much contracted between the bases of 'the ocular peduncles and somewhat spatulate in form. Gelasimus heterophthalmus, sp. nov. Plate II, figure 6, 6^ Plate III, figure l-l^ Male. The carapax is somewhat quadrilateral iu outline, but the antero-lateral angle on the side of the larger cheliped is much produced laterally, so that the orbit is much longer on that side than on the other and the lateral border strongly divergent. The dorsal surface is smooth and shining, and convex longitudinally but not at all late- rally. The branchial regions are very slightly SAVollen, scarcely high- er than the gastric and cardiac regions, and are separated from them by slightly marked sulci. The front is spatulate, contracted between the bases of the ocular peduncles and much expanded below. The superior border of the orbit is much excavated at the base of the ocu- lar peduncle, and strongly arcuate in the middle, and has a very slight- ly upturned and entire margin. The antero-lateral angle on the side of the smaller cheliped, is angular but does not project either anteri- orly or laterally, while on the side of the larger cheliped it is broad, obtuse and projects very much laterally, as described above. The lateral margin is obtuse and its posterior part only is indicated by a faint granulous line. The upper part of the inferior branchial region is oblique, flat and very smooth, and is separated from the lower por- tion by a slightly raised line running straight from the antero-lateral angle to the base of the third pair of ambulatory legs. The inferior border of the orbit is denticulate with minute, flattened and truncate teeth. The jugal regions are smooth and shining. The ocular peduncles are rather slender, slightly enlarged at the cornea, and the one on the side of the larger cheliped is consider- ably the longer and is terminated beyond the cornea by a very slen- der filiform stylet, much longer than the peduncle itself, and slightly flattened and expanded at the tip. There is no trace of a terminal stylet on the peduncle of the other side. In the larger cheliped, the anterior surface of the merus is smooth, narrowly triangular in outline and considerably convex, the inferior margin is sharp and denticulate, and the superior margin is armed with a slight crest which is very low and entire for most of its length but quite high, and in some specimens slightly dentate, at its distal extremity. The carpus is short and its upper surface is slightly ver- rucose. The basal portion of the propodus is rounded and coarsely aS. I. Smith on American Crustacea. 117 and densely verrucose externally, the superior and inferior margins are thin and dentate, and the inner surface is nearly smooth, excepting three, high, tuberculose crests, of which one runs obliquely upward from tlie inferior margin, one from the base of the dactylus along the margin of the depression into which the carpus folds, meeting the first in nearly a right angle, and another along the margin next the base of the dactylus, leaving a rectangular, depressed area between it and the lower crest. Both the fingers are smooth on the inside, quite long, compressed and high, and the prehensile edges are evenly tuber- culated and each armed with a single, stout, median tooth. The oute r sui-fiTce of the propodal finger is somewhat roughened with irregular, shallow punctures, the inferior edge is granulated and has a submar- ginal, granulous line on the outer side, and the prehensile edge is armed with a stout tooth considerably within the tooth on the dactylus ; the edge beyond this tooth is straight and closes evenly against the dactylus, but between the tooth and the base it is deeply excavated, leaving a short and broad opening between the bases of the fingers. The dactylus is smooth on the outside, except a small space at the base, its superior edge is entire and smooth, and the prehensile edge is nearly straight, tuberculated and armed with a stout tooth a little beyond the middle. In the smaller cheliped the merus is slender and somewhat trique- tral, and the superior and exterior angles are sharp. The carpus is short, ovoid in form, and smooth and rounded externally. The hand is slender, and the fingers long, flattened at the tips, and the angles clothed with hairs. The ambulatory legs are smooth and unarmed. The abdomen is contracted at the articulation of the first with the second segment, and the edges are straight from the second segment to the terminal, which is broad and obtusely rounded at the extremity. Four specimens gave the following measurements : — 1. 2. 3. 4. Length of carapax, 18-7mm iS-Sm™ 18-2mm le-gmm Breadth of " Ratio of length to breadth, . - . - ] Length of larger hand, Length of ocular peduncle on side of smaller cheliped, 140 Length of ocular peduncle on side of larger cheliped, excluding stylet, Length of terminal stylet of ocular peduncle, In numbers 3 and 4 the ocular stylets are broken and partly wanting. Quite a number of specimens are in the collection of the Peabody 3'2-2 32-3 30-0 27-2 1-V2 : 1:1-75 1: 1-65 1: 1-61 48-4 53-5 430 37-5 140 14-3 12-9 12-3 16-2 16-3 15-0 13-8 19-4 20 10 + 118 )S. T. Smith on American Crustacea. Academy of Science, all obtained at the Gulf of Fonseca, west coast of Central America, by J. A. McNiel. This species is apparently closely allied to the G. styliferus, but the ocular stylets in that species are very short, and the hand, as figured by Edwards, is shorter and higher in proportion than in our species. The description of G. styliferus is, however, too short to permit of a detailed comparison of the species. Gelasimus styliferus Edwards. G'etosmMS^Zail/dociyZMs Edwards, Regne animal de Cuvier, 3™e ^(jit.j Crust, pi. 18, fig. 1 ''■, Tion Histoire naturelle des Crust, tome ii, p. 51, 1837, {teste Edwards). Gelnsimus styiferus Edwards, Annales des Sciences naturelle, 3™«' serie, Zoologie, tome xviii, 1852, p. 145, pi. 3, fig. 3. The following is the description given by Edwards : — '• Espece tr6s voisine du G. platy dactyl a s.^ mais ayant le Crete marginale du bras moins developpee et les podophthalmites tei-mines par un petit stylet comme chez les Ocypodes. — Guayaquil." Gelasimus heteropleurus, sp. nov. Plate II, figure 7. Plate III, figure 2-2". Male. The carapax is quadrilateral in outline, but the antero-lateral angle on one side is produced as in G. heterophthalmus. The dorsal surface is slightly granulous, quite flat anteriorly and only slightly convex posteriorly. The branchial regions are not at all swollen but are separated from the gastric and cardiac regions by deep sulci. The front is spatulate and expanded below the bases of the ocular pedun- cles. The superior border of the orbit is arcuate in the middle and has an upturned and slightly crenulated margin. The antero-lateral angle, on the side of the smaller cheliped, is acute and projects slightly forward, while on the side of the larger cheliped, it projects laterally as a very prominent obtuse tooth. The lateral margins are angular and armed with a very marked line of sharp granules. The upper part of the inferior branchial region is smooth and nearly perpen- dicular. The inferior border of the orbit is thin and denticulate with minute, flattened and truncate teeth. The jugal regions are granulous. The ocular i)eduncles are slender, much enlarged at the cornea and the one on the side of the larger cheliped is much longer than the other and is terminated by a slender flattened stylet about as long as the cornea. In the larger cheliped, the anterior surface of the merus is narrow, somewhat convex, and smooth, its margins are minutely denticulate, 3nd the superior one is armed with a narrow crest-like process at the S. I. Stnith on American Crustacea. 119 distal extremity. The superior surface of the carpus is flattened and granulous. The outer surface of tlie basal portion of the propodus is thickly verrucose, the verrucae near tlie upper margin being coarse and tuberculiform, the inner surface is armed only with the oblique tubercular crest running from the inferior marsjin. Both fingers are smooth on the inside, compressed and short, being but little longer than the basal portion of the pi-opodus ; their prehensile edges ai'e evenly tubercular, each armed with a tooth a little way from the tip, and nearly straight, but widely separated at base, leaving a broad, open space within the teeth, but beyond the teeth, the edges meet and the tips hook by each other. The outer surface of the propodal finger is granulous or minutely verrucose and the inferior edge is minutely tuberculated and has a submarginal crest on the outer side. The outer surface of the dectylus is granulous like the other finger and the superior edge is somewhat tuberculated or denticulate. The smaller cheliped and the ambulatory legs are very much as in G. heterophthalmus. The abdomen is quite similar to that of G. heterojyhthdlmus, but is more narrowed toward the tip and the edges are slightly concave. Length of carapax, ..... Breadth of " Ratio of length to breadtli, . . . . l Length of larger hand, ..... Length of ocular peduncle on side of smaller cheliped, Length of ocular peduncle on wde of larger cheliped, excluding stylet, ....... Length of terminal st3'let of ocular peduncle, I have seen but two specimens, both obtained, with the other spe- cies mentioned, by Mr. McNiel, at the Gulf of Fonseca (Collection Peabody Academy of Science). In the lengtli of the ocular stylet this species agrees with the G. styllferus, but the merus and hand in the larger cheliped are very different, and at once distinguish it from that species. The Gelasimus vocans of Desmarest (Considerations generales sur la Class des Crustaces, p. 123) seems to be distinct from any of the species descibed by recent authors and apparently belongs in this section, as it is distinctly stated that the ocular peduncles are ter- minated by stylets. Edwards refers it to his G. jxihtstris, to which it evidently cannot belong, but, as the character of the front is not stated, it may possibly belong in section B, forming in that case a sub- section with ocular stylets. 15.8mm 15-2mm 25-0 25-6 1-58 1 : 1-68 320 36-0 10-1 ... 12-0 123 2-5 2-8 120 S. T. Smith on American Crustacea. Desmarest's description is as follows : — " Carapace unie, avec le bord anterieur sinueux ; serre droite ordinairement phis grande que la gauche ; toutes les deux etant finement ehagrinees en dehors, avec une ligne en- foncee courte, pres de leur extremite, et ayant leurs doigts longs, etroits, tr^s-ecartes entre eux, unis, comprimes ; pedoncules oculaires pourvus a leur extremite d'une pointe aigue. Des Antilles." G-elasimus princ.ps, sp- nov. Plate II, figure 10. Plate III, figure 3-3^. Male. The carapax is in the form of a trapezoid much contracted behind, and the dorsal surface is smooth and shining. The branchial regions are somewhat gibbous, are higher than the gastric and cardiac regions and are separated from them by deep sulci. The fi'ont is spatulate and much contracted between the bases of the ocular peduncles. The su])erior margin of the orbit is strongly curved, the jiosterior margin is slightly raised and minutely denticulated, and the outer angle projects laterally as a very prominent triangular tooth, which is considerably larger on the side of the greater cheliped than on the other side, so that the carapax is somewhat un symmetrical. The lateral margins are marked by sharply granular lines which curve slightly inwai"d and rapidly converge posteriorly. The upper portion of the inferior branchial region is quite oblique, flat and smooth, and is separated from the lower portion by a slight, granu- lated line. The inferior margin of the orbit is armed with about twenty-five small, compressed and truncate teeth. The ocular peduncles are unequal in length, the one on the side of the larger cheliped being the longer, very slender but considerably enlarged at the the cornea and shorter than the broad, open orbits. The larger cheliped is enormously developed, the hand being nearly three times as long as the carapax. The anterior surface of the merus is flat and smooth, and its superior margin projects into a thin, high, evenly arched and sharply dentate crest, and the inferior angle is armed with a line of small and closely set spines. The upper surface of the carpus is rounded and verrucose and the inner margin is angu- lar and denticulate. The basal portion of the propodus is rounded and coarsely verrucose externally, the superior margin projects as a thin crest beneath which the carpus closes, the inferior margin is dentate, and the inner surface is smooth, excepting two tuberculose crests, of which one runs obliquely upward, from the base of the dactylus, along the margin of tlie depression into which the carpus folds and meets the first crest in a rigfht anirle. The fingrers are much jS. I. Siiiith on American Crustacea. 121 compressed and very long, the inner surfaces are smooth, and the pre- hensile edges are very tuberculose and each is armed with a stout tooth near the middle, the tooth on the dactylus being a little nearer the base than the other; within these teeth the prehensile edges gape widely leaving an ovate space, while beyond the teeth, the edges meet and are nearly straight almost to the tips, which, however, are strongly curved. The outer surface of the digital portion of the propodus is nearly smooth but has a submarginal, crenulated crest below, and the inferior margin is denticulate. The outer surface of the dactylus is somewhat verrucose and the superior edge is denticulate and slightly margined toward the base. In the smaller cheliped, the merus is slender and somewhat trique- tral and the superior and exterior angles are sharp and granulated. The hand is very similar to that of G. heterophthalnms. "^riie ambulatory legs are stout and nearly naked and the meral seg- ments are somewhat compressed and their edges shar]) and minutely denticulate. The abdomen is broad, the basal segment is considerably shorter til an the second and third, the edges approach each other somewhat at the junction of fifth and sixth, and the terminal segment is nearly twice as broad as long and its extremity is rounded. Five specimens give the following measurements : — ,h of carapax. Breadth of carapax. Eatio. Length of larger hand. 24-linm 41-imm 1: 1-71 64-omm 24-0 39-8 1: 1-66 70-0 23-4 39-8 1: 1 70 71-4 220 36-4 1: 1-65 64-4 213 360 1:1-69 60-4 I have examined a large number of specimens of this species col- lected at Corinto, on the west coast of Nicaragua, by J. A. McNiel, (Collection Peabody Academy of Science). There are three female specimens of Gelasimiis collected at the same locality by Mr. McNiel, which probably belong to this species although they differ quite remarkably from it. The carapax (Plate II, figure 8) is not so much narrowed behind as in the males, the dorsal surface is evenly convex and thickly covered with rounded granules, which are quite coarse along the lateral borders, and the branchial regions are not raised above the gastric and cordiac regions, and are separated from them only by slight sulci. The sides of the carapax are perfectly symmetrical, the anterior angles are prominent and sharp, and the lateral margins are marked by sharp crests of bead-like 122 S. J. Smith oti American Crustacea. granules. The jngal regions are granulous. The chelii)eds resemble very much the smaller cheliped of the males but are rather smaller in proportion. The abdomen is broadly elliptical and there is a line of granules on the basal segment. Two of these specimens give the following measurements : a of carapax. Breadth of carapax. Ratio. 21-8min 33-8 1 : 1-55 15-2 23-4 1: 1-54 Under the name of G. platydactylus, Saussure* mentions a species from iVIazatlan, Gulf of California, which I sliould refer to this species without hesitation, did he not state that the carpus was bitubereulate, a character which does not apply to any species of Gelaslmus which I have seen. Saussure's notice is as follows : " Gelaslmus 2olatydactijlus^ Latr. — Presque entierement semblable aux individus de Cayenne, si ce n'est que le carpe est bitubercule, et que la grande Crete du bras est dentelee, non entiere." G-elasimus platydactylus Edwards. f Cancer vocans major Herbst, Naturgeschichte der Krabben und Krebse, Band i, p. 83, Band iii, erstes Heft, p. 29, Tab. 1, tig. 1 1 (after Seba). ? Ocypoda heterochelos Bosc, Histoire naturelle des Crustaces, tome ii, p. 197, 1802. ? Gelasimus maracoani Desmarest, Considerations gonerales sur la Class des Crnstaces, p. 123, 1825, (non Latreille). Gelisimus phUydctylus Edwards, Histoire naturelle des Crust., tome ii, p. 51, 1837 ; Annales des Sciences naturelle, 3^^ serie, Zoologie, tome xviii, 1852, p. 144, pi. 3, fig. •-'. The synonymy of this species is in much confusion. Edwards quotes Ilerbst's and Seba's figui-es without query as belonging to his G. platydactylus and refers the Ocypoda heterochelos of Bosc to the G. viaracoaiii. Bosc's description however appears to have been drawn up from Herbst's or Seba's figure, and if these figui'es really belong to Edwards' species, the name heterochelos should be restored and the species should stand as Gelasimus heterochelos. The rough- ened or veiTucose character of the carapax in Ilerbst's figure is a marked feature which is not mentioned in either of Edwards' descrip- tions, so that it is quite likely that Bosc's heterochelos may be distinct from Edwards' species. Edwards gives Cayenne as the habitat of G. platydactylus. As described and figured by Edwards, this species difiers from G. princeps in having the superior crest of the merus of the larger * Description de quelques Crustaces nouveaux de la cote occidentale du Mexique. Eevue et Magasin de Zoologie, 2^ serie, tome v, 1853, p. 362. S. J. Smith on American Cmstacea. 123 cheliped entire, the hand much shorter and the fingers gaping for the whole length, and wanting the stout tooth on the prehensile edge of the propodus. G-elasimus maracoani Latreiiie. Maracoani, Marcgrave de Liebstadt, Ilistoire rerum naturalium Brasilise, figure. Ocypoda maracoani Latreiiie, Histoire des Crust, et Insectes, tome vi, p. 46, 1803. GonojHax maracoani Lamarck, Histoire naturelle des animaux sans vertebres, 2^ edit., tome V, p. 465. Gelasimus maracoani Latreiiie, Nouveau Dictionnaire d'Histoire naturelle, 2^ edit , tome xii, p. 517, 1817; Encyclopedie methodique, pi. 296, fig. 1; Edwards, His- toire naturelle des Crust., tome ii, p. 51, 1837; Annales des Sciences naturelles, 3>ne serie, Zoologie, tome xviii, 1852, p. 144, pi. 3, fig. 1; Dana, United States Ex- ploring Expedition, Crust, p. 318, 1852. Said to inhabit Cayenne and Brazil. Very likely two or more species are still confounded under the name of maracoani. Neither Edwards nor Dana mention any spines on the meral segments of the ambulatory legs, while in Latreille's figure in the Encyclopedie methodique there are short spines repre- sented on the posterior legs, Gelasimus armatus, sp. nov. Plate II, figure 5. Plate III, figure 4-4 ^i. Male. The carapax is only slightly convex and very little narrow- ed posteriorly, and the dorsal surface is naked and deeply areolated. The gastric and cardiac regions are smootli and shining, and the car- diac is large and very prominent. The branchial regions are promi- nent and their surfaces smooth but covered by very distinct, raised, vein-like markings which branch off in an arborescent manner from a conspicuous central trunk. The front is small, spatulate, contracted between the bases of the ocular peduncles and expanded below. The superior border of the orbit has a strongly raised margin, its edge is slightly sinuous and the antero-lateral angle prominent, the one on the side of the smaller hand being directed forward and the one on the side of the larger hand being more prominent than the other and di- rected strongly outward. The anterior part of the lateral margin is longitudinal, so that the breadth of the carapax is scarcely more be- tween the antero-lateral angles than a short distance posteriorly ; at the posterior extremity of this longitudinal portion, there are two small, but prominent, marginal tubercles, from which a graiudated line extends to the bases of the posterior legs, where there is another small rounded tubercle. The posterior margin is straight, smooth and 124 JS. J. Smith on American Crustacea. unarmed. The inferior margin of the orbit is armed with fifteen to eighteen slender, compressed and truncated teeth. The jugal regions are swollen and smooth, but their surfaces are veined somewhat as the regions above. The ocular peduncles are unequal in length, the one on the side of the larger cheliped being the longer, are very slender, but considera- bly enlarged at the corina, and shorter than the broad and open orbits which they only partially fill. The larger cheliped is enormously developed, the hand being high and lamellar, and exceeding, in length, twice the length of the cara})ax. The ischium is armed above and below with a small, marginal tuber- cle. The merus is smooth and rounded posteriorly, the anterior sur- face is flat and smooth, the inferior angle is armed with scatt^^red tu- bercles, and the superior angle rises into a low crest toward the distal portion, and is armed with slender tubercles. The carpus is smooth and rounded, but is armed with one or two small tubercles at the prox- imal extremity of the inner margin, and there are several low tuber- cles on the outer surface. The basal portion of the propodus is short; the inner surface is smooth and ixnarmed, except with a jjrominent tu- bercle near the middle, from which a line of obscure tubercles extends along the slight, oblique ridge to the inferior margin ; the outer sur- face is covered with very large, depressed, smooth tubercles which are separated by considerable spaces; and the inferior margin is thin and armed Avith dentiform tubercles. The digital portion of the propodus is thin and very broad toward the base ; the inner surface is smooth and somewhat concave ; the outer surface is flat and very coarsely punctate ; the inferior edge is denticvilate and slightly margined on the outside ; and the prehensile edge is straight, except a slight exca- vation at the base, is armed witli very small marginal tubercles and a high, tubercular, median ridge, and at the extremity, with a slender tooth. The dactylus is broadest toward the extremity ; the inner sur- face is concave and smooth ; the outer surface is flat and nearly smooth ; the superior edge is arcuate, thin and slightly denticulate ; the prehensile edge is straight, closes closely against the propodal fin- ger, except the slightly excavated portion at the base, and is armed with three lines of tubercles, like the propodal finger, except that the inner, mai-ginal line is separated from the median line by quite a wide space toward the tip, and that one of the tul)ercles, about two-fifths of the way from the base to the tip, is much larger than the rest ; and the tip is armed Avith a tooth projecting perpendicularly downward. In the smaller cheliped, the merus is slender and its anterior edge is S. I. Smith on American CrusUicea. 125 armed with three spimiles. The liaiid is slender, and the fingers are long, flattened at the tips, and the angles clothed with long hairs. The ambulatory legs are stout. The merus is smooth and unarmed in the lirst pair, but in the three last pairs, its posterior edge is armed with slender spines, — live in the second pair, six or seven in the third, and three short ones on the fourth or last. The abdomen is quite similar to that of G. princeps. Length of carapax, 25*2"""; breadth of carapax, 35 'S"""; ratio of length to breadth, 1 : r41. Total length of propodus in larger cheli- ped, 60-0'""\ Length of dactylus, 45 -6'"'"; breadth of dactylus, 11-8""". The only specimen of this species which I have seen is in the col- lection of the Peabody Academy of Science, and was obtained at the Gulf of Fonseca, West Coast of Central America, by J. A. McNiel. The larger hand in this specimen resembles very much the figure of the hand of G. maracoani given by Edwards in the Annales des Sci- ences naturelles, 3""^^ serie, tome xviii, 1852, pi. 3, fig. 1'', but the car- apax and ambulatory legs seem to be very difterent from that species, as neither Edwards nor Dana mention, in their descriptions of G. mar- acoani^ the peciiliar sculpturing of the branchial regions, the tuber- cles of the lateral margins or the spines of the ambulatory legs which are so conspicuous characters in G. armatus. In these characters it approaches the genus Acanthoplax, as described by Edwards. Gelasimus ornatus, sp. oov. Plate II, figure 9-9^ Plate III, figure 5-5<=. Female. The carapax is narrow and the greatest breadth is be- tween the antero-lateral angles, it is convex longitudinally, but only slightly laterally, and the dorsal surface is verrucose, some of the ver- rucae, especially on the branchial regions, being large and depressed. The regions are not swollen or protuberant, but the cervical and bran- chio-cardiac suture is very distinctly indicated. The front is narrow and spatulate, but only slightly expanded below" the bases of the ocu- lar peduncles. The superior border of the orbit is slightly and regu- larly arcuate, as seen from above, the margin is slightly raised and minutely denticulate, and the lateral angle projects forw^ard and out- ward as a slender and prominent tooth. The antero-lateral margin is longitudinal for a short distance anteriorly, but the posterior jjortion curves inward to the base of the posterior leg, and is ornamented with eight to ten bead-like tubercles. The latero-inferior, branchial regions are nearly vertical, and are divided by a granulated crest 126 S. I. Smith on American Crustacea. which starts a little way from tlie antcro-lateral angle and extends obliquely backward to the bases of the penultimate legs. The poste- rior margin is ornamented with a line of low tubercles. The inferior margin of the orbit is armed with about fifteen compressed and trun- cate teeth. The jugal regions are rough and sparsely clothed with short hairs. The ocular peduncles are equal in length, slender, slightly enlarged at the cornea and very little shorter than the broad and very open orbits. The chelipeds are like the smaller cheliped of G. armatus, except that the merus has but one spine and that the ischium has a slight tooth on the lower side next the articulation with the merus. The ambulatory legs are quite similar to those of G. armatus, but all of them have a tooth or spine on the lower side of the ischium, and the merus is armed in the first i)air with one or two spines, in the second with three, in the third with five, and in the last with two or three. The abdomen is broadly elliptical, and the basal segment is orna- mented with a line of small tubercles. Length of carapax, 26-6™'" ; breadth of carapax, 36-0'""" ; ratio of length to breadth, 1 : 1'35. The single specimen above described is in the collection of the Pea- body Academy of Science, and was brought home, with the G. arma- tus and several of the foregoing species, by J. A. McNiel, but unfor- tunately has no label to indicate the exact locality from which it came. It is however undoubtedly from some part of the west coast of Cen- tral America. This species is allied to the Acanthoplax insignis Edwards, but is at once distinguished from it by the verrucose dorsal surface of the carapax. It has also considerable affinity with G. armatus, and it is possible that it may be the female of that species, but this seems very improbable, when the great diflerences in the ornamentation of the carapax and in the armature of the chelipeds and ambulatory legs are considered. G-elasimus insignis. Acanthoplax insignis Edwards, Annales des Sciences naturelles, 3™e serie, Zoologie, tome xviii, 1 852, p. 151, pi. 4, fig. 23 ; Archives du Museum d'Histoire naturelle, Paris, tome vii, p. 162, pi. 11, fig. 1, 1854. Edwards states that this species was known to him only from a sin- gle, female specimen brought from Chili by M. Gay, but the figures which he has given in the Annales des Sciences and in the Archives S. I. Smith 071 American Crustacea. 127 du jNIuseum, differ so much that it would scarcely be supposed that they were intended to represent the same species, much less the same specimen. The only generic characters which are given by Edwards to distin- guish Adinthoplax from Gehtsinius, the proportions of the carapax and the tuberculation of the branchial regions, appear to me to be of slight importance. In the proportions of the carapax, the difference between Acanthoplax as figured in the Aunales des Sciences and the ordinary narrow fronted Gelasimi is scarcely, if any, greater than the difference between the two figures of A. inslgnis, for the figure of the carapax in the Annales is 19-0'"'" in length and 27'5'"'" in breadth, giv- ing tlie ratio of length to breadth, 1 : 1 '45, while the carapax in the figure in the Archives du Museum is 25-2™'" in length and 32-0"'"' in breadth, giving the ratio, 1 : r27, and this when both figures are stated to be of natural size. No measurements are given in the text in either place. The tuberculation of the branchial regions appears to be merely a character of ornamentation to which there is a consid- erable approach in the females of many of the large Gelasimi, and in the male G. armatus described in this article, there is a still closer approach to it. The armature of the ambulatory legs, however, may prove to be a character of some importance, and would unite in one group with A. insignis, G. ornatus and G. arinatus, and perhaps also G. niaracoani. B. — Species in ivliich all the segments of the abdomen are separated by distinct articula- tions, hut in luhich the front is broad and evenly arcuate between the bases of the ocular peduncles. Gelasimus palustris Edwards. (?) Cancer vocator Herbst, op. cit., Band iii, viertes Heft, p. 1, Tab. 59, fig. 1, 1804. Gelasimus vocans Edwards, Histoire naturelle des Crust., tome ii, p. 54; et Eegne an- imal de Cuvier, 3™^ edit, Crust., pi. 18, fig. 1 (teste Edwards). Gelasimus pahMtris Edwards, Annales des Sciences naturelles, 3™e gerie, Zooloo-ie tome xviii, 1852, p. 148, pi. 4, fig. 13. [Xon Cancer vocans Linne, Systema Naturae, editio xii, tome i, p. 1041). As figured in the Annales des Sciences naturelles, this species is quite different from any species which I have examined, and is distin- guished by the form of the terminal segment of the male abdomen which is as long as its breadth at base, with the sides straight and slightly divergent and the extremity broad and rounded, and by the anterior margin of the orbital border being symmetrical and not more rapidly curved above the base of the ocular peduncle than on the out- side, as it is in most of the allied species. It is described in the fol- 128 K I. Smith on A?nerican Crustacea. lowing brief terms : — " Crete sourciliere posterieure presque droite, I'anterieure tres courbe ; cretes marginales tres marguees sur les lobes mesobranchiaux. — Antilles." It is quite apparent that Edwards confounded at least two species under the name of palustris. The figure of G. vocans, which he has given in the Regne animal and which he refers to his pahtstris, evi- dently represents a different and distinct species, as the front is quite narrow, the basal portion of the propodus of the larger cheliped much longer in proportion and the terminal segment of the male ab- domen entirely diifereut in form. It is very likely the same as the G. vocafis of his Histoire naturelle des Crustaces, which is said to inhabit Brazil. Stimpson, in tlie Annales of the Lyceum of Natural History, New York, vol. vii, p. G2, refers the G. vocans of Dana and the G. minax of LeConte to the palustris of Edwards, and he evidently had more than one species before him, as he mentions tliat the tubercles on the outer surface of the larger cheliped were minute or obsolete in speci- mens from the Mexican and Central Amei'ican shores. G-elasimUS macrodactylus Edwards et Lucas. Voyage de d'Orbignj' dans FAmeriqiie meridionale, Crust, p. 27, pi. 11, fig 3, 184.3; Edwards, Annales des Sciences naturelles, ^^^ serie, Zool., tome xviii, 1852, p. 149. " Cotes du Valparaiso " (Edwards and Lucas). G-elasimus minax LeConte. Gelasimus minax John LeConte, On a new species of Gelasimus, Proceedings Acad- emy Nat. Sci., Philadelphia, vol. vii, 1855, p. 403. Gelasimus palustris {pars) Stimpson, Annals Lyceum Nat. Hist., New York, vol. vii, p. 62, 1859. Plate II, figure 4. Plate IV, figure l-l^ Male. The carapax is quite convex longitudinally and slightly transversely, and in large specimens the branchial regions are some- what gibbous above. The dorsal surface appears smooth, but is very minutely granulous, and there are a few small tubercles on the ante- rior part of the gastric region near the lateral margin. The front is broad and regularly arcuate. The posterior, or upper, edge of the superior orbital border is transverse and nearly straight, and has a smooth upturned margin. The anterior, or lower, edge is marked by a sharply raised and minutely denticulated margin which curves rap- idly downward above the base of the ocular peduncle, then gradu- ally upward and joins the posterior margin a little way from the an- S. J. Smith on American Crustacea. 129 tero-lateral angle, which is obtuse and not at all prominent. The lat- eral border is marked by a sharply upturned and finely denticulated margin, which is arcuate anteriorly so that the breadth of the carapax is considerably less between the antero-lateral angles than a little pos- tei-iorly, and the posterior portion is strongly incurved and terminates opposite the cardiac region. The posterolateral border is crossed by an oblique raised line or plication. The inferior orbital margin is finely toothed and the jugal region is rough and hairy. The larger cheliped is stout, and the length of the hand in large specimens is nearly or quite three times as great as the length of the carapax. The anterior surface of the merus is smooth, narrowly tri- angular in outline and its margins are nearly straight, the inferior armed with minute tubercles, and the superior with slender tubercles on the distal portion ; the upper surface is roughened with shorty irreg- ular, transverse rows of small tubei'cles. The superior surface of the carpus is covered with depressed tubercles, the proximal portion of the inner edge is tubercular and the inner surface is crossed by an ob- lique ridge armed with tubercles. The basal portion of the propodus is much shorter than the digital portion, and its superior and exterior siirface is covered with depressed tubercles, which are large and sepa- rated by smooth spaces on the upper portion, but below are smaller and crowded, and, along the inferior border, almost obsolete ; the inner surface is armed, on the inferior border, with a ridge of large tubercles extending from the base of the propodal finger obliquely upward to the border of the deep depression into which the carpus folds, and there are also a few tubercles between this depression and the base of the dactylus, and a line of tubercles extending upward, from the inner edge of the propodal finger, parallel to the base of the dactylus ; the superior edge is tuberculose and has a crenulated margin on the out- side and the inner margin is curved downward at the extremity of the depression into which the carpus folds ; and finally, the inferior edge is smooth and rounded, but with a slight margin on the outside. The propodal finger is nearly straight ; the inferior edge is smoothly round- ed, the prehensile edge is broad and ai-med with marginal lines of small tubercles, and a median one of irregular tubercles, of which one, about the middle of the finger, is very much larger than the rest ; and the tip has an excavation into which the dactylus fits. The dactylus is much curved, especially toward the tip, which hooks considerably by the tip of the propodal finger, and the prehensile edge is much as in the other finger, but the tubercles of the median line are nearly obsolete, except two or three large ones near the base, and as many more between the middle and the tip. Trans. Connecticut Acad., Vol. II. 9 March, 1870. 130 S. J. Smith on American Crustacea. The ambulatory legs are stout and very hairy along the edges, and the meral segments are quite broad, those of the posterior pair being nearly three times as long as broad. The abdomen is slightly narrowed at the first segment and is broad- est at the second and third. The distal margin of the penultimate segment is somewhat excavated for the reception of the terminal seg- ment, which is much narrower than the penultimate and broadest at the base, from which the margin is regularly arcuate, forming scarcely more than a semicircle. Both in alcoholic and dry specimens the points of the articulation of the merus with the cai'pus, the carpus with the propodus and the propodus with the dactylus, in the larger cheliped, are marked by red spots, and there are similar, but smaller, spots on the ambulatory legs, at the articulation of the meral with the carpal segments. The females differ from the males in being narrower and more evenly convex above, and in having the branchial regions more swollen and thickly covered with rounded tubercles. A number of specimens give the following measurements : — Locality. Sex. Length of carapax. Breadth of carapax. Ratio. Length of hand. Breadth of hand. New Haven, Ct. Male. 26-5inin 38-imm 1: 1-44 75-Omm 23-omm li ii It 22-9 340 1:1-48 61-0 20-8 U 1( 11 22-9 32-8 1:1-43 11 11 " 22-2 30-0 1:1-35 53-0 18-0 Bluffton, S. 0. 11 19-0 28-2 1:1-48 45-0 15-8 (1 11 It 17-6 25-2 1 : 1-43 40-5 14-8 II 11 It 17-2 24-5 1:1-43 40-0 14-2 New Haven, Ct. Female, . 24-9 34-3 1:1-37 11 11 11 21-8 29-2 1 : 1-34 This species is found at New Haven, Conn., on salt-marshes. There are specimens in the collection of the Peabody Academy of Science from Blufiiton, South Carolina, and also, from St. Augustine, Florida. LeConte's specimens were fi-om New Jersey. This is a very large species and I have not seen young specimens. It has perhaps been considered an adult form of G.pugnax; LeConte, however, recognized it as a distinct species and pointed out the differ- ences, having very naturally mistaken t\\Q pugnax for G.pugilator. The tubercles on the anterior portion of the branchial region of the male are probably only an adult character, but the very coarse tubercula- tion of the basal portion of the propodus and the red markings on the larger cheliped of the male, and the tubercular branchial regions of the female, are quite enough to distinguish it from the allied species. /S. I. Smith on American Crustacea. 131 GelasimiTs brevifrons stimpson. Annals Lyceum Nat. Hist, New York, vol. vii, p. 229, 1860. Of this species, which was found in a lagoon at Todos Santos, near Cape St. Lucas, Lower California, I have seen only a single, female specimen, which was kindly loaned from the collection of the Chicago Academy by Dr. Stimpson. As far as can be judged from the female alone, it is very distinct from any other species with which I am acquainted and seems to be most closely allied to G. minax. It differs from the female of Q. m,i- nax, in having the carapax broader in proportion and not nearly so much narrowed behind, and the dorsal surface less convex ; the cari- nae of the lateral margins ai'e more prominent and, from the form of the carapax, are not so much curved ; the front is shorter and more perpendicular, and the anterior margin of the orbital border is more convex, leaving a broader space between it and the posterior margin ; and finally, the meral segments of the ambulatory legs are much nar- rower in proportion, and are marked with conspicuous, transverse pli- cations. Length of carapax, 17"5"""; breadth of carapax, 25'0"'"^; ratio of length to breadth, 1 : 1*43. Gelasimus pngnax, sp. nov. Gelasimus vocans {pars) Gould, Report on the Invertebrata of Massichusetts, p. 325, 1841 ; G. vocans, var. A, DeKay, Natural History of New York, Crust., p. 14, pi. 6, fig. 10, 1844 (raore Cancer vocans Linne). Gelasimus pugilator LeConte, loc. cit., p. 403 {non Bosc). (?) Gelasimus palusfris {pars) Stimpson, Annals Lyceum Nat. Hist., New York, p. 62, 1859 {non Edwards). Plate II, figure 1, Plate IV, figure 2-2''. Male. The carapax is quite similar to that of G. minax but it is broader, the dorsal surface is smooth and there are no tubercles on the branchial regions, the front is narrower and projects farther down- ward, the anterolateral angle is sharp and the anterior part of the lateral margin is not at all, or only very slightly, arcuate. In the larger cheliped, the anterior surface of the merus is usually somewhat granular or finely tuberculose, especially along the inferior border, its outline is triangular and much broader toward the carpus than in G. m,inax, and the distal portion of the superior margin is high and arcuate and not tuberculated as in that species. The superior surface of the carpus is covered with small, rounded tubercles and the inner surface is crossed by an oblique, and more or less tuberculated, 132 >S. I. Smith on American Crustacea. ridge. The basal portion of the propodus, even in quite small speci- mens, is shorter than the digital portion and its superior and exterior surface is covered with small, depressed tubercles of unequal sizes and so thickly crowded together that there are scarcely any spaces between them, the oblique ridge on the inferior border of the inside is armed with numerous very small tubercles, the whole space between the upper portion of this ridge and the base of the dactylus is finely tuberculose, and the inferior edge is very distinctly margined on the outside. Both the propodal finger and the dactylus are more slender than in G. minax but offer no distinctive characters. The ambulatory legs are rather stout, very hairy along the edges of the carpal and propodal segments and the meral segments are broad, those of the posterior pair being about one and a half times as long as broad. The abdomen is scarcely at all narrowed at the basal segments. The terminal segment is very much as in G. minax but slightly broader in proportion and very similar to that of G. pugilator, figured by Edwards in the Annales des Sciences naturelles, 3™" serie, tome xviii, 1852, pi. 4, fig. 14'\ and not at all like his figure of G.palustris^ fig. 13'' on the same plate. The females differ from the males in being slightly narrower in proportion and in having the dorsal surface of the carapax more con- vex and minutely granulous. In life, the dorsal surface of the carapax of the male is very dark greenish olive, the middle and anterior portion, mottled with grayish white, the front, between and above the bases of the ocular peduncles, lio-ht blue varying somewhat in intensity in different specimens, and the anterior margin tinged with brown. The larger cheliped is lio-hter than the carapax, is marked with pale brownish yellow at the articulations and along the upper edge of the dactylus, and both finoers are nearly white along the prehensile edges. The exposed portions of the the ocular peduncles and the eyes are like the dorsal surface of the carapax. The smaller cheliped and the ambulatory leo-s are somewhat translucent and thickly mottled and sj^ecked with dai'k grayish olive. The sternum and abdomen are mottled ashy gray. The females differ from the males in having the dorsal surface of the carapax less distinctly mottled with whitish and in wanting the blue on the front. This description of the colors was taken, in November, from about a dozen specimens from New Haven. aS'. I. Smith on American Crustacea. 133 A series of specimens give the following measurements : — Locality. Sex. Length of carapax. Breadth of carapax. Ratio. Length of hand. Breadth of liand, New Haven, Conn. Male. 15.3mm 23-2ram 1 :l-52 mm mr u 11 14-8 22-6 1: 1-51 40-5 13-8 :i (1 " 14-4 21-9 1: 1-52 410 13-5 Bahamas. " 14-3 22-0 1: 1-54 39-5 13-4 New Haven, Conn. 11 l.S-8 20-7 1: 1-50 40-0 13-0 11 i( a 13-7 20-3 1: 1-48 370 12-4 K i: i( 12-8 19-3 1: 1-51 34-5 12"2 u 11 12-1 18-1 1:1-49 32-2 11-0 East Florida. " 10-6 16-6 1: 1-57 260 8-8 New Haven, Conn. • 1 10-4 15-5 1 : 1-49 22-0 8-3 East Florida. ti 10-3 15-7 1 : 1-52 24-5 8-6 u " 8-8 13-2 1: 1-50 15-2 6-5 Bahamas. 11 8-7 12-8 1:147 21-0 6-8 ■ 1 (( 7.4 11-0 1: 148 16-4 55 New Haven, Conn. 11 11 Female. 12-8 12-5 18-6 17-8 1 : 1-45 1 : 1-42 (1 11 (( 12-0 17-1 1 : 1-42 U 11 " 9-6 13-7 1: 1-43 Bahamas. 11 ii 8-6 7-3 12-4 10-2 1 : 1-44 1: 1-40 New Haven, Conn. " 7-0 10-0 1 : 1 -43 This species is common upon the salt-marshes about New Haven, Conn,, and there are specimens in the Museum of Yale College from St. Augustine, Florida (Col. W, E, Foster). In the collection of the Boston Society of Natural History there are specimens from Bahamas (Dr. Henry Biyant), and in the collection of the I^eabody Academy of Science, from Hayti (Dr. D. F. Weinland). At first sight this species might be mistaken for the youno- of G. minax, but when specimens of each, of nearly equal size, are compared there is no danger of confounding them. G. pvgnax is much smaller than G. minax, the carapax is considerably broader, is not so much contracted at the antero-lateral angles and is perfectly smooth, the tubei'cles of the outer surface of the larger cheliped are very much smaller and more crowded together, and the coloration is quite dif- ferent, the red on the chelipeds and ambulatory legs being entirely wanting. A male of tliis species, collected at New Plaven by W. C. Beecher, presents a remarkable anomaly in having tlie chelipeds nearly equal in size, while in other respects it is exactly like ordinary individuals. This specimen is briefly noticed in the American Naturalist, vol. iii, p. 557, ixnder the name of G. palustris. The left cheliped is exactly like the larger cheliped of ordinary specimens, and the right one 134 >iS'. I. Smith on American Crustacea. differs only in being, somewhat smaller and in having the fingers slightly more incurved at the tips so as to fit nicely the buccal area. Length of carapax, 11-2"""; breadth of carapax, 16"4""" ; rato, 1 : 1-46. Length of left cheliped, 25 -0'"'". Length of right cheliped, 21-0""", The specimen, which was examined while alive, was very active and used both hands with equal facility. With this single remarkable exception, I have found only the slightest variations in examining carefully more than a liundred specimens, G-elasimus rapax, sp. nov. Plate II, figure 2. Plate IV, figure 3. Male. The carapax is very much like that of G. pugnax^ but the front is narrower, the upper edge of tlie superior orbital border is sinuous and not so transverse as in that species, being directed some- what backward, the border itself is wider and its lower edge is not so abruptly curved above the base of the ocular peduncle. In the larger cheliped, the anterior surface of the merus is smooth. The superior surface of the carpus is minutely tuberculose and the inner surface is crossed by a slight, oblique ridge which is nearly smooth. The basal portion of the propodus is much stouter tlian in G. pugnax and considerably longer than the digital portion, the superior and exterior surface is thickly covered with small tubercles and the inner surface is much as in G. pugnax, but the superior margin is curved more abruptly, and farther downward at the extrem- ity of the depression into which the carpus folds, and there is a line of bead-like tubercles, along the border next the base of the dactylus, which are very much larger than in G. pugnax. The projiodal finger is short and stout and considerably curved upward, the inferior edge is smooth and rounded, and the prehensile edge is much as in G. pugnax, but the tubercles are larger. The dactylus is slout, curved toward the extremity and the tip hooked by tlie end of the other finger, the superior margin is tuberculose toward the base and mar- gined on the outside for nearly half its length, and the prehensile edge is as in G. pugnax but there are four or five large tubercles close together near the base. The ambulatory legs are quite similar to those of G. pugnax but seem to be much less hairy. The abdomen is as in G. pugnax. Length of carapax, 12-6'"™; breadth of carapax, 190'""'; ratio, 1 : 1-51. Length of hand, 28-2""'' ; breadth of hand, 10-8'""\ S. 1. Smith on American Crustacea, 135 I have seen but a sinole specimen,' which was collected at Aspinwall by F. H. Bradley. Although closely allied to G.minax andpuffnax, it is vei-y different from any specimens which I have seen, of either of those species, and is readily distinguished from them by the very short and stout fingers, the tubercles on the basal portion of the upper mar- gin of the dactylus, the long basal portion of the propodus and the line of bead-like tubercles along its border next the base of the dactylus. The differences in the carapax are however very slight, and it may possibly prove to be a variety of G. pugnax. G-elasimus mordax, sp. nov. Plate II, figure 3. Plate IV, figure 4, 4\ Male. The carapax is convex both transversely and longitudinally The dorsal surface is jmnctate and the space between the puncta is smooth and naked, but the puncta themselves give rise to short hairs which are very easily removed. The front is much less deflexed than in the allied species, its dorsal surface is divided by a distinct median sul- cus and its inferior surface, between the margin and the epistome, is quite high. The upjjer edge of the superior orbital border is directed somewhat backward as in G. rapax, but is straight and not sinuous ; the border itself is much more oblique than in the allied species, so that it appears very large as seen from above. The anterior part of the lateral margin is thin and projects somewhat laterally. In the larger cheliped, all the segments are more elongated than in the allied species. The anterior surface of the merus is smooth, nar- row in outline and its margins are tuberculose. The superior and exterior surface of the carpus is obscurely tuberculose, and its inner surface is crossed by an oblique ridge which is nearly smooth. The basal portion of the propodus, as seen in front, is narrowed toward the articulation of the carpus and is very much shorter than the digital portion ; the superior, and the upper part of the exterior, surface is obscurely tuberculose while the lower portion is smooth ; the oblique ridge on the inferior border of the inside, is much higher and extends farther back toward the articulation of the carpus than in the allied species, and is tliickly covered with very large, rounded tubercles, and all the space between its upper portion and the base of the dactylus is covered with depressed tubercles; the superior edge is somewhat carinated, slightly tuberculose and margined on tlie outside, and the inner margin is turned abruptly downward at the extremity of the depression into which the carpus folds ; and finally, between this abruptly curved portion and the base of the dactylus and just 136 S. J. Smith on American Crustacea. below the superior margin, there is an oblong, depressed space which is very conspicuous as seen from above. This dejDression exists in 6r. minax but is not at all conspicuous. The propodal finger is very long and slender, curved upward at the extremity, and the prehensile edge ai'med with a large tubercle near the middle and another near the tip, which is deeply excavated for the reception of the dactylus. The dactylus is very slender, the basal portion nearly straight, the extremity strongly hooked downward and inward, the superior edge smooth, and the prehensile edge armed with several large tubercles. The ambulatory legs are long and much more slender than in the allied species, the meral segments being quite narrow. The abdomen is quite similar to the abdomen of G. pugnax, but is somewhat narrower. The females differ from the males in having the carapax narrower and more convex, and in the branchial regions being tuberculose along the lateral margins. Several specimens give the following measurements : — Sex. Length of carapax. Breadth of carapax. Ratio. LenRth of hand. Breadth of hand. Male. 16-9mm 25-5mm 1: 1-51 45-Omm 12-5mm 11 15-4 23-2 1:1-51 45-0 13-0 11 15-3 23-0 1: 1-50 46-5 13-0 11 14-5 21-5 1 : 1-48 42-0 12-6 " 10-6 lo'5 1: 1-46 20-5 7-0 emale. 12-9 18-1 1: 1-40 « 12-5 16-7 1: 1-34 11 10-8 14-3 1 : 1-32 " Canals at Para, South America, October or November, 1858 ; Caleb Cooke" (Collection Peabody Academy of Science). G-elasimus pugilator LatreOie. Ocypoda pugilafor Bosc, Histoire naturelle des Crust., tome i, p. 197, 1802; (pars) Say, Journal Academy Nat. Sci , PhQadelphia, vol. i, p. 71, 1817, p. 443, 1818. Gelasimus pugilator LatreiUe, Nouveau Dictionnaire d'Histoire naturelle, 2 ^ edit, tome xii, p. 520, 1817; Desmarest, op. cit., p. 123 ; Edwards, Annales des Sciences natu- relles, 3"^e serie, Zoologie, tome xviii, 1852, p. 14, pi. 4, fig. 149; Stimpson, Annals Lyceum Nat. Hist., New York, vol. vii, p. 62. Gelasimus vocans, DeKay, Natural History of New York, Crust., p. 14, pi. 6, fig. 9 ; (pars) Gould, Report on the Invertebrata of Massachusetts, p. 323 (non Cancer .vocans Linne). Plate IV, figure 7. This IS at once distinguished from any of the east coast species, except G, subcj/lindricus, by the rectangular outline, swollen and S. I. Smith on American Onifitacea. 137 highly polished, dorsal surface of the cara})ax, and by the inner sur- face of the basal portion of the propodus of the larger cheliped being evenly rounded and beset with small scattered tubercles, but with no indication of an oblique tuberculose ridge. From G. suhcylindricus, it is readily distinguished by the carapax being narrower and its pos- terior margin straight, by the hand in the larger cheliped of the male being margined with a slight crest on the outside of the superior edge, and by the narrow male abdomen. It seems to be abundant from the Giilf States to Massachusetts. At New Haven, Conn., it is very common upon muddy beaches, but is not usually associated with G. pugnax, which prefers salt-marshes. There are specimens in the Museum of Yale College, collected at Egmont Key, West Florida, by Col. E. Jewett, and at St. Augustine, by Col. W. E. Foster and H. S. Williams ; and in the collection of the Peabody Academy of Science, there are specimens from Savan- nah, Georgia, from Bluffton, South Carolina, and from Nantucket, Massachusetts, those from the last locality collected by Dr. A. S. Packard, Jr. A series of specimens give the following measurements : — Locality Sex. Length of carapax. Breadth of carapax. Ratio. Length of hand. Breadth of hand. West Florida, Male. 15-Omm 21-6™ni 1: 1-44 38-omm 12-5mm 11 " u-t 210 1 : 1-43 33-0 10-5 New Haven, Conn. 11 14-2 20-6 1 : 1-44 36-5 11-8 " 11 a 13-6 194 1 : 1-43 340 11-0 a 11 " 13-4 18-8 1 : 1-40 30-2 11-4 u 11 It 12-5 17-4 1: 1-39 27-0 10-6 11 11 " 11-7 16-2 1: 1-38 23-8 9.6 11 i( 11 7-6 10-2 1: 1-33 9-5 4-8 "West Florida. Female. 14-6 20-4 1 : 1-40 New Haven, Conn. 11 11 a 12-5 10-8 16-4 14-3 1: 1-31 1 : 1-32 K 11 .1 91 12-0 1 : 1-32 G-elasimus snbcylindricus Stimpson. Annals Lyceum Nat. Hist, New York, vol. vii, p. 63, 1859. Plate IV, figure 6-6''. This species has a general resemblance to G. pugilator, but the body is much broader, not so much narrowed behind and veiy con- vex, being in foct much like G. gibhosus. The male abdomen and its appendages are, moreover, very unlike any other species wliich is known to me. 138 S. J. Smith on American Crustacea. Male. The dorsal surface of the carapax is minutely granulous, very convex longitudinally and swollen along the branchial regions, which, however, do not project above the middle of the carapax, and the regions are not separated by distinct sulci. The front is evenly rounded and strongly deflexed. The superior border of the orbit is nearly perpendicular, and its posterior, or upper, margin is sinuous, curving forward in a slight prominence in the middle. The antero- lateral angle is obtuse and not at all prominent. The lateral margins converge slightly anteriorly and are only feintly indicated on the posterolateral border. The posterior margin is divided into two broad lobes by a very marked median immargination. The inferior border of the orbit is slightly curved and finely denticulate. The external maxillipeds are proportionately smaller than in the allied species, the ischium is only very slightly wider than the merus and its outer margin is nearly straight. Corresponding with the form of the external maxillipeds, the buccal opening is smaller and more rectangular than in the other species. In the larger cheliped, the angles of the merus are obtuse and granulous and the anterior surface is slightly convex. The outer sur- face of the carpus is slightly granulous. The basal portion of the propodus is nearly as long as the digital portion ; the inner surface is not armed with a tuberculose ridge along the inferior margin, that portion being rounded and only obscurely tuberculose, but on the border next the base of the dactylus, there are two, sharp, tubercular, parallel ridges, the inner one highest and separated from the other by a deep, narrow groove ; the outer surface is densely covered with small, depressed tubercles which are more uniform in size and more promi- nent than in G. piignax, or G. pugilator / the superior edge is tuber- culose but not distinctly margined on the outside as in G. nimax, pt(g}iax, and pi/gilator ^' the inferior edge is armed with a prominent, tubercular margin on the outside, and the flat, oblique space between the inner and outer margins is smooth and shining, while in G. pugil- ator it is covered with rounded granules. The propodal finger is considerably curved upward, its outer surface is armed, on the basal portion, with a distinct, median ridge, the inferior margin is smooth, and the prehensile edge tubercular and armed with a single, large tooth near the middle. The dactylus is strongly and evenly curved, the superior margin is smooth and the prehensile edge is tubercular and armed with several larger tul)ercles toward the base. The smaller cheliped and the ambulatory legs do not differ notably fi'ora those of the allied sj^ecies. ->. 1. Smith on American Crustacea. 139 The abdomen is very broad, its breadth being fully equal to two- thirds its length, while, in G. pugilator and allied species, the breadth is not equal to more than half the length. The terminal segment is very small, being rather less than half as broad as the penultimate and very much shorter than broad. The appendages of the first seg- ment are very stout and nearly straight organs, reaching to the middle of the penultimate segment, and the tips are horny and slightly hairy, while in G. pugilator these organs are longer, very slender, and strongly curved outward at the tips. The female difi^ers from the male in having the posterior margin of the carapax only slightly immarginate in the middle. Sex. Length of carapax. Breadth of carapax. Ratio. Length of hand. Breadth of hand. Male. 12imm 18-5mm 1 : 1-53 25-Omm 11-Omm II 105 160 1: 1-52 20-5 90 Female. 100 15-5 1:1-55 The above description and measurements were made from three of the original specimens, collected at Matamoras on the Rio Grande, by M. Berlandier, and loaned by Dr. Stimpson. Gelasimus Stenodacylus Edwards et Lucas. Voyage de d'Orbigny dans I'Araerique meridionale, Crust., p. 26, p. 11, fig. 2, 1843 ; Edwards, Annales des Sciences naturelles, 3me serie, Zool., tome xviii, 1852, p. 149. " Trouve sur les cotes du Valparaiso par M. d'Orbigny," (Edwards and Lucas). In the Annales des Sciences naturelles, Edwards gives the habitat as, " Chili, Bresil," but there is very likely some mistake in regard to the latter locality for very few, if any, species of Crusta- cea are common to Chili and Brazil. Gelasimns Panamensis stimpson. Annals Lyceum Nat. Hist, New York, vol. vii, p. 63, 1859. Plate IV, figure 5. Stimpson had only the young of this species and did not give the characters of the larger cheliped of the male, but a good series of specimens collected at Panama by Mr. Bradley, shows that it is very difierent from any of the east coast species and is not allied to any from the west coast, unless it be to G. stenodcatylus which I have not seen. Male. The carapax is broadest between the antero-lateral angles and is much less convex than usual. The dorsal surface is very minutely granulose, and there are a few coarse granules or small tubercles on the front and on the anterior part of the branchial region 140 S. I. Smith ou Anierlcan Crustacea. near the lateral margin. The upper edge of the superior orbital border is sinuous and the border itself is quite narrow. The antero- lateral angles are sharp and project prominently forward. The inferior orbital margin is thin and sharply dentate and its outer angle is jDrominent and angular, and is separated from the superior margin by a deep and broadly rounded sinus. In the larger cheliped, the merus is slender, and its anterior surface is narrow and smooth and the mai'gins are unarmed and rounded. The carpus is evenly rounded and nearly smooth externally. The basal portion of the propodus is smooth or microscopically granu- lose and flat and entirely unarmed within ; the depression into wliich the carpus folds is very short, not extending half way to the base of the dactylus ; and the superior and inferior margins are evenly rounded. The propodal finger is slightly upturned at the tip, the inferior edge is perfectly smooth and evenly rounded, and the tuber- cles of the prehensile edge are nearly obsolete except a large de- pressed one near the middle. The dactylus is strongly curved down- ward at tip, the superior edge is smooth and rounded and the pre- hensile edge is obscurely tubercular In very young specimens the hand is quite granulose above but becomes smooth with age. In the smaller cheliped the tips of the fiugers are densely clothed with soft hair. The ambulatory legs are slender, smooth and almost entirely naked. The females differ from the males in the carapax being a little narrower in proportion, and in the branchial regions being slightly inflated and more granular or even tuberculose. Several specimens give the following measurements : — Locality. Sex. LeiiKth of carap.ix. Breadth of carapax. Ratio. Lcngtli of liand. Bmadth of liauu. 'anaina. Male. 12-5'nm IS-Qram 1: 1-44: 27-5mm 9-4mni •• 11 12 1 18-0 1: 1-49 32 11-0 II 11 8-3 11-1 1: 1-34 9-4 4-8 (1 Female. 13-6 18-5 1 :l-36 (1 11 12-2 17-0 1: 1-39 11 ■ 11 11-5 16-0 1: 1-39 ti (1 Q-t 13-8 1 : 1-42 C. — Species in which the fourth, fifth and sixth segmtnts of the male abdomen completely anchylose, and in which the carapax is very transverse, and the branchial regions are gibbons. Gelasimns gibbosns, sp- nov. Plate II, figure 11. Plate IV, figure 8. Male. This is a small species quite different in general appearance from any of the foregoing. The body is very short and broad, very S. I. Smith on American Cuustacea. 141 little contracted behind, and, in general form, a short cylinder trun- cated at each end. The chelipeds and ambulatory legs are slender and elongated. The dorsal surface of the carapax is naked, smooth and shining, convex longitudinally, deeply areolated and nearly symmetrical. The cervical sutiire is slightly curved and very distinctly marked by a deep sulcus. The median portion of the gastric region is triangular, and is separated from the antero-lateral lobes by very distinct but shallow sulci, which meet in an acute angle on tlie front. The cardiac region is large, quite prominent and distinctly separated from the gastric. The branchial regions are very prominent and swollen, pro- jecting much above the median regions, and a narrow portion next the cervical suture is cut oft* by a straight and sharp sulcus. The front projects well forward and is quite narrow, but not contracted between the bases of the ocular peduncles. The superior border of the orbit is nearly on a plain with the anterior part of the carapax, its anterior edge is strongly arcuate and is marked by a very slight, but sharply raised and continuous margin, and the posterior edge is marked by a faintly raised line, which is transverse and nearly straight toward the front, but, toward the side of the carapax, falls off posteriorly, so that the antero-lateral angle, which is right-angular, but not at all prominent, is considerably posterior to the rest of the anterior margin. The faintly margined lateral borders are parallel anteriorly but approach slightly posteriorly. The inferior border of the orbit is denticulate, the teeth being very minute on the portion toward the front but much larger, and very slender on the outer portion, and round into the external hiatus. The jugal regions are much swollen and are separated from the buccal area by a deep depression. The ocular peduncles are quite stout and as long as the orbits, which they nearly fill. The ischial segments of the external maxillipeds are very broad and the outer edges are arcuate to fit the expanded buccal area, and thus resemble the species of section A. The larger cheliped is remarkably developed for so small a species, the merus being as long as the carapax, while the hand is almost three times as long, and nearly twice as long as the breadth of the cai-apax. The anterior surface of the merus is smooth, flat and quite narrow, and its angles are smooth and unarmed. The superior and exterior surface of the carpus is evenly rounded and very slightly granulous, and the inner margin is sharp and dentate. The basal portion of the 142 S. I. Smith on American Crustacea. propodus is short and compressed, the outer surface is flat and granu- lous, the inferior edge is angular and has a very slight, granular margin on the outside, the superior edge is rounded and granulated, and the inner surface is armed with a slight, oblique, tuberculose ridge extending from the inferior edge to the shoi-t depression into which the carpus folds. The digital portion of the propodus is much compressed, straight and very slender, the inferior edge is nearly smooth, the prehensile edge is only very obscurely tuberculate and has a single, very slight tooth near the middle, and the tip is slender, acute and slightly upturned. The dactylus is compressed, very slender, straight for two-thirds its length and the terminal portion regularly curved downward, the superior edge is rounded and slightly granulous toward the base, and the prehensile edge is as in the other finger, except that the tooth is smaller and nearer the base. The smaller cheliped is smooth and unarmed, the merus is slender and triquetral, the carpus is short and rounded, the basal portion of the propodus is quite short and thick, and the fingei's are slender. The ambulatory legs are long, very slender and nearly naked, and the meral segments are very narrow. The sternum is very broad and very convex. The abdomen is scarcely at all contracted at the second segment, and it tapers slightly to the extremity of the sixth ; the first and second are very short, the the third is about twice as broad as long, the fourth, fifth and sixth are completely anchylosed into one piece, and the seventh, or last, forms very nearly a semicircle. Length of carapax, S'S"'"' ; breadth of carapax, 14-4'"'" ; ratio, 1 : 1*79. Length of hand, 24-8™"^ ; breadth of hand, 8 •2""". I have seen only one specimen, which was collected at the Gulf of Fonseca, west coast of Central America, by J. A. McNiel (Collection Peabody Academy of Science). Family, Gecarcinid^. Cardiosoma Latreille. In this genus the abdominal appendages of the male present, in some cases at least, good specific characters. In all the species which I have examined, the appendages of the first segment are very stout and nearly straight organs reaching beyond the middle of the abdo- men, articulated at their bases Avith a large and hard semicircular plate, which arches round the intestinal canal and joins the abdomen on each side, and armed at their extremities with slender, horny tips. S. J. Smith on American Crustacea. 143 The appendages of the second segment are small and inconspicnous, and their slender tij^s are flexiljle and folded within a little groove on the inside of the bases of the appendages of the first segment. Cardiosoma guanhumi Latreiiie. Cardisoma g-MawAw^m Latreiiie, Encyclopedie methodique, tome x, p. 685, 1824, {teste Edwards); Edwards, Histoire natureUe des Crust., tome ii, p. 24, 1837; Regne animal de Cuvier, 3™^ edA.\,., pi. 20, fig. 1 ; Annales des Sciences naturelles, S^ie serie, Zoologie, tome xx, 1853, p. 204, pi. 9, fig. 1 ; Gibbes, On the Carcinological Col- lections of the United States, Proceedings American Association, 3d Meeting, p. 179, 1850; Stimpson, Proceedings Academy Nat. Sci., Philadelphia, 1858, p. 100; Saiissure, Crustaces nouveaux des Antilles et du Mexiqiie, p. 21, 1858. Ocypode {Cardisoma) cordata DeHaan, Fauna Japonica, Crustacea, p. 27, 1835 [non Cancer cordatus Linne). Ocypoda ruricola Freminville, Annales des Sciences naturelles, 2e serie, Zoologie, tome lii, 1835, p. 217 (non Cancer ruricola Linne). Ocypoda gigantea PreminviUe, loc. cit, p. 221, 1835. Plate y, figure 3. The abdomen of the male is broadest at the third segment, from which the margins converge rapidly to the sixth, which is considerably- longer than broad. The terminal segment is narrow and its extremity is rounded. The first pair of abdominal appendages reach to the middle of the sixth segment, are triquetral, straight and stout, and their tips are rounded and slightly flattened laterally, and each is armed with a very small, scale-like appendage directed obliquely out- ward, and on the upper edge, just above this appendages, there is a small process which is straight and does not reach beyond the rounded extremity of the thickened portion of the organ. A male from the Florida Keys gives, length of carapax, 65""" ; breadth of carapax, 78'""'; ratio of length to breadth, 1:1-20. Length of merus in right cheliped, 31""°; in left cheliped, 49"'"\ Length of right hand, 45""" ; breadth, 19. Length of left hand, 88'"™ ; breadth, 44. Cardiosoma quadratum Saussure. See these Transactions, vol. li, p. 16. Plate V, figure 4. In this species the male abdomen and its appendages are almost ex- actly like those of C. guanhumi except that the hoiny extremities of the appendages of the first segment are a little longer and more slen- der. There is a remarkable difference between the male abdominal appendages of this species and the species from the west coast of 144 S. I. Stnith on Amerlcmi Crustacea. Africa, with which it is compared on page 16 of this volume. In the African species the first pair of these appendages are very much like those of the following species, the horny tips being long, slender and somewhat spiral, and the process on the iipper edge extending much beyond the thickened portion of the organ. Cardiosoma crassum, sp nov. Plate V, figure 5. In general appearance this species is closely allied to C. qiiadratum. The carina of the lateral margin of the carapax is, however, much more strongly mai'ked and the ambulatory legs are clothed with long hair, while in G. quaclratum they are nearly naked. The male abdom- inal ai:)pendages are entirely unlike in the two species. The dorsal surface of the carapax is naked, very minutely granu- lous, regularly and strongly convex longitudinally, but only slightly transversely, and the areolation is not strongly marked, the cardiac region and the median portion of the gastric alone being indicated ; the anterior extremity of the mesogastric lobe, however, is distinct, long and slender and reaches nearly to the front. The front is broad and high and the epigastric lobes protuberant, leaving, between them and the front, a depressed space which is thickly covered with coarse granules. The superior margin of the orbit is slightly sinuous, as seen from above, and the lateral angle projects forward as an angular tooth. Just back of this tooth the antero-lateral margin is broken by a sharp notch, above which the carina of the lateral margin begins in a sharp prominence. This carina through its entire length is very high and distinct, being much more strongly marked than in C. quadratum. The epistome and nasal lobe are very much as in C. quadratum, but the labial border of the epistome is armed with a line of granules which is more sharply raised and composed'of smaller granules than in that species. The jugal regions are densely clothed with short, soft hair. The inferior branchial regions are naked, but are roughened with numerous, short, sharp rugte. The chelipeds are very unequal in both sexes, and the ischial seg- ments are armed, on the anterior side, with a few small tubercles. In the larger cheliped, the merus is triquetral, very stout and reaches slightly beyond the lateral margin of the carapax, the anterior sur- face is flat and both its margins are armed with very large and prom- inent tubercles directed forward, and on the outer surface and the pos- terior angle, which is obtuse, there are short granulous rugre which are very conspicuous on the angle. The larger hand is very short and /S. I. Smith on American Crustacea. 145 stout, the breadth being about equal to four-sevenths of the length ; the outer surface of the propodus is flattened and smooth ; the inner surface, in the middle and toward the base of the dactylus, and the margins, are armed with scattered tubercles ; and finally, the fingers are very stout, the outer edges are armed with small horny tubercles, and the prehensile edges gape but slightly, and are armed with large, irregular teeth. In the smaller cheliped, the merus is more slender and does not quite reach the lateral margin of the carapax, and the hand is very much smaller and more slender. The ambulatory legs ai-e stout and the carpal and propodal seg- ments, and the meral on the angles below, are clotlied with long black hairs, which are very conspicuous and fasciculated on the carjial and propodal segments of the first and second anterior pairs. In the male, the abdomen is broadest at the third segment, from which the margins converge regularly to the sixth, which is nearly or quite as broad as long and only slightly narrowed for most of its length, but sharply contracted just before the articulation with the small and narrow terminal segment. In the female, the abdomen is broadest near the articulation of the fifth with the sixth segment, and the margins of the sixth segment are arcuate and converge rapidly to the small, obtusely triangular terminal segment. The first pair of male abdominal appendages reach to the middle of the penultimate segment of the abdomen, and their extremities are slightly flattened laterally, thickly clothed with hair on the out- side and terminated by a long, slender, hard and horny tip, which curves outward for nearly half its length, then rapidly iipward, and again outward at the end, forming thus about the third of a very elongated spiral. From the under edge, just below the base of this horny tip, there is a stout, straight process, which is soft and flex- ible, and clothed at the extremity with hair. Four specimens give the following measurements : — Sex. Male. ngth of carapax. Breadth of carapax. Katie. - 50-7mni 62-Omm 1:1-22 54-0 66-3 1: 1-23 - 56-4 68-0 1: 1-21 530 64-5 1: 1-22 Female. I have examined a large number of specimens collected at the Gulf of Fonseca, west coast of Central America, by J. A. McNiel, and in the Museum of the Peabody Academy of Science. Traks. Connecticut Acad., Vol. II. 10 April, 1870. 146 S. I. Smith on American Crustacea. Family, Bosciad.e. Pseildothelphusa Saussure. Potamia Latreille, Cours d'eutomologie, p. 338, 1831 {teste Edwards); Edwards et Lucas, Voyage de d'Orbigny dans I'Amerique meridionale, Crust., p. 22, 1843; White, List of the Crustacea iu the British Museum, p. 30, 1847 ; Dana, United States Exploring Expedition, Crust, p. 293 ; Saussure, Crustaces noveaux des An- tilles et du Mexique, p. 19, 1858 {iion Robineau-Desvoidy). Boscia Edwards, Histoire natureUe des Crust , tome ii, p. 14, 1837 ; Annales des Sci- ences naturelles, 3™® serie, Zoologie, tome xx, 1853, p. 207 ; A. Edwards, Annales de la Societe entomologique de France, 4™e serie, tome vi, 1866, p. 203. Pseudothelphusa Saussure, Revue et Magasin de Zoologie, 1857, p. 305 {teste Saus- sure). Latreille's name, Potamia, given in 1831, was properly rejected by Edwards on account of its previous use, in 1830, by Robineau-Des- voidy, for a genus of Diptera, but the name Hoscia, proposed by Edwards in 1837, is quite as objectionable, having been used, accord- ing to Agassiz's Nomenclator Zoologicus, by Leach, in 1813, for a genus of Cimpedia, by Schweigger, in 1820, for a genus of Polyps, and by Leach again, in 1824, for a genus of Coleoptera. Pseudothel- phusa, although at first proposed as a new genus, does not dilFer es- sentially from the species of Edwards' JBoscia which have no superior frontal crest, and was finally united with Potamia by Saussure him- self, so that it may properly be adopted for the genus as defined by Edwards. Pseudothelphusa, as here limited, includes the following American species : — P. Americana Saussure, from Hayti. P. gracilipes {Poscia gracilipes A. Edwards, Annales de la Societe entomologique de France, 4""^ serie, tome vi, 1866, p. 204), from Haute Vera-Paz, Gautemala. P. plana, sp. no v., from Peru. P. rnacropa {Boscia macropa Edwards, Archives du Museum d'Histoire naturelle, Paris, tome vii, p. 175, pi. 12, fig. 3), from Bo- livia. P. Chilensis {Potamia Ghilensis Edwards et Lucas, Voyage de d'Orbigny dans I'Amerique meridionale. Crust,, p. 22, pi. 10, fig. 1), from Lima, Peru. P. denticulata {Boscia denticulata Edwards, Annales des Sciences naturelle, Zoologie, 3"'*^ serie, tome xx, 1853, p. 208), from Guiana. P. Bocourti, {Boscia Bocourti A. Edwards, loc. cit., p. 203), from the River Coban, Haute Vera-Paz, Gautemala. S. T. Smith on American Crustacea. 147 P. dentaia {Boscla dentata Edwards, Histoire naturelle des Crust., tome ii, p. 15, pi. 18, fig. 14-16), from the West Indies. The only other described species is the jR sinutifrons [Boscia sinu- tifrons A. Edwards, loc. cit., p. 205), the habitat of which was not known. Pofamia latifrons Randall (Journal Academy Nat. Sci., Philadel- phia, vol. viii, p. 120, 1839), supposed to have come from Surinam or the West Indies, probably belongs here, but the description is too indefinite to determine its aflSnities with any degree of certainty. Pseudothelphusa plana, sp. nov. Female. The carapax is very broad and its dorsal surface is flat in the middle and posteriorly, but convex along the anterior border, and is punctate, but the surface between the widely separated punctures is glabrous. The gastric region is undivided, except by a short and shallow median sulcus, which separates the slightly indicated anterior lobes and extends down the front. The anterior portion of the cer- vical suture, from the median lobes of the gastric region to the antero- lateral margin, is well indicated by a straight, broad and deep sul- cus. There is no sulcus between the gastric and hepatic regions. The branchial regions are very prominent and undivided. The front is deflexed and the narrow inferior margin is perpendicular, and has a distinct submarginal groove. The orbits are well filled by the stout ocular peduncles. The antero-lateral margin is evenly and very strongly arcuate, and its edge is sharp and finely denticulated. The postero-latei'al margin is concave in outline. The external maxillipeds, as well as the sternum, are punctate like the carapax but the punctures are much larger. A single cheliped is quite small ; the merus scarcely reaches beyond the carapax, is triangular, the anterior angle slightly dentate, and the posterior angle rounded and granulated ; the upper side of the carpus is punctate like the carapax, evenly rounded and armed with an angu- lar tooth on the inner margin ; the basal portion of the propodus is punctate, slender and evenly rounded ; and finally the fingers are long, slender, cylindrical, nearly stiaight, and slightly toothed within. The ambulatory legs are naked, slender and rounded, and the dactyli are nearly straight, cylindrical and sparsely spinulose. The color of alcoholic specimens is uniform dark olive brown above and lighter beneath. Sex. Length of carapax. Female. IS-emm " 16-5 wJth of carapax. Ratio. 22-4nam 1: 1-65 27-7 1:1-67 148 S. I. Smith on American Crustacea. Tliere are two rather badly preserved specimens, collected at Paita, Peru, by Prof. Jaraes Orton, in the Museum of Yale College. The smaller specimen wants both chelipeds, and the larger specimen, one. This species is closely allied to _P. macropa, but is easily distin- guished from it by the denticulated antero-lateral margin, by the short merus of the chelipeds, and by the flattened carapax — the carapax of _P. macropa being represented in Edwards' figure as quite convex transversely, while in P. plana it is flat in that direction. Moreover the front seems to be much more deflexed in our species, the orbits are much smaller and are well filled by the eyes, and the antero-lateral margin is not " creuses en dessous d'un sillon bien marque." In the depressed form of the carapax, it is apparently closely allied to JP. gracilijyes, but the ambulatory legs are not longer in proportion than in jP. macropa., and the front is almost straight, as seen from above, and not lobed as in P. Americana, with which the tront of P. gra- cilipes is compared. In the denticulated antero-lateral margin it re- sembles P. C'hilensis, but in the form of the carapax, and in other characters it is much nearer to P. macropa. Opisthocera,* gen. nov. The carapax is much as in PseudothelpJuisa ; the dorsal surface is not distinctly areolated ; the front is deflexed, smooth and unarmed, and the edge is not reflexed beneath a superior crest as in Epilohocera and Potamocarcinus ; and the lateral margins are not armed with strong teeth or spines. The epistome is deeply channeled transversely and the labial bor- der is divided into three very prominent lobes projecting far forward, and of which the lateral ones are bilobed at tip and are separated from the antero-lateral angles of the buccal opening by broad and very deep efierent orifices. The external maxillipeds are as in Epilohocera, the merus trans- verse, the anterior margin rounded, and the palpus goniarthroid. In the single species upon which the genus is based, there is a long and slender spine projecting from the upper side of the expiratory canal near the external orifice. In the character of the front, this genus agrees with the species of Pseudothelphusa which have no superior frontal crest and diflers from Epilohocera, while, in the position of the antennse, it agrees with Ejn- lohocera and differs from Pseudothelphusa. * 'OtvcgOe, po7ie ; Kcpag, cornu. iS. I. Smit/) on American Crustacea. 149 Opisthocera Gilmanii, sp. nov. Plate V, figure 1, Male. The dorsal surface of the carapax is evenly convex in two directions and nearly smooth, but very minutely granulated and con- spicuously punctate with widely scattered punctures. There is no indication of areolation except two minute lunate impressions in the middle. The front has a smooth, revolute margin, which is continu- ous with the upper margin of the orbits, and a distinct, submarginal groove, which extends slightly along the inner portion of the supe- rior orbital border. The orbits are large, open and shallow, only par- tially filled by the ocular peduncles, and the inferior margin is sharp and minutely denticulate. The antero-lateral margin is evenly con- vex in outline, is broken by a small, oblique groove near the angle of the orbit, and its edge is sharp and very slightly and obtusely dentic- ulated anteriorly, but smooth posteriorly. The postero lateral mar- gin is concave in outline and rounded. The inferior lateral regions are naked and smooth. The labial border of the epistome is deeply divided ; the lobes are very prominent, and neai'ly horizontally, the median lobe being longest and its extremity triangular. The external maxillipeds are nearly smooth externally, but are marked with a few scattei'ed punctations. The chelipeds are very unequal ; in both, the merus is triquetral, the inferior angle rounded, but armed with a few small tubercles toward the carpus, and the superior angles are obtuse and armed with numer- ous tubercles, which are somewhat spiniform on the anterior angle ; the carpus is smooth and rounded externally and has a prominent spine on the inner margin. The basal portion of the propodus in the larger hand, is very stout, the superior margin is quite high, but rounded, and the inferior margin is armed with a few small tubercles near the base, the fingers are long, rather slender, and irregularly toothed within, and the dactylus is strongly curved so that the fingers gape very widely. The smaller hand is quite slender, the fingers are nearly cylindrical, very long, nearly straight, and but slightly gaping. The ambulatory legs are slender, naked and nearly smooth, the meral segments are nai-row, and the dactyli are armed with three rows of spines above and two below. The abdomen is widest at the third segment, and the first and sec- ond segments are only slightly narrower; from the third segment, the margins converge quite rapidly to the sixth, which is nearly twice as broad as long and its lateral margins only slightly converging ; the terminal segment is much broader than long and its extremity som 150 S. I. Smith on American Crustacea. what acutely arcuate. The appendages of the first segment are very stout and nearly straight organs reaching to the middle of the sixth segment, and articulated at their bases to a hard plate, which arches round the intestinal canal much as described under the genus Cardl- osoma. A deep groove extends from the basal articulation along the inside of each of these organs, curving round to the outside and ter- minating at the tip, which is truncate, turned sharply outward and armed with sharp, hooked spinules, and, on the inferior edge, with a small, curved process. The appendages of the second segment are as long as those of the first, are widely separated at theii- bases, and the terminal portions, which are lodged in grooves in the appendages of the first segment, are long, very slender and taper to acute points. The color, in alcohol, is uniform dirty yellowish brown, lighter beneath. Length of carapax, 38-7"'"' ; breadth of carapax, 57-2"'"' ; ratio, 1:1-48. Length of larger hand, 61-0"'™; breadth, 24-5; length of dactylus, 37*0. Length of smaller hand, 41'0"'™; breadth, 12 '8; length of dactylus, 24-5. The single specimen, which furnishes the above description, is in the collection of the Boston Society of Natural History, and was col- lected in a small stream near the center of the Isle of Pines by S. H. Scudder and Winthrop S. Gilman, Jr. At the suggestion of Mr. Scudder, the species is named for his friend. Epilobocera stimpson. Epilobocera Cubensis stimpson. Annals Lyceum Nat. Hist., New York, vol. vii, p. 234, 1860. This species, discovered in fresh water streams on the Island of Cuba, near Santiago, has close generic relations with the last species, but the character of the front and of the epistome is very difierent. I have seen only a single, imperfect, female specimen loaned by Dr. Stimpson. In this specimen, the dorsal surface of the carapax is arm- ed, along the lateral border, with small, tuberculiform granules, and the inferior lateral regions are armed, toward the lateral margin, with similar granules which are conspicuous on the anterior part of the in- ferior branchial region. The superior frontal crest projects consider- ably beyond the inferior one and is divided into two, slightly convex lobes by a well marked, median sulcus which extends back upon the carapax to the raesogastric lobe. The inferior margin of the front is straight, as seen in a front view, and its edge is slightly crenulated. S. I. Smith on American Cirustacea. 151 The inferior margin of the orbit is finely crenulated, and the crenula tions cease near the external angle, but there is no hiatus. The labial border of the epistome has a prominent, triangular tooth in the middle and smaller ones each side ; they all project downward and very slightly forward, and the median one has one or two small denticles toward its base. There is a quite broad, but very short, pro- cess projecting from the upper side of the expiratory canal, nearly in the position of the slender spine in Opisthocera. The abdomen is very similar to that of the male Opisthocera just described, except that the first and second segments are scarcely nar- rower than the third. It is remarkably narrow for a female, and the specimen is probably a sterile individual of that sex. Epilobocera armata sp. nov. Plate V, figure 2. The carapax is flattened above and the dorsal surface is neai-ly smooth, but very minutely granulous and punctate with widely scat- tered punctures. The epigastric lobes are just indicated by slight ele- vations and are separated by a very distinct, broad and shallow me- dian sulcus which extends forward and breaks through the superior frontal crest in a smooth sinus. There are no other marks of areola- tion except two minute lunate impressions in the middle of the cara- pax. The superior margin of the front projects slightly beyond the infei'ior one, is nearly straight, as seen from above, but curved down- ward in the middle, as seen in a front view, and is closely armed with conspicuous, rounded tubercles. The inferior margin of the front is straight and its edge is raised into a prominent crest and is distinctly crenulated. The superior margin of the orbit is continuous with the inferior margin of the front and is crenulated like it, and, at the outer angle is armed with one or two spiniform tubercles. The inferior mar- gin of the orbit is finely dentate and is broken beneath the outer angle by a broad, smooth sinus. The antero-lateral margin is separated from the angle of the orbit by a slight hiatus and is armed with sharp, spiniform teeth, which are prominent and slender on the anterior por- tion, but decrease in size posteriorly and are quite small at the broad- est portion of the carapax. The postero-lateral margin is concave in outline, as seen from above, smooth and rounded. The labial border of the epistome is divided into three lobes as in the last species. The median lobe is very prominent, projects out- ward nearly as far as the superior crest of the front, is acutely trian- gular and armed with two or three spiniform tubercles on each side, 152 /S. I. Smith on American Crustacea. of which the ones toward the base are very prominent. The lateral lolies are obtusely rounded, their outer margins are unarmed and the inner margins are armed somewhat as the median lobe, but the tuber- cles at the bases are slightly separated from the lobes, and stand par- tially between the lateral and median. There is a process pro- jecting from the upper side of the expiratory canal, as in the last spe- cies. The external maxillipeds, the chelipeds, and the ambulatory legs are very much as in E. Cubensis. The abdomen is very broad, nearly covering the whole sternum, the greatest breadth being at the fifth segment, and the fourth and sixth but little narrower. Sex. Leagth of carapax. Breadth of carapax. Ratio. Female. 43-8mm 70-4 1: 1-61 It 47-2 "77-5 1 : 1-64 The two specimens from which this description was taken are in the collection of the Boston Society of Natural History, and without labels to indicate from whence they came, but they are probably from the Bahamas. Although closely allied to JEJ. Cubensis, it is readily distinguished from the only specimen of that species which I have seen, in wanting wholly any granulations or tubercles along the lateral margins of the carapax, either above or below, by the more tuberculose superior fron- tal crest, in having tubercles at the outer angles of the orbits and a marked hiatus beneath it in the inferior margin, by the much longer teeth of antero-lateral margin, and by the quite difierent labial bor- der of the epistome. Family, Tkichodactylid^e. Dilocarcinus Edwards. Dilocarcinus pictns Edwards. Annales des Sciences naturelles, S^e serie, Zoologie, tome xx, 1853, p. 216; Archives du Museum d'Histoii'e naturelle, Paris, tome vii, p. 181, pi. 14, fig. 2, 1854. There are specimens in the collection of the Peabody Academy of Science and of the Museum of Yale College, from the River Amazon, at Nauta, Peru, which I refer to this species, although they do not agree perfectly with Edwards' figures and description. The speci- mens from Nauta are alcoholic and both females, and are considera- bly larger than the figure given by Edwards, one of them giving the following measurements: — Length of carapax, 29'0'"'"; breadth of S. I. Smith on American Crustacea. 153 carapax, including teeth, 34'6 ; ratio, 1:1'19. The carapax in our specimens is somewhat broader and the lobes of tlie front, as seen from above, are more prominent and their summits nearer together, leaving the orbit larger, than in the figure. The propodi and dactyli of the ambulatory legs are thickly ciliated along both edges, while Edwards' figures 2'^ and 2'^ represent only a few cilia on the posterior edges ; in the text, however, the dactyli are said to be " a bords cilies." The abdomen is quite remarkable for a female, the third and the three following segments being united into a single piece, as in the figure of the male abdomen, given by Edwards,* but, unlike the figure, it is broadest at the middle and the margins are convex in outline. Family, Grapsid^e. G-lyptograpsus, gen. nov. The carapax is much broader than long and the dorsal surface is distinctly areolated. The front is arched and nearly horizontal above the antennoe and antennula^, but excavated and defiexed in the mid- dle. The lateral margins are strongly arcuate and are dentate ante- riorly. The epistome is high and nearly perpendicular and is crossed trans- versely by a sharp groove, and the labial border is straight, as seen in a front view, but broken by a distinct notch in the middle, as seen from below. At the sides of the epistome, in the antero-lateral angle of the buccal area, there is a deep and narrow notch, which serves as an efferent orifice. There are no longitudinal ridges on the palate. The basis of the antenna is movable and fills the whole space be- tween the small, triangular, inner suborbital lobe and the front, and its summit is excavated on the inner side for the reception of the suc- ceeding segments, which are within the orbit. The external maxillipeds are not crested and their inner margins are closely approximated ; the ischium and merus are of nearly equal length and are both very broad, the merus being broader than long, and its antero lateral angle not expanded. The ambulatory legs are long and the dactyli are quadrangular and the angles armed with spines. None of the segments of the male abdomen are anchylosed. * This figure is marked 3 on plate 14, as if it belonged with fig. :!, Z>. spinifer, and on p. 180 it is referred to under that species, but in the explanation of the plates on p. 192, no fig. 3e is mentioned, while under D. pictus is placed, "Fig. 2^ . Abdomen du male," yet there is no fig. 2e on the plate, and Be is the only abdomen there figured. The abdomeu is not referred to in the description of D. pictus. 154 S. T. Smith on American Crustacea. The aspect of the single species upon which this genus is founded is quite peculiar. The body is thick, the dorsal surface is uneven and the lateral margin is armed with five teeth (including the angle of the orbit), the last and smallest of which is on the postero-lateral margin. The form of the carapax, the arching of the front above the antennu- Ise, and the number of teeth on the lateral margin, recall the genus Cryptogropsus^ from which, however, it is widely separated by the form of the external maxillipeds and of the epistome. In the form of the maxillipeds it is allied to Ileterograpsus. The form of the epistome and the peculiar, deep efferent orifice are very marked and distinctive characters. Glyptograpsus impressus, sp. nov. Male. The dorsal surface of the carapax is uneven, with numerous, irregular, shallow punctures, and along the lateral borders, with small, tuberculose elevations. The cervical suture is indicated by a very dis- tinct sulcus. The median portion of the gastric region is separated from the protogastric lobes by deep sulci, which unite between these lobes and extend down the front as a broad and deep depression. The epigastric lobes are very prominent and their anterior margins are transverse and precipitous. The protogastric lobes are well indi- cated, and an outer lobule is separated as a small, but very distinct, tuberculiform elevation opposite the inner angle of the orbit. The epibranchial lobes are uneven and partly separated from the meso- branchial by well marked, but short, depressions. The posterior por- tion of the branchial region is divided by a longitudinal ridge into a flat inner area and a broad precipitous portion between the ridge and the lateral margin. The front, as seen from before, is very sinuous, and broken in the middle by a broad, deep, rounded sinus ; its outer angles, as seen from above, are obtusely rounded, and the margin is continuous to the inner angle of the orbit, where it passes abruptly downward beneath the ocular peduncle as a sharp ridge, leaving a dis- tinct notch, above which the margin begins again and is continuous to the acutely triangular antero-lateral tooth, which is prominent and directed straight forward. The second tooth of the lateral margin is broad and obtusely rounded and situated above the plain of the ante- rior tooth ; the third and the fourth are slender and acute ; the last is on the postero-lateral margin and is small, acutely pointed and some- what below the level of those just in front of it. The inferior margin of the orbit is straight and finely dentate. The inferior lateral re- gions are granulous and slightly hairy. S. J. /Smith on American Crustacea. 155 The chelipeds are short and very unequal ; in both, the merus is short, not extending beyond the margin of the carapax, and trique- tral, with the angles denticulate, and the carpus is small and its outer surface granulous and sliglitly margined on the inner edge. In the larger hand, the propodus is short and very stout, the outer surface is convex and finely granulous, and the digital portion is very short, and its prehensile edge directed obliquely downward; the dactylus is straight, rather slender, and granulous like the propodus ; both fin- gers are obtusely tubercular on the prehensile edges and have horny, slightly excavated tips. The smaller hand is slender, somewhat cylin- drical, the basal portion is granulous externally, and the fingers are very slender, with the prehensile edges minutely toothed and the tips as in the larger hand. The ambulatory legs are nearly naked ; the meral segments are flat and each is armed with a small spine on the anterior edge near the distal extremity ; the carpi are slightly bicarinated along the anterior edges ; the propodi are broad, somewhat expanded in the middle, the anterior edges carinated like the carpi, and the posterior edges spinu- lous. The dactyli are slender, slightly curved, somewhat flattened, and the angles armed with sharp spinules. The abdomen is broadest at the base, from which it tapers to the last segment, which is longer than broad and rectangular, except that the extremity is slightly rounded. Length of carapax, including lobes of frontal margin, 1 2 -4"^™ ; breadth of carapax, including lateral teeth, IS'O"'™; ratio, 1:1*21. Breadth between antero-lateral angles, 1 1 -5""". Length of ambula- tory legs, first, 19™™ ; second, 25 ; third, 25 ; fourth, 21. I have seen only a single specimen, which was collected at Acajutla, west coast of Central America, by F. H. Bradley. The appendages of the first abdominal segment in the male are widely separated at their bases, which are articulated to a slender plate arching round the intestinal canal, and converge toward their tips, but do not meet, although they extend to the middle of the sixth segment. Each of the organs is nearly straight and rather stout for two-thirds its length, and the terminal portion is suddenly constricted on the under side and curved outward and strongly downward to the tip. The appendages of the second segment are small and are lodged in grooves at the bases of the first pair. 156 S. J. /Smith on American Crustacea. Sesarma Say. Sesarma reticulata Say. Ocypode (Sesarma) reticulatus Say, Journal Academy Nat. Sci., Philadelphia, vol. i, p. 7.3, 76, pi. 4, fig. 6, 1817, and p. 442, 1818. (Sesarma re^zcwZaiffi Gibbes, Proceedings American Association, 3d meeting, p. 180. 1850 ; Edwards, Annales des Sciences naturelles, S^ie serie, Zoologie, tome xx, 1853, p. 182; Stimpson, Annals Lyceum Nat. Hist., New York, vol-, vii, p. 66, 1859. This species is found at New Haven, Conn., inhabiting salt-marshes and associated with Gelasimus pugnax. Sex. Lengtl 1 of carapax. Breadth of carapax. Ratio. Breadth of front. Male. 14-Omm 17-imm 1:1-22 9-4mm 15-2 18-3 1: 1-20 9-9 17-2 21-0 1 : 122 11-4 19-7 24-2 1: 1-23 13-2 22-4 27-5 1:1-23 150 23-0 28-3 1: 1-23 15-4 Female. 19-7 24-6 1: 1-25 13-5 In this species, the first segment of the male abdomen projects lat- erally considerably beyond the second segment, and beyond tlie pos- terior margin of the carapax, and tlie third segment is as wide as the first and its lateral margins are strongly arcuate ; at the fourth seg- ment, the abdomen is suddenly contracted and the remaining portion is quite narrow and the margins are slightly concave to the sixth seg- ment ; the terminal segment is scarcely more than one half as wide as, but considerably longer than, the sixth, much longer than broad, and its extremity rounded. The appendages of the first segment extend nearly to the extremity of the sixth segment, are articulated at their bases to a slender, arched plate, much as m Glyptograpsus impressus, are triquetral, quite stout, nearly straight and widely separated even to their tips, which are slightly flattened and hairy. The appenda- ges of the second segment are short and slender and are lodged in grooves at the bases of the appendages of the first segment. Sesarma sulcata, sp. nov. Female. The carapax is quadrihiteral in outline and much broader than long. The dorsal surface is convex in both directions, but some- what more so longitudinally than laterally, and is clothed anteriorly and along the sides with scattered fascicles of short hairs. The pro- togastric lobes ai'e divided, for half their length anteriorly, into nearly equal lobules by well mai-ked sulci, and are limited next the orbits by deep depressions which extend to the antero-lateral angle of the cara- pax. The median portion of the gastric region is surrounded by a S. I. Smith on American Crustacea. 157 broad depression and is somewhat separated from the rather broad mesogastric lobe, which extends forward, in the median sulcus between the protogastric lobes, nearly to the front. This median sulcns is broad and very deep, with precipitous sides and cuts through the whole height of the frontal crest. The branchial regions are trav- ersed by sharp transverse plications. The fi'ont is perpendicular and low, and the inferior margin is broken by a broad excavation in the middle, where it scarcely projects beyond the epistome ; above the antennulffi the edge projects, but toward the orbit slopes otf again. The antero-lateral margin is armed with two stout teeth (including the angle of the orbit) and with the trace of a third. The first tooth is acute, dii'ected forward and situated below the level of the rest of the margin, the second is prominent, acute, and projects forward par- tially over the deep, rounded incision which separates it from the first tooth, and the third is only indicated by a slight emargination. The chelipeds are equal and rather small ; the merus is rough ex- ternally, the angles are sharp and the anterior ones serrate ; the cai*- pus is very granulous externally ; and the hand is slightly compressed, smooth externally, and the superior margin armed with a sharp crest. The ambulatoiy legs are stout and much compressed, the meral seg- ments are very broad, the breadth being equal to half the length, and rough with short transA^erse plications, the propodi and dactyli ai'e hairy along the edges, and the dactyli are stout, curved and acumi- nate. Length of carapax, 25-0™'"; greatest breadth of carapax, 3rO'"'^; ratio of length to breadth, 1:1*24. Breadth of carapax between antero-lateral angles, 29-5""". Breadth of front, 16-4'"'" ; height of front, 3 •4""". The single specimen described was obtained at Corinto, west coast of Nicaragua, by J. A. McNeil, and is in the collection of the Pea- body Academy of Science. Sesarma cinerea Say. Grapsus cinereus Bosc, Histoire naturelle des Crust., tome i, p. 204, pi. 5, fig. 1, 1802 ; Latreille, Histoire naturelle des Crust, et Insects, tome, vi, p. 72, 1803. Grapsus {Sesarma) cinereus Say, Journal Academy Nat. Sci., Philadelphia, vol. i, p. 442, 1818 {non Grapsus cinereus Say, loc. cit, p. 99, 1817). Sesarma cinerea Edwards, Histoire naturelle des Crust, tomeii, p. 75, 1837; Annales des Sciences naturelle, 3™^ serie, Zoologie, tome xx, 1853, p. 182 ; Gibbes, Proceed- ings American Association, 3d meeting, p. 180, 1850; Stimpson, Annals Lyceum Nat. Hist, New York, vol. vii, p. 65, 1859. 158 S. I. Stnith on American Crustacea. There are specimens before me collected at Egmont Key, west coast of Florida, by Col. E. Jewett ; at Bluffton, South Carolina, by Dr. J. H. Mellichamp (collection Peabody Academy of Science), and at Fort Monroe, Virginia, by Dr. Kneeland. Several specimens give the following measurements : — Locality. Sex. Length of carapax. Breadth of carapax. Ratio. Breadth at orbital angles. Breadth of front. Bluffton. Male. 1 2 ■ 1 mni 13-8mm 1:1-14 13-9mm 8-2mm Ft. Monroe. u 12-8 14-4 1: 1-13 14-0 8-3 Bluffton. " 15-2 n-4 1:113 17-0 100 Ft. Monroe. " 16-4 18-6 1:]13 17-8 10-8 Egmont Key. Female. 11-0 12-8 l:l-lfi 12-6 7-2 " " 12-8 15-0 1:1-1'7 14-5 8-7 The abdomen of the male is broadest at the third segment, the first and second are much narrower and of equal length ; from the fourth to the sixth, the abdomen is broad and the lateral margins converge regularly ; the terminal segment is scarcely a third as wide, but about as long, as the sixth, and very little longer than broad. The appen- dages are similar to the appendages of S>. reticulata^ but those of the first segment are a little shorter and much stouter. Sesarma occidentalis, sp. nov. A species closely allied to S. cinerea Say. Male. The carapax is quadrilateral in outline and considerably broader than long. The dorsal surface is flat in the middle and pos- teriorly, but somewhat convex in front and along the sides. The pro- togastric lobes are convex and divided by slight depressions anterior- ly, and the surface is rough with coarse, sharp granules arranged in very short, irregular, broken lines. The median portion of the gas- tric region is sparsely granulous, surrounded by a shallow sulcus, and the mesogastric lobe is very narrow and extends far forward in the well marked, median sulcus between the protogastric lobes. The branchial regions are traversed by indistinct transverse plications, and the posterior regions are punctate with indistinct, shallow puncta. The front is nearly perpendicular, quite high and slightly concave, the concave surface is irregularly and coarsely granulous, and the inferior margin is curved forward somewhat beyond the crest and its edge is nearly straight. The antero-lateral tooth is acute and projects well foi'ward. The lateral margin is sharp, continuous, and nearly straight as seen from above. The chelipeds are equal, short and stout ; the anterior angle of the merus is sharp, dentate and raised into a thin crest at the end next S. I. Smith on American Crustacea. 159 the carpus ; the carpus is thickly beset externally with sharp gi-an- ules ; the basal portion of the propodus is short, and the outer surface is evenly rounded and very granulous and the superior margin is armed with a sharp crest ; and finally, the dactyl us is granulous on the upper side at base. The ambulatory legs are rather slender, the meral segments are sharply granulous above, and the propodi and dactyli are clothed with a few short, stiff hairs along the margins. Two males give the following measurements : — Length of Breadth of Breadth at Breadth Height carapax. carapax. Eatio. orbital angles. of front. of front. 11-6™™ 13-ln"n 1:1-13 12'9ni™ 7-omm 21™™ 15-8 17-6 1:1-12 16-9 9-4 3-0 I have seen only two specimens, both males, which were collected at Acajutla, west coast of Central America, by F, H. Bradley. Although closely allied to S. cinerea, it is very readily distinguish- ed from all specimens of that species which I have seen, by the gran- ulous anterior regions of the carapax, the coarsely granulous front, and by the crested and granulous hands. The carapax also is more convex anterioi'ly and along the branchial regions. The male abdomen and its appendages are almost exactly as in S. cinerea, except that the last segment of the abdomen is somewhat larger in proportion. Sesarma angustipes Dana. United Stales Exploring Expedition, Crust, p. 353, pi. 22, fig. 7, 1852; Stimpson, Proceedings Academy Nat. Sci., Philadelphia, 1858, p. 106; Annals Lyceum Nat. Hist., New York, vol. vii, p. 66, 1859. Six specimens give the following measurements : — Locality. Sex. Length of carapax. Breadth of carapax. Eatio. Breadth at orbital angles. Breadth of front. AspinwalL Male. 8-7mm 9.3mm 1: 1-07 9-5mm 4-7mm ii " 15-2 16-2 1: 1-07 15-3 8-5 Florida. 11 17-0 18-2 1:1-07 17-0 10-2 " " 18-9 20-2 1:1-07 18-4 10-6 K Female. 11-2 12-0 1: 1-07 11-5 6-8 II ■ u 16-6 18-2 1:1-10 16-8 9-5 Sesarma angusta, sp. nov. Female. The carapax is quadrate, longer than broad and depress- ed. The protogastiic lobes are very little convex, slightly divided anteriorly and their surfaces beset with sharp granules. The median portion of the gastric region is surrounded by a well marked sulcus, 160 S. I. Smith on American Crustacea. and the anterior portion of tlie meso-gastric lobe extends forward, almost to the line of the front, as a very narrow ridge in the deep sul- cns between the protogastric lobes. The median and posterior regions are punctate with irregular, coarse punctations, and the branchial re- gions are sliglitly plicate transversely. The front is nearly perpen- dicular, but low and very concave, tlie superior crest projects almost as far forward as the inferior margin, and is divided into four equal lobules by a deep median groove and slight lateral ones, and the infe- rior margin is strongly reflexed, its edge sinuous, as seen from above, wnth a broad and shallow sinus in the middle, and a very slight one each side. The antero-lateral tooth is nearly right-angular, and pro- jects but slightly forward. The lateral margin is straight and entire. The chelipeds are equal and very small, the merus and carpus are sharply granulous externally, the hand is about half as long as the breadth of the front, slender, the inferior edge evenly rounded, and the superior edge more angular and sparsely granulous, but not crest- ed, and the fingers are slender, nearly cylindrical, and very slightly toothed within. The ambulatory legs are very long and slender, even longer than in /S. angustipes, and the meri and propodi are rough above. Length of carapax, from its posterior margin to superior lobes of the front, 14-1""" ; breadth of carapax, 13-8'"™ ; ratio, 1 : 0-98. Breadth of carapax between antero-lateral angles, 13*6™"\ Breadth of front, '7-2; height of front, 1"8. Length of ambulatory legs, first, 22*0; second, 28*4; third, 32'0; fourth, 25-0. Length of jjropodus in first pair of ambulatory legs, 5 '6 ; second pair, 8*0 ; third pair, 9-0 ; fourth pair, 6*6. I have seen only one specimen, a female, collected at the Pearl Isl- ands, Bay of Panama, by F. H. Bradley. It is readily distinguished from all the other described American species of the genus by the narrowness of the carapax, the low, per- pendicular and excavated front, and the great length of the ambula- tory legs. GONOPLACID^. Prionoplax Edwards. Prionoplax ciliatus, sp. nov. A species similar to P. spinicarpus Edwards, Archives du Museum d'Histoire naturelle, Paris, tome vii, p. 167, pi. 11, fig. 3. JS. I, Smith on American Crustacea. 101 Male. The carapax is very convex longitudinally, but scarcely at all transversely. The dorsal surface is thickly beset with small, tuber- culiform granules, but the space between the granules is smooth and shining. The areolation is similar to that of -R spinicarpus j the cer- vical suture is indicated by a very distinct, smooth sulcus, which is sharp and deep in the longitudinal portions in the middle of the cara- pax ; the mesogastric and the metagastric lobes are united ; there are no distinct sulci between the protogastric lobes and the hepatic regions ; tlie branchial regions are undivided and only indistinctly separated from the cardiac. The front is lamellar, very strongly de- flexed and its edge divided into two prominent, rounded lobes, which, when seen in a front view, project below the inferior margins of the orbits. The antero-lateral margin is thin and is divided by deep rounded sinuses into four slightly upturned lobes or teeth, of which the anterior, the hepatic, and the epibranchial are broad and truncate and their truncated edges finely denticulated, while the posterior, or mesobranchial, is acutely pointed. The inferior lateral regions are grauulous like the dorsal sui-face, and, along the lateral borders, are clothed with long cilia which project beyond the margins. There are also, some hairs along the lateral margins of the dorsal surface, but they are very easily removed. The outer surface of the external maxillipeds is minutely granulous. The chelipeds are stout and slightly unequal. The merus is trique- tral and armed with a spine on the posterior angle near the distal ex- tremity. The upper side of the carpus is flat, somewhat roughened, and armed on the middle of the inner side with a long spine. The hands are stout, slightly compressed laterally, and perfectly smooth ; the upper edge is angular, but not crested, and the fingers are com- pressed, deflexed, somewhat incurved, coarsely and irregularly toothed within, and do not gape. The ambulatory legs are slender and thickly hairy along the edges, especially on the dactyli, which are long, very slender, and cylindrical. The sternum is granulous like the carapax, only more minutely. The abdomen is smooth ; the first and third segments are very much wider than the second, and the penultimate is much broader than lono- and its lateral margins are deeply concave in outline. The appenda- ges of the first segment are long, slender, triquetral, and nearly straight organs reaching almost to the extremity of the abdomen. The appendages of the second segment are short and inconspicuous. I have seen only males. Trans. Connecticut Acau., Vol. II. ii April 1870 162 *S'. I. Smith on A^nerlcan Crustacea. Lengfth of carapax, IS-imm Breadth of carapax, 22-9mni Ratio, 1 : 1-44 " " " 15-5 " " " 23-9 " 1:1-47 Collected at Panama by F. H. Bradley. This species is closely allied to P. sphiicarpus, and it may possibly prove to be identical with the species from Panama mentioned under that name by Stirapson, Annals Lyceum Nat. Hist., New York, vol. vii, p. 59. Edwards states, howevei', that, in his species, the teeth of the antero-lateral margin are " aplaties et aigues," and they are so figured on his plate, while in our species, all, except the posterior one, are broad, truncate and denticulated. The carapax in his figure is considerably broader, and the chelipeds seem to be much less robust, than in P. ciliatus. Moreover, there are no hairs or cilia indicated in the figure, on the carapax or the ambulatoiy legs, and they are not mentioned in the description. The specimens, when received, were completely covered with fer- ruginous mud. Their cylindrical form is well adapted for living in holes, and this is quite probably the habit of the species, as it is of Speocracinus, according to Stimpson. Euryplax Stimpson. Euryplax nitidus stimpson. Annals Lyceum Nat. Hist, New York, vol. vii, p. 60, 1859. Of this species, there is a specimen, in the Museum of Yale Col- lege, collected at Egmont Key, west coast of Florida, and there is another in the collection of the Peabody Academy labeled New Or- leans, but probably from some part of the Gulf of Mexico. Both these specimens are adult males and agree perfectly with Stimpson's description. The pit on the anterior surface of the merus is exactly alike in both chelipeds and in each specimen. The antero- lateral margins converge anteriorly so that the breadth of the cara- pax between the anterior angles, is very much less than between the posterior teeth. The anterior angle is obtuse, the second tooth is tri- angular, but blunt, and the last is slender and acutely pointed. The male abdomen is broadest at the second segment, the sides of which extend in narrow projections quite to the coxae of the posterior legs. The first segment is narrow and is only exposed in the broad excavation of the posterior margin of the carapax. The third seg- ment is very broad and its sides project in acute angles, over the chan- nel between the sixth and seventh segments of the sternum, nearly to the coxse of the posterior legs. From the third segment, the abdo- S. I. Smith on American Crustacea. 163 men is nan-ow and tapers to a very narrow terminal segment, which is two-thirds longer than broad, and obtuse at tip. The appendages of the fii'st segment extend a little beyond the sixth segment. They are widely separated at base, strongly incurved till they meet a little way from the tips, which are again curved strongly outward. They are slender and taper to slender and acute tips, and the terminal third is shining black in color. The apj^endages of the second segment are situated within those of the first, are short, slender, straight, and white. Alcoholic specimens are pale yellowish white, and the fingers white at tips. Lenprth of Breadth of Breadth Locality. Sex. carapax. carapax. Katio. of front. Florida. Male. 13-4mm 22-Omm i ; i-64 10-2mm New Orleans? " 14-6 240 1 : 1-65 10-4 Euryplax politus, sp. nov. This species is allied to the last, but wants wholly the pits on the meral segments of the chelipeds, and the antero-lateral margins are parallel instead of converging anteriorly. Male. The carapax is glabrous, convex longitudinally and very slightly transversely. The dorsal surface is not distinctly areolated, although the cervical suture can be traced by a slight depression. The fi-ont is nearly straight and has a distinct marginal groove upon the upper edge and is deeply notched each side at the insertion of the antennae, as in M nitidtis. Tlie antero-lateral margins are parallel, very short, and each is armed with three acute teeth. The postero- lateral margin is slightly incurved. The posterior margin is slightly concave in the middle. The chelipeds are nearly equal, stout, smooth and glabrous. The merus is armed with a small spiniform tooth, as in E. nitidus, and the carpus, with a small tooth within. The hands are slightly swol- len, the superior margins are quite high, but smooth and rounded, and the fingers are slender and slightly deflexed. The ambulatory legs are smooth, nearly naked, and very slender. The abdomen is quite similar in form to that of E. nitidus, and the appendages are very much as in that species, but those of the first segment are not as strongly curved at the tips, and the terminal por- tion is brown instead of black. An alcoholic specimen is pale yellowish white, ■with the fingers brown at tip. Length of Breadth of Breadth Sex carapax. carapax. Ratio. of front. Male. Q-9rnm ll-2mm 1:1-63 4-4mm A single specimen was collected at Panama by F. H. Bradley. 164 S. I. Smith on American Crustacea. This species agrees perfectly with all the characters assigned to the genus Euryplax by Stimpson, except in wanting wholly the pit on the front side of the merus of the chelipeds. This character might, perhaps, be considered generic, but, in the absence of any knowledge in regard to its functional importance, it seems best to refer this spe- cies to Euryplax, and especially, since it agrees so closely in most of its specific characters with the type of that genus, G-lyptoplax, gen. nov. The carapax is cancroid in form and similar to Eucratopsis* The dorsal surface is deeply areolated, the front is prominent and nearly horizontal, and the antero-lateral margin is dentate and about as long as the postero-lateral. The basis of the antenna is long and joins a slight process from the side of the front. The epistome is much as in Panopeus. There is a sharp carina on each side of the palate, along the efterent canal, but it is interrupted a little way from the border of the epistome. The external maxillipeds are approximated along their inner mar- gins. The ischium is longer than broad, and its anterior extremity projects farther forward on the inside than the outside. The merus is somewhat triangular, the antero-lateral angle is very prominent, the anterior margin is very short and nearly parallel with the inner mar- gin, which slopes oft* rapidly toward the antero-lateral angle. The palpus is endarthroid. The chelipeds are short, but the hands are very stout. The ambu- latory legs are slender and smooth. The seventh segment of the male sternum is exposed on each of the abdomen. The verges pass from the coxje of the posterior legs to the abdomen, through canals beneath the sternum. The sides of the first segment of the abdomen extend in triangular projections to the coxaB of the posterior legs; the second segment is much narrower than either the first or the third ; the sides of the third segment do not reach the margins of the sternum; and the third, fom-th, and fifth seg- ments are anchylosed. This genus is allied to Eucratopsis, but differs very much from it in the form of the external maxillipeds, in the more prominent and hori- zontal front, and in the longer antero-lateral margins of the carapax. From Speocarcinus Stimpson (Annals Lyceum Nat. Hist,, New York, * Eucrate Dana. See these Transactions, vol. ii, p. 35. S. I. Smith on American Crustacea. 165 vol. vii, p. 58), it differs in the appioximation of the external maxilli- peds and in the form of the carapax. Glyptoplax pugnax, sp. nov. Male. The dorsal surface of the carapax is slightly convex longi- tudinally, but not at all transversely, and is thickly granulous. The mesogastric lobe is not distinct from the metagastric, but is well sep- arated li-om the protogastric, and its anterior portion is narrow and extends well forward. The protogastric lol)es are prominent and un- divided, and are not distinctly separated from the epigastric, which are very slight elevations separated by a marked median sulcus. The hepatic region is prominent, undivided, and separated from the gastric and branchial regions by deep sulci. The mesobranchial and meta- branchial lobes are separated by a very slight sulcus, and the anterior portion of the branchial region is divided into three lobules, — one at the base of the epibranchial tooth, a larger one just within this, and a small, indistinct one next the gastro-cardiac sulcus. The front is thin and horizontal, its edge is slightly convex, as seen from above, and divided by a very slight notch in the middle. At each side of the front, there is a deep antennal notch, above which, the inner angle of the sujjerior orbital border projects as a prominent tooth. The superior margin of the orbit is divided by two deep notches. The anterolateral margins are arcuate. The outer angle of the orbit pro- jects only slightly beyond the second tooth and is separated from it by a slight sinus. The remaining portion of the margin is divided into three, prominent, triangular teeth, of which the middle one, or epibranchial, is most prominent. The ocular peduncles are armed with a granulous tubercle on the anterior side near the cornea. The chelipeds are slightly unequal and the hands are very large. The merus does not project beyond the lateral margin of the carapax. The carpus is short and the outer surface is granulous, has a slight groove along the margin next the propodus, a tooth upon the inner margin, and a small tubercle near the articulation of the propodus. The hand is compressed, very broad, and nearly smooth. The basal portion of the propodus is slightly convex on both sides, the lower edge is rounded, and the upper edge is slightly crested ; the digital portion is very broad at base and very much deflexed, so that the pre- hensile edge is parallel with the margin at the base of the dactylus, the inferior edge is slightly margined on the outside, and the tip is slender and upturned. The dactylus is long and slender, the upper 166 S, J. Smith on American Crustacea. edge is slightly crested and the tip is hooted by the tip of the pro- podus. The prehensile edges of both fingers are sharp, very slightly dentate, and do not gape, or only very slightly. The ambulatory legs are slender and minutely granulous ; the pro- podi are slightly hairy on the posterior edges ; and the dactyli are slender, slightly compressed, those of the posterior pair considerably shorter than the others, and all clothed with very short hair. The sternum is minutely granulous. The terminal segment of the abdomen is about as broad as long, and the extremity is obtusely rounded. The appendages of the first abdominal segment are long, slender, nearly straight, and reach to the terminal segment. The appendages of the second segment are short and very small. The females differ from the males in being more convex and in the front being less prominent and very sliglitly deflexed. The young males approach the females in these characters. The fingers ai-e black in both sexes. No. 1. Sex. Lenj Male. jth of carapax. Br 4-8mm eadth of (!ara 6-4uim 2. 11 5-7 7-8 3. 11 60 8-3 4. (1 6-8 9-4 5. " 7-7 11-0 6. (1 8-6 121 7. Female. 4-4 61 8. 11 4-8 6-7 9. 11 51 7-2 Ratio. Breadth of front. 1 : 1-33 2 ■6mm 1 : 1-37 2-8 1 : 1-36 3-0 1: ]-38 3-5 1:1-43 3-7 1 : 1-41 41 1: 1-39 2-3 1 : 1-40 2-6 1 : 1-41 2-7 The chelipeds of numbers 2, 4, 6, and 9, give the following mea- surements: — No. 2. Length of hand. Right. Left, gfmin 62™™ Breadth of hand. Right. Left. 4-7mm 3-8™m Length of dactylus. Right. Left. 5-imm 4-6nim 4. 7-2 8-4 4-2 5-0 5-3 60 6. 10-2 110 5-8 6-2 8-0 8-4 9. 50 51 2-6 30 31 3-2 Collected at Panama by F. H. Bradley. Family, Pinnotherid.e. Pinnotheres Latreiiie. Pinnotheres margarita Smith. I_ Verrill, American Naturalist, vol. iii, p. 245, July, 1869. This is a stout, thick species, with a firm integument, and every where covered, except the dactylus of the right ambulatory leg of the S. I. Smith on American Crustacea. 167 second pair in tlie female, and the tips of the others in both sexes with a very short and close, clay-colored pubescence, looking much like a uniform coating of mud. Female. The carapax is very strongly convex in all directions and the dorsal surface, beneath the pubescence, is smooth and shining. The cardiac region is protuberant and is separated from the gastric region by a conspicuous sulcus, and from the branchial regions, by very marked and deep depressions, which extend along the cervical suture to the hepatic region. The branchial regions are protuberant along their inner sides. The front is not protuberant, is strongly detiexod, and has a slight median depression. The external maxillipeds are more longitudinal and of a firmer con- sistency than is usual in the genus. The merus is short and broad, and the inner margin is angulated in the middle, the portion toward the base fitting the anterior margin of the sternum and the distal por- tion being slightly concave and fitting closely the terminal segments of the palpus. The second segment of the palpus is large, broadest in the middle at the attachment of the terminal segment, and the outer surface is flattened. The terminal segment is slightly spatulate in form and reaches almost to the tip of the second segment. The chelipeds are equal and very stout and the hands are long and nearly cylindrical. The fingers are somewhat cylindrical, nearly straight almost to the tips, which are hooked by one another, and the prehensile edge of the dactylus is armed, near the base, with a small tooth, which fits a slight excavation in the propodal finger. The ambulatory legs are stout and all the ischial segments, and the posterior margins of the propodi and dactyli in the last pair, are clothed with a long, woolly pubescence. The dactyli in the three anterior pairs are short, curved, and pubescent nearly to the tips, except in the right leg of the second pair, where the propodus is con- siderably longer than in the corresponding leg on the other side, and the dactylus very long, almost straight, and entirely naked. In the posterior legs, the dactyli are long, straight, slender, and pubescent. The anterior margin of the sternum is excavated into a bi'oad, rounded sinus for the reception of the tips of the palpi of the exter- nal maxillipeds. The abdomen is orbicular and completely covers the sternum. Male. The only male which I have seen is much smaller than the females, and is not so thickly pubescent. The cardiac and branchial regions are less protuberant and are separated from the gastric by a 168 S. I. Smith on American Crustacea. slight depression only. The front projects slightly and is not so much deflexed as in the female. The chelipeds and ambulatory legs are like those of the female, except that the ambulatory legs of the right side are like those of the left. The abdomen is broadest at the third segment, fi'om the third to the sixth, the margins are straight and converging, the sixth is abruptly contracted, and the terminal segment is nearly square. The appendages of the first segment are rather stout organs, somewhat hairy along the margins, and reach to the terminal segment. They curve inward for about two-thirds of their length and then outward again to the tips. Tlie appendages of the second segment are short and are lodged in grooves at the bases of the first pair of appendages. Locality. Sex. Length of carapax. Breadth of carapax. Ratio. Pearl Islands. Male. 5 Smni Q-imxa i ; in La Paz. Female. 8-1 8-9 1:M0 Pearl Islands. ' 8-8 9-7 1:1-10 •' " 100 11-0 1:110 " " 10-3 11-4 1:111 " " 10-9 12 1:110 " " 11-8 13-4 1 : M4 This species was found living in the Pearl Oyster (Margarito- phora fimhriata Dunker), at the Pearl Islands, Bay of Panama, by F. H. Bradley, It has also been sent from La Paz, Lower California, by Capt. J. Pedersen. A sterile female Pinnotheres, found in an alcoholic specimen of the Pearl Oyster collected at the Pearl Islands by Mr. Bradley, probably belongs to this species. It agrees closely with specimens of P. m.ar- garita, described above, in the form of the external maxilli]»eds and the firm integument. The carapax is more like the male than the ordinary female, but is narrower and more depressed. The front is more prominent and scarcely at all deflexed. The dorsal surface is very slightly areola- ted, quite flat, and is clothed, except the cardiac region and a small space in the middle of the gastric, with a very dark, almost black, velvety pubescence. A single cheliped is stouter in proportion than in the ordinary male and female, and the pubescence upon the upper surface of the carpus and a small space at the base of the hand, is black as on the dorsal surface of the carapax. The ambulatory legs are less pubescent than in the male, while the propodus and dactylus of the right leg of the second pair are longer S. I. Smith on American Crustacea. 169 than in the corresponding leg of the left side, but are not as long as in the female. The abdomen is not broader than in the male, but the margins are slightly convex, it is not contracted at the sixth segment, and the extremity is rounded. Length of cirapax, 51™™; breadth of carapax, S-S™™; ratio, 1 : 1-04. Pinnotheres Lithodomi, sp. nov. Female. The carapax, in the single specimen examined, is much crushed out of shape, but the dorsal surface is smooth and naked. The raerus of the external maxilliped is broadest at the distal extremity, and both margiiis are nearly sti'aight. The chelipeds are equal, smooth, and naked. The hands are cylin- drical, and the fingers are short, nearly straight, the tips are slightly hooked by each other, and the prehensile edge of the dactylus is armed, near the base, with a small tooth, which fits a slight excava- tion in the propodal finger. The ambulatory legs are very slender and wholly naked, except the dactyli. In the first pair, the dactyli are very short and only slightly curved ; in the second, they are considerably longer than in the first, and nearly straight ; in the third, they are very long, being nearly as long as the propodi, slender, and slightly curved ; and in the posterior pair, they are about as long as in the second and are ciliated along the posterior edges. Breadth of carapax, about, 4'""". The only specimen seen, was found in a specimen of Lithodomus aristatus Forbes and Hanley which was in its excavation in the shell of a Spondylus collected at the Pearl Islands by F. H. Bradley. Although the specimen is very small, it has a large number of eggs beneath the abdomen. OstraCOthereS Edwards. Ostracotheres politiis, sp. nov. Female, The carapax is depressed, naked, smooth, and shining. The dorsal surface is flat and the borders are smoothly rounded. There is a short median sulcus on the front, and a very slight U-shaped one extending from the orbits to the middle of the carapax. The front does not j^roject beyond the anterior margins. The external maxillipeds are smooth and almost entirely naked, and, in fomi, are considerably like the figure of 0. affinis given by Edwards (Annales des Sciences naturelles, 3""^ serie, Zoologie, tome xx, 1853, 170 S. J. Smith on American Crustacea. pi. 11, fig. 11), but the merus is wider at the distal end and the outer margin is not so arcuate. The chelipeds are equal and all the segments are rounded, smooth, and glabrous. The hands are small and much compressed. The fin- gers are shorter than the basal portion of the propodus, do not gape, and the dactylus is slightly curved and is armed, near the base, with a small tooth, which fits a slight excavation in the propodal finger. The ambulatory legs are short, slender, cylindrical, and smooth. Those of the first pair are shorter than those of the second, and the dactyli, in both the first and second pairs, are very short and curved, and close against the expanded end of the propodus, which is clothed at that point with a little tuft of short, stiff hair. Those of the third pair are about the length of those of the second pair, and the dactyli are short and curved, but the distal ends of the pi-opodi are not expanded for their reception. The posterior legs ai*e sliorter than those of the second or third pair, are much more slender than any of the others, and the dactyli are only slightly curved and are very long and slen- der, their length being about equal to that of the propodi. The abdomen is very broad and covers the whole sternum. Length of carapax, 5*4™™; breadth of carapax, 7'3™"i; ratio, 1 : 1'35 a a a g.3 a a a g.3 a 1.1-32 " » " 6-4 " •' •• 8-5 " 1 : 1-33 Collected at Callao, Peru, by F. H. Bradley. The integument is quite thin and yielding, and the species undoubt- edly lives protected within some bivalve mollusk (probably Mytilus algosus Gould). It appears to difier remarkably from the other species of the genus in the depressed carapax and naked ambulatory legs, and I refer it to Edwards' genus with some doubt, although it agrees in the two-jointed palpus of the external maxillipeds. The other described species of Ostracotheres are : — 0. Savignyi Edwards {Pinnotheres veterum Savigny), from the Red Sea ; O. Tri- dacnm Edwards (Ruppell), also from the Red Sea; and 0. affinis Edwards, from the Isle of France. Pinnaxodes Heller. Pinnaxodes Chilensis Smith. Pinnotheres CMe?ms Edwards, Histoire naturelle des Crust., tome ii, p. 33, 1837; Edwards et Lucas. Voyage de d'Orbigny dans I'Amerique meridionale. Crust., p. 23, pi. 10, fig. 2, 1843. Fahia Chilensis Dana, United States Exploring Expedition, Crust., p. 383, 1852. Pinnaxodes hirtipes Heller, Reise der osterreichischen Fregatte Novara um die Erde, p. 68, pi. 6, fig. 2, 1865. Pinnaxodes Chilensis Smith, in Verrill, American Naturali'tt, vol. iii, p. 245, 1869. )S. 1. Smith on American Crustacea. 171 The parasitic liabits of this species have been fully described by- Prof. Verrill.* It inhabits JEuryechinus imhecillis Verrill, living in a sac formed by the distention of the intestine near the anal orifice. The females, after they have arrived at any considerable size, must remain permanently within the same echinus, since the anal orifice is much smaller than the body of the crab. I have examined quite a number of individuals obtained from spe- cimens of the Eiiryechinus collected by Mr. Bradley at Paita and Callao, Peru, and by Prof. James Orton at Paita, and have little doubt that the species figured by Edwards and Lucas and by Heller ai-e identical, although the figui-es given by these authors are quite different. The specimens l^efore me agree very well witli the figure in the work of Edwards and Lucas, except that the outer margin of the carpus of the external maxillipeds is not quite so much curved toward the distal extremity as in the figure. On account of the soft and yielding nature of the carapax, many of the specimens do not show distinctly the sulci in the dorsal surface. The figure given by Heller seems to have been drawn from such a specimen, for no sulci are represented. The carpus in the figure of the external maxilliped in Heller's work, is quite different from Edwards' and Lucas' figure ; but the figure of the latter authors represents the whole maxilliped removed from the rest of the animal, while Heller's figure represents only the exposed portion, and Avas evidently drawn from the maxilli- ped while in place, and, if the carpus were seen in a slightly oblique position, it would account for its narrower Ibrm in his figure. The dactyli of the ambulatory legs, as represented in Heller's figure, are somewhat longer than in oixr specimens. The peculiar habit is also a confirmation of the identity of the spe- cies. Heller's specimens were from Ecuador, and he says of them : — " Diese in zwei weiblichen Examplaren vorliegende Art soil nach Dr. Scherzer in einer Echinus-Art vorkommen." Neither Edwards nor Edwards and Lucas give anything in regard to the habits of the spe- cies, but merely state that it was found at Valparaiso. Dana, how- ever, mentions it as " from an Echinus on the coast of Chili, near Val- paraiso." A single specimen of a male, which evidently belongs to this spe- cies, was found upon the outside of an echinus which contained within it a female. This male is very small, the carapax is rather narrower * These Transactions, vol. i, p. 306, American Journal of Science, 2d series, vol xliv, p. 126, 1867, and American Naturalist, vol. iii, p. 245, 1869. 172 S. T. Smith on American Crustacea. than in the female, the chelipeds are stouter in proportion, and the ambulatory legs are somewhat less hairy. The carapax is of the same weak consistency, and the external maxillipeds of the same form, as in the female. The abdomen is quite narrow and all the seg- ments are distinct. The margins are very straight to the sixth seg- ment, which is slightly contracted, and the extremity is broadly rounded. A number of specimens give the following measurements, which are only approximately correct, on account of the soft and flexible nature of the carapax. Locality. Sex. Length of carapax. Breadth of carapax. Ratio . Callao. Male. 2 -6111111 2- 5mm 1:0-96 Paita. Female. 7-2 7-8 1: 1-08 Callao. " 90 9 2 1: 1-02 Paita. K 12-2 12-7 1 : 1-04 The genus Plnnaxodes is quite distinct from the typical species of Fabia Dana, in the form of the external maxillipeds, which are nearly longitudinal and much as in Pinnixa, with which, in fact, Heller com- pares them, while in Fabia suhquadrata, they are oblique and resem- ble those of Pinnotheres. The carapax also is quite difierent in form, and in Fahia^ the sulci which extend back from the orbits are very deep and there is no median sulcus on the front, while in Plnnaxodes, the sulci from the orbits are very slight, not more distinct than the median. Family, Dissodactylid.^. This family, which is here established for the following genus, appears to be most nearly allied to the Pinnotheridm, but diifers from that family, and in fact from all other Ocypodoidea, in the structure of the palate, or endostome, which is not divided by a median ridge separating the efferent passages. Dissodactylus,* gen. nov. The carapax is depressed, the dorsal surface is smooth and not areo- lated, and the front is narrow and horizontal. The eyes are very minute, being much smaller even than in the Plnnotheridae. The epistome is very short, so that the labial border approaches very near to the front, leaving only a narrow space which is nearly filled by the antennulse. The labial border is regularly concave, as seen in a front view, is not interrupted in the middle by any projec- * AfCCTof, duplex; (Jd/crf/lof, digitus. S. I. Smith on Ainerican Crustacea. \73 tion or emarsijination, and is continuous with the lateral margin of the buccal area, which is broad behind as in the Pinnotherldm. The pal- ate is not divided longitudinally either by lateral ridges or even by a median one, so that the efferent passages are not distinctly separated at their external orifices. In the external maxillipeds, the ischium is coaleseent with the merus as in the Pinnotheridce, and the palpus is composed of only two segments, of Avhich the terminal one is large and spatulate. The chelipeds are small and equal and the hands short and rounded. The ambulatory legs are small and slender and the dactyli in the three anterior pairs are short and deeply bifurcate, while those of the posterior pair are simple and slender. In the male, the sternum is flat and very broad, the breadth between the posterior legs being much more than twice as great as the breadth of the basal segments of the abdomen. The male abdomen is narrow and only three-jointed, the first and second segments anchylosing into one piece, the third, fourth, fifth, and sixth into another, and the terminal being free. The verges are sternal and the appendages of the first segment are large and stout, while those of the second segment are very small. Dissodactylus nitidus, sp. nov. Male. The carapax is broad posteriorly, the breadth at the poste- rior margin being but little less than that between the lateral angles, and the postero-lateral margins are about as long as the antero-lateral. The dorsal surface is naked and polished, and is slightly convex in front and along the lateral margins, but flat in the middle and poste- riorly. The antero-lateral border is slightly arcuate and is armed with an upturned margin which curves suddenly inward at the lateral angle, and extends a third of the way to the middle of the carapax. The postero-lateral border is nearly straight and is armed with a slight upturned margin. The merus in the external maxillipeds is of about equal width at base and summit, the inner and outer margins are nearly straight, and the angles at the summit are rounded. The segments of the palpus are quite long, and, when folded down, the tip reaches to the anterior margin of the sternum ; the terminal segment is spatulate and its dis- tal end quite broad and squarely truncated. In the chelipeds, the merus extends but little beyond the margin of the carapax; the carpus is short, smooth, and unarmed; the hands are smooth, rounded, somewhat swollen, and the fingers are slender 174 S. I. Smith on American Crustacea, acutely pointed, slightly detlexed, and the prehensile edges minutely dentate. There is a small tuft of dense pubescence on the inferior edge of the propodal finger near the base. The ambulatory legs are slightly hairy along the edges, and the meri, carpi, and propodi are somewhat compressed. In the first, sec- ond, and third pairs, the dactyli are smooth, naked, and divided half- way to the base; the divisions are cylindrical, acutely pointed, slightly curved, and the anterior one of each leg somewhat longer than the other. In the posterior pair, the dactyli are nearly straight, slightly compressed, sulcate above and below, and naked. The first and second segments of the abdomen are narrower than the third and are completely anchylosed, but the suture which sepa- rates them is slightly shown for a little space in the middle and each side. The succeeding piece, composed of the third, fourth, fifth, and sixth normal segments, is slightly expanded at base, considerably con- tracted at the distal end, and does not show the slightest trace of any sutures. The terminal segment is small and forms a nearly equilate- ral triangle. The appendages of the first segment reach almost to the terminal segment, they are straight for the basal two-thirds, and the terminal portion is turned sharply outward at an obtuse angle. The basal por- tion is hairy along the outer edge, and the terminal portion, on both edges. The color, in alcohol, is dirty white, the carapax marked with ii-reg- ular, transverse bands of purplish brown, and the divisions of the dactyli in the first and third pairs of ambulatory legs tipped with dark brown. Length of carapax, 4*7"""; breadth of carapax, 5'1'""'; ratio of length to breadth, 1 : 1'08. Collected at Panama by F. H. Bradley. Unfortunately only a single specimen was sent home by Mr. Brad- ley, and on this account, as well as from the minuteness of the species, the description is not so complete as might be wished. Although so small, the integument is firm and indurated, and the sexual organs are fully developed, so that it is evidently an adult. The structure of the endostome shows a very remarkable approach to the Oxystomata. The efferent canals do not, however, issue in a deep and narrow median opening as in that group, but seem to be spread out over the whole, broad, concave surface of the endostome, while the external maxilli- peds retain the form peculiar to the Pinnotheridse. The form of the S. I. Smith on American Crustacea. 175 carapax, the minute eyes, the peculiar, Ostracotheres-like, external maxillipeds, the broad male sternum with the verges arising from it, and the narrow male abdomen, show close affinity with the Pinnothe- ridfe but the union of so many segments of the male abdomen sepa- rates it again from that family. . EXPLANATION OF PLATES. Plate II. All the figures are natural size, except 2, 11, and 11* and all are copied from photo- graphs, except 6* . Figure 1. — Gelasirrms pugnax. Carapax of a male, from New Haven. Figure 2. — G. rapax. Anterior portion of the carapax of the male, seen partly iu a front view and enlarged two diameters, r Figure 3. — G. mordax. Carapax of a male, from Para. Figure 4. — G. minax. Carapax of a male, from New Haven. Figure 5. — G. armatus. Carapax of the male, from the Gulf of Fonseca. Figure 6 — G. heterophthalmus. 6, carapax of a male, from the Gulf of Fonseca. 6**, terminal portion of the ocular peduncle, on the side of the larger cheliped, with its stylet, seen in a front view. Figure 7. — G. heteropleurus. Carapax of a male, from the Gulf of Fonseca. Figure 8. — G. princeps. Carapax of a female, from Corlnto. Figure 9. — G. ornaius. 9, carapax of the female. %^ , facial region of the same specimen. Figure 10. — G. princeps. Carapax of a male, from Corinto. Figure 11. — G. gihhosus. 11, carapax of the male, enlarged two diameters. 11**, out- line of the front of the same specimen, enlarged two diameters. Plate III. All the figures are natural size, and all from photographs, except 4^ , 5, and b^ . Figure 1. — Gelasimus heterophthalmus. 1, outer surface of the hand of the larger cheli- ped. \^ , inner surface of the hand of another specimen. 1^ , anterior surface of the merus of the same cheliped as figure 1 . Figure 2. — G. heteropleurus. 2, outer -surface of the hand of the larger cheliped. 2^, inner surface of the hand of another specimen. 2^ , anterior surface of the merus of same cheliped as figure 2=^ . Figure 3. — G. princeps. 3, outer surface of the hand of the larger cheliped. 3* , ba- sal portion of the inner surface of the hand of another specimen. 3^ , anterior sur- face of the merus of the same cheliped as figure 3» . S^j, external maxilliped. Figure 4. — G. armatus. 4, outer surface of the hand of the larger cheliped. 4" , anterior surface of the merus of the same cheliped. 4*5 , 4c , 4, outer surface of the hand of the larger cheliped in two males. 2*= , ante- rior surface of the merus of the larger cheliped of a male. 2^^ , abdomen of a male. All the specimens from New Haven. Figure 3. — G. rapax. Inner surface of the hand of the larger cheliped of the male. Figure 4. — G. mordax. 4, inner surface of the hand of the larger cheliped of a male. 4a , outer surface of the hand of the larger cheliped of a young male. Both speci- mens from Para. Figure 5. — G. Pana/mensis. 5, inner surface of the hand of the larger cheliped of a male, from Panama. 5^ , anterior surface of the merus of the same specimen. Figure 6. — G. subcylindricus. 6, outer surface of the hand of the larger cheliped of a male, from Matamoras. 6* , inner surface of the basal portion of the same hand. e** , abdomen of the same specimen. Figure 7. — G. pugilator. 7, outer surface of the hand of the larger cheliped of a male. 7*, abdomen of a male. Both specimens from New Haven. Figure 8. — G. gibhosus. 8, outer surface of the hand of the larger cheliped of the male from the Gulf of Fonseca. 8* , abdomen of the same specimen. Figure 9. — G. princeps. Abdomen of a male from Corinto. Plate V. All the figures are natural size. Figures 1, !''■, 2, and 2* are copied from photographs, all the others from drawings. Figure 1. — Opisthocera Gilmanii. 1, dorsal view of the whole animal. 1*, facial region, l^" , abdomen. I*', one of the first pair of abdominal appendages. \^ , one of the second pair of abdominal appendages. All the figures from the male collected at the Isle of Pines. Figure 2. — Epilobocera armata. 2, facial region of one of the female specimens in the collection of the Boston Society of Natural History. 2* , outline of the antero-late- ral margin of the carapax of the same specimen. 2^ , external maxilliped. Figure 3. — Cardiosoma guanhumi. 3, one of the appendages of the first segment of the abdomen of a male, from the Florida Keys. 3* , side view of the same. Figure 4. — Cardiosoma quadratum. 4, one of the appendages of the first segment of the abdomen of a male, from Pemambuco, Brazil. 4* , side view of the same. Figure 5. — Cardiosoma crassum. 5, one of the appendages of the first segment of tlie abdomen of a male, from the Gulf of Fonseca. 5* , side view of the same. IV. — Ox SOME ALLEGED SPECIMEN'S OF INDIAN OnOMATOPCEIA. By J. Hammond Trumbull. Professor D. Wilson, in ^'■Prehistoric Man'''' (2d ed,, p. 63), has remarked, that " primitives originating directly from the observation " of natural sounds are not uncommon among the native root-words " of the New World." In proof of this, or as " specimens of Indian onomatopoeia," he has given twenty-six names of animals, which he had " noted down chiefly from the lips of Indians speaking the closely allied Chippewa, Odahwa and Mississaga dialects of the Algonquin tongue." Such evidence, introduced on so respectable authority, is of suffi- cient importance to invite scrutiny. Its importance was evidently not underrated by Prof. Wilson himself, for he tells us that, in " the " names of animals clearly traceable to imitation," — " this nearest ap- " proximation to verbal creation," — is to be found that which " car- " ries us back to the very foundation of language, and helps to solve " one of the profoundest problems in ])hilology." {lb., p. 55). The position that onomatopoeic primitives are not uncomonon in North American languages, will be generally conceded, — even by " those who share Prof Wilson's conviction that " the onomatopoeic " theory will neither account for the origin of language, nor supply a "complete series of roots /or any port'ion of the vocabulary T (p. 56). So far, then, as these selected specimens serve to establish that posi- tion, it matters little whether they are well or ill chosen. But so ser- viceable a collection is not likely to escape the notice of those who maintain, with more zeal and less discretion than the author of Pre- historic Man, the universality of the imitative principle in languaofe. Several of Prof Wilson's examples have already been appropriated by a well-known writer (the Pev. F. W. Farrar, in his Chapters on Language, pp. 24, 25,) to sustain the position, that, in the vocabulary of almost every savage nation, '■^almost every name for an animal is a striking and obvious onomatopana.'''"^ To this sweeping generali- zation, I shall have a word or two to say, presently. First, however, I propose to examine some of Prof. Wilson's specimens, for the pur- * Tliis assertion is quoted by Mr. Wedgwood in his volume "On the Origin of Language," (p. 29). Trans. Connecticut Acad., Vol. II. 12 July, 1870. 178 Oti some alleged specimens of Indian Onomatopma. pose of estimating the value of the whole collection, as evidence of the predominance of the onomatopoeic element in the vocabulary of the North American languages. These languages, it must be premised, have — even in their principal dialects — been so superficially studied and are so imperfectly known, that it is not always possible to trace derivatives to jjiimitives, even when the fact of derivation is obvious,^ — or to prove the negative against every assumed onomatopoeia, by exhibiting the true etymol- ogy. Of some of the names under consideration, I can say no more than that their onomatopoeic origin is wot, prima facie, apparent, and that they are quite as likely to be proved holophrastic or descriptive, as mimetic. Of others, I can more positively affirm that they have not the least claim to inclusion with specimens of onomatopoeia. Take first, the name '^ koo-Jcoosh, the sow." This is specially no- ticed by Dr. Wilson (p. 62) as " purely onomatopoeic." It is, in fact, one of a considerable group of derivatives from a well-defined Algon- kin root. When the hog was introduced by European colonists, the Algonkin tribes of the Atlantic coast adopted its English name, — modified by the characteristic affixes of the Indian animate-nouns. In Eliot's translation of the Bible in the language of Massachusetts, ' swine ' is rendered by pigs for the singular, p>W^'^9 ^^r the plural. Roger Williams, in the Narragansett, wrote, sing, hogs, and ^:>/^sy pi. h6gs-uck, pigs-uck ; Rasles, for the Abnaki, pikess, pi. p'tks-ak. Sometimes, howevei-, the Indians transferred to this (as to other newly introduced species) the name of some animal previously known, which the new-comer was thought most nearly to resemble, or they compounded a new name which denoted such a resemblance. The Narragansetts occasionally called swine by the name of the Wood- chuck or Ground Hog, Ockqutchaun, — which R. Williams describes as " about the bigness of a pig and rooting like apigP {Indian Key, ch. xvii.) This name signifies ' burrower ' or ' digger.' Similarly, the Shyennes — an off-shoot of the Algonkin stock — call the pig, the ' sharp-nosed dog ' (e Mi si ^i o tum), and the domestic cat ' the short- nosed dog ' {ka esio turn). Koo-koosh is a Chippewa form of a descriptive name which was perhaps first used by the Delawares or Nanticokes. It is found (as kwskiis) in the vocabulary of New Sweden compiled by Canipanius before 1606. The root, ko or koo, has its place in nearly all Algon- kin languages. It signifies ' sharp-pointed.' Hence, in the Massa- chusetts as written by Eliot, ko-tis, ' a thorn, or briar ;' ko-uhquodt, 'an arrow' \lit. 'sharp-tipped,' or 'sharp at the end,'J and kdwa (Narr. c6 waw ; Del. cu we), ' a pine tree,' named, as in other Ian- On some alleged specimens of Indian Onomatopoeia. 179 guages, from its pin-like leaves. Hence too, the Algonkin name of the only native animal which has a pin-like or bristly covering, — the ■porcupine. [Abn. Jcanais, ' thorn,' ' spine ;' kanuia/i, a porcupine's skin, lit. 'pin skin;' Cree, kaiok'wu, porcupine; Chippewa, kaagk; Blackfoot, ka'i ska.] In nearly all dialects, the affix 'sA (strongly aspi- rated) denotes aversion or depreciation. Fur example, in the Chip- pewa, chimaun means ' a canoe ;' chimauntsJi, ' a had or worthless canoe ;' kaugk, ' a porcupine,' and kaugk-bsh [gag-osh., Baraga,] ' a bad porcupine.' This name is etymologically identical with '■Jcoo koosh,'' a hog, — and the latter, so far from being a true specimen of onomatopoeia, is seen to be built up, from its monosyllabic primitive, to describe " a bad animal with a bristly (or, pin-like) skin." Only one other name of a quadruped appears in Dr. Wilson's list : '''■ JPe-zheio or Bi-zhew, the Lynx or Wild Cat." The Indians of Mas- sachusetts called the domestic Cat poopohs* and Dr. Pickeringf thought that this name might have been "formed from the English poor puss.^^ But Roger Williams gives pussdugh as the Narragansett name of the Wild Cat, and Rasles's Abnaki Dictionary has pesouis for ' Chat,' — which, again, Dr. Pickering thought might be a corrup- tion of " the familiar English pws5 or jowssy." Without accepting this derivation, it seems plain enough, at least, that the Xarr. pussough and Abnaki pesouis are equivalents of the modern Chippewa pe shoe orpe zhew, ' the Wild Cat,' and of the Menomenee ^«2/ shay ew. The Chijjpewa name of the Panther is mis'si-pe zhew, ' great pezhew ' or 'great Cat.' It is not impossible, certainly, — but it is hardly proba- ble, that a name which appears in so many forms and which has been given to the domestic Cat, to the Lynx, and to the Panther, origina- ted by imitation of the cry of one or another of these animals. Those who maintain the universality of onomatopoeia, are entitled to the benefit of the doubt. Of the twenty-six specimens presented, nineteen (or neai'ly three- fourths) are names of hirds. Four or five of these are apparently mimetic ; six or seven are possibly so ; and nearly all the rest are demonstrably derivative, and independently significant. As might have been anticipated, the nanus of Owls and of the Crow are among those which are least doubtfully onomatopoeic. The Chippewa lo-ko'- ko-o (Mass. kook kook haiis ; Narr. ko ko' ke hoin / Mohawk, o-ho-ho- wah ; Onondaga, ke kd a ;) represents very nearly the call of the Cat * Cotton's Vocabulary, 3 Mass. Hist. Coll., ii. 156. f In note to Rasles' Abnaki Dictionary, s. v. Chat. 1 80 On some alleged specimens of Indian Onomatopoeia. Owl {JStryx Virginiana). " Kah hah he' sha, the Screech Owl " ought not however to be separately counted as an ' onomatopoeic primitive,' for it is merely the diminutive of the name which Dr. Wilson writes "^c/A kau bail* a small owl which repeats the crj gah kaii,'''' — perhaps the Long-eared Owl {S. otus). So also, " Oo-oo-m.e-see^'' another ' screech owl,' is a regularly formed diminutive, — ' the little Oo-oo^ — denoting probably the Gray Screech Owl {&. ncevia). " Aund a gosh' kwan^ the crow " and " Gah gau ge shin, the raven," are both derivatives, in the dialect of the Saginaw Chip^^ewas, from the primitives ahn daig and ka gd gi. The former is perhaps onoma- topoeic ; the latter, obviously so. '■'■ Tchin dees, the blue jay," and ^''Denddai, the bull-frog," are counted as two specimens. The former (in Chippewa proper, dain da' see or tin dese), is a diminutive of the latter; and the jay is the " bull-frog bird." So, of the two names of ' the gull,' one ' gah yaush ko shan ' is a derivative of the other, gai ashk (or, as Dr. Wilson writes it, kuh yanshk), the more common Chippewa form, which may or may not be onomatopoeic. The first specimen in the list (and the first which is borrowed by Mr. Farrar,) is " /Shi sheeh, the duck." In the Massachusetts language. Cotton wrote this name ' se sep.'' It has the same sound in the Cree, ' see' seep.'' In Chippewa, both sjbilants are aspirated, ^ shee sheeh'' or, as Dr. Wil- son has it, ' shi sheeh!' The root, seep or sheeh enters into the compo- sition of the names of several species of water-birds or divers. In the Labrador dialect one species of duck is called masheshep [i. e. ' great sheeh '] ; Cotton gives qunusseps [evidently compounded of qunni 'long' and sep'], as one name for 'duck:' in the Chippewa, muk ud a chih (from muk ud a, ' dark ' or ' black,' and sheeh), is the name of the ' black duck,' and was the title of a famous warrior of that nation, on the Upper Mississippi, some fifty years ago ; and in the same language, the cormorant is called ka-ga-gi-v)e sheh, or ' raven- like duck;' &c. Shee' -sheeh, or se-sep, is the frequentative or inten- sive form ; and, in some of the western Algonkin dialects, this receives one or more additional syllables, as in the Shawnee, see' see' hah ; Sag- anaw-Chippewa, shi shee' he an, — both which fonns are unmistakably verbals. The root, sep, signifies primarily, to extend, to stretch out, and secondly, to dive. (Eliot wrote ' se sep a' eu^ ' he stretches him- self) The Massachusetts '■sesep^ the Cree see^ seep, the Chippewa * " Noctua lucifugans cucuhat in tenebris."— Auct. Philomela. On some alleged specimens of Indian Onomatopoeia. 181 shee sheeb, and the Sliawnee see' see' halt, are names of a diving bird, — literally, a duck. Compare, Lat. mergus horn. 7nergere ,' Dutch duycker (a dob-chick) from duycken (to bow the head). The name ' ah ah wa, a diver, a kind of duck,' is less doubtfully ono- matopteic. This bird is the totem of one of the principal families of the Chippewa nation, from which have come some of their most renowned sachems. Dr. Wilson supposes that the '*' Pau-jxm-say, the common spotted woodpecker," is so called " from the sound it makes in striking a tree with its bill." Perhaps so; but, to uncultivated ears, the name does not so exactly reproduce the sound as to compel belief in its mimetic origin. He describes the woodpecker as " spotted." Wliy may not the Indian have fixed upon the same distinguishing mark? Paupatc say is the Saganaw name. In the Menomenie, we find pah pah nch for the woodpecker, pah pah nay eio for the robin, and pahpe quoh kah for the toad. In the Chippewa, paA be Jco dain' dai is the " speck- led toad " {dain-dal meaning ' toad ' or ' frog '). In the Delaware, pa pa chees (as Zeisberger wrote it), is 'woodpecker,' and popocus, 'partridge,' or quail. In the Abnaki, the verb pepesagh i gon signi- fies 'he is spotted'' (" il est mouchete," Rasles). The modern Cree, papa tay oo, has the same meaning. If paupau say is onomatopoeic, it is certainly descriptive, as well, — and marks a ' spotted ' bird. " Moosh-kah-oos, a kind of crane which frequents marshy places, " and makes tliis sound, with a choking cry, in the evening." Moosh- kah-oos, or mooshkowese, is the Chippewa name of the bittern {Ardea lentiginosa). " Frequenting marshy places," it derives its name from Chip. maJis Icoosch, ' a marsh or bog,' or nioos-keeg, ' a swamp,' — both words being nearly related to mnsh-koo-deh, a meadow or prairie, and more remotely to Chip, mush koos ieio or mezh usk, ' green grass.' Why " No-no-no-caus-ee^'' as a name for the Humming Bird, is put among specimens of onomatopoeia, is not easily guessed. Was it sup- posed to ' imitate ' the little creature's length of bill ? No polysylla- bic name in any American language is less doubtfully synthetic and independently significant. The root nok, or nonk (with on nasal), is nearly equivalent to the Latin tener. It means ' tender,' ' delicate,' 'soft;' hence, 'light of weight' (levis), 'slender,' and sometimes, 'young.' Roger Williams translates nduk-i, by 'light.' Cotton's Vocabulary has nonk-ke, and (as a prefix or in composition) nonk-, for 'light.' Eliot wrote 7ioohk-i [it is], tender, or soft ; with an ani- mate subject, noohk-esii, [he is J tender, or soft, — applied to the flesh of a young animal, as in Genesis xviii. 7 : and in composition, he 182 On some alleged specimens of Indian Onomatopoeia. wrote nunk-omp (= light male, or young male) for ' boy,' or stripling. The modern Chippewa has nearly the same form of the animate verb- adjective, )io-1ce-see^ which, by intensive reduplication, becomes no-no- ke-see, ' he is very tender, or light.' So, in the old Alnaki, we find nan-nank-es-es-oo \j=. no-7igk-es es-ti], ' il est leger ' (Rasles), In a Chip- pewa Vocabulary published by Schoolcraft [History, &c. of the Indian Tribes, v. 599), I find " no no' Icaio s'e, the humming bird." With the double augment, as Dr. Wilson wrote it {jio-no-no-eaus-ee), the name becomes a superlative, and denotes " an exceedingly light, or slight, or delicate creature," — as if we should say, ' the tiny-tavmiest little crea- ture.' If we prosecuted our examination through the whole list of names, we should find that not more than one-fourth of them could be fairly set down as onomatopoeic. And if this is true of a few carefully selected specimens, gleaned from three dialects, how much less is likely to be the proportion of such names, in the whole vocabulary of any one tribe? It may be safely afiirmed, that by far the greater number of names of animate beings are 7iot, in any Algonkin language, onomatopoeic primitives, but are descriptive derivatives from predicative roots ; that some names of birds, reptiles and insects are ap>parently formed by imitation of natural sounds, but that the species so named are generally those which are more often heard than seen, and conse- quently more easily indentified by their cries, or by sound, than by peculiarities of form, color, or habit ;* and finally, that it is yet doubtful if any Indian name of a quadruped can be shown to be purely onomatopoeic. Of many animal-names, the composition or derivation is sufficiently obvious. Of others, the form of word or observation of changes which it has undergone in passing from dialect to dialect, enables us to say confidently that they are compounds or derivatives, and not primitives formed by imitation. How utterly unfounded is Mr. Farrar's assertion of the universality of onomatopoeia in the vocabularies of savage nations, may be shown by a few examples taken for the most part from eastern Algonkin dialects. * Thoreau, ia an account of a canoe-voyage up the Penobscot, remarked that his guide (an Abnaki Indian) " sometimes could not tell the name of some small bird which [Thoreau himself] heard and knew, but he said, "I tell all the birds about here, — this country; camU tell littlum noise, hut I see 'em, then I can tell." — Maine Woods, p. 112. On some alleged specimens of Indian Onomatopoeia. 183 The Beaver (Mass. tiunmunk ; Narrag. tummock ; Abn. tema 'koue;) is a ' Cutter-off ' or ' Feller ' [of trees]. Another name (Abn. ameskou; Del. amochk ; Cree amisk; Chip, amik;) signifies the ani- mal which ' puts his head out of the water,' i. e. the air-breathing water-animal. The Otter (Narr. nkeke ; Alg. nikik ;) is a 'Biter,' or rather, 'He ' who tears with his teeth.' The Delaware name {gtainamochk, Zeis- berger) means ' Long beaver-like animal.' The Raccoon, was called by the Delawares ' Soft hands ' {wtacke- Unsche, Zeisb.), and ' Scratcher' [nac/ieniim). The latter name is the equivalent of the Al)nalvi areskane, and the Virginian aroughcun or arocoun, corrupted by the English to ' Raccoon.' The Bear Avas sometimes called a ' night-walker ' (Narr. paitkim- nawdic); and the same name was given to the constellation Ursa Major, perhaps because it was seen to ' travel by night ' about the pole star. Another and the more common name of the Bear, signifies, I think, the ' Hugger ' or ' Squeezer ' (Cree, muskwah ; Mass. mosq ; Chip, makiod ; Del. m,achk). The Panther, in some eastern dialects, was ' Long Tail ;' in Chip- pewa and other western languages, he was the ' Great Lynx.' The Moose (Abn. moos; Narr, moos;) was a ' Smoother ' or 'Trim- mer' of trees ; so called from his manner of feeding by stripping the young bark and the twigs from the lower branches. The Oppossum, in Delaware, was ' White Face,' or ' Great "White Face.' The Horse received from the Indians of New England and Dela- ware a name which might pass, better than some of Dr. Wilson's specimens, for onomatopoeic (Narr. nay-nay-o-vm-e-wot^ R. W. ; Mass. nah-nai-ye-um-oo-adt, Cotton) ; but it is in fact a verbal, and signifies " one who cari-ies on his back an animate burden." The Chippewas called him " The animal with undivided hoofs," and sometimes " my "servant" or "my domestic animal," par txcellence {n\li). The Blackfeet named him " elk dog " {ptu no kd mi ta), and the Sioux, the " marvellous (or supernatural) domestic animal." The Bald Eagle was ' White Tail ' (Del. icoa2)alamie, Zeisb.). The Red-tailed Hawk, F. borealis, was ' Red Tail ' (Del. meechga- lanne, Z. ; Mass. mashquanon). The Swallow-tailed Hawk, K {Nau- clerus) furcatus, was the Delaware 'Fork tail' {cluinwalmine) proba- bly, which Zeisberger calls " an Eagle with a forked tail." The Turkey, in eastern dialects was 'Scratcher' (Abn. neheme; Narr. neyhom). 184 On some alleged specimens of Indian Onomatopo'.ia. Tlie King-bird, Tyrannus mtrepidus^ was called by the Narragan- setts and other New England tribes, " the Sachem." Examples might be multiplied to hundreds, but enough have been given to answer the present piirpose. If time permitted, I would direct attention to some curious features of Indian nomenclature of animals and plants that are not without interest to students of language. Just now, I will mention only one of these, namely, the generic affix, or formative, by means of which a specific or individual name is referred to a known class, family or group. For example ; the names of certain aquatic air-breathing animals, such as the Beaver, the Otter, the Muskrat, &c., receive, in some dialects, a common suffix, derived from a verb which signifies " to put the head oiit of water " or " to come to the surface ," some rodents are characterized by a generic affix as " biters," and others are, in the same way, classed with " scratchers " or " tearers." In the Algonkin, these generics follow, in some other languages they are prefixed to the sjjecific names. Thus, in Dakota nouns, the prefix ta- limits the signification to ruminating animals ; loa-, to animals of ' bear kind ;' ho-, to ' fish kind.'* Similar affixes are employed for the classification of vegetables and plants. One distinguishes such fruits (melons, cucumbers, squashes, etc.) as may be ' eaten i-aw ' or 'before they are ripe;' another [min or minne), which may be regarded as an inseparable noun-generic, makes part of the names of edible ripe fruit, grain, nuts, &c., — especially of herries and other small fruit ; a third refers to one class all plants which produce edible tubers, (potatoes, the several species of ground-nuts, &c.) ; and so on. It is true that the American languages are deficient in general names, but it is likewise true that this deficiency is in great measure com- pensated by the number of inseparable generics which enter into the composition of specific names. Sometimes this affix is purely gram- matical, — the formative of the participial or verbal which is used as a noun, — and has no independent significance. Such is the termina- tion -gun or -jegun, which characterizes a numerous class of nouns in the Chippewa and other nearly-related languages. This is the forma- tive of a participle of causative verbs, and denotes the instrument by which the action of the verb is caiised or effected. Mr. Schoolcraft was led into the error of regarding this terminal -gun or -jegun as a primitive noun, " denoting, in its modified forms, the various senses implied by our words ' instrument,' ' contrivance,' ' machine,' &c.* * Riggs, Dakota Grammar, § 62. f Information respecting tlie Indian Tribes, &e., vol. ii, p. 390. On some alleged specimens of Indian Onomatopoeia. 185 Mr. Farrar, in Chapters on Language, (p. 34), has fallen into a worse mistake. In illustration of the assumed fact that, " in some cases the " onomatopoeic instinct is so strong, that it asserts itself side by side '■'■ with the adoption of a name'''' from a foreign language, — he tells us that " the North American Indian will speak of a gun as an ut-to- ^aA-gun, or •:s. paush-ske-zi-^ww.'''' ZTt-to-tah-gun, as Mr. Farrar miglit have learned by a more careful reading of the page of ' Prehistoric Man ' from which the word was borrowed, signifies — not ' a gun,' but ' a bell' 3Ioreover, the final -gmi which Mr. Farrar mistook for an ' adopted ' English name was, as I have pointed out, merely the formative of the instrumentive participial. The Chippewa name for ' gun,' — paush-kiz'-zi-gtiti, literally ' instrument of explosion ' or ' ex- ploding instrument,' — is not more indebted to the English for its last syllable than is (in the same language) op wail gun, ' a tobacco pipe ' [smoking instrument], ne hau gun, ' a bed,' pug gi mau gun, ' a war club ' [striking instrument], or ni mi bagun, ' a water pail.' It would be easy to prove that neither ut-to-tah-gun nor 2^<'f-(sh-Jciz-zi-gini is directly or purely onomatopcsic, but the demonstration is uncalled for. It is plain enough that as illustrations of the exercise of " ouo- matopceic instinct," Mr. Farrar's examples were not well taken. V. Ox THE MOLLUSCAX FaUNA OF THE LATER TeRTIARY OF Peru.* By Edward T. Nelson, PI1.D. The followmg pages give the results of an examination of a collec- tion of fossil MoUusca from Zoi'ritos, Peru, presented to the INIiiseum of Yale College, in 1867, by Mr. E. P. Larkin and Prof F. H. Bradley. The paper is simply a preliminary one, giving a catalogue of the genera found in the collection, with descriptions of a part of the species. It is to he hoped that other collections may be received from that very interesting region, both in order to complete the fauna and to afford the means for the description of many species, which, in this collection, are too imperfectly preserved for satisfactory description. GASTEROPODA. Bulla, sp. iad. A single specimen was found, resembling Bulla Adamsii Mke., but differing in the following points. Shell less convex above and propor- tionally broader at the extremities. Aperture, below, also appears broader than in any specimen of B. Adamsii that I have seen. Fur- ther specimens may prove this to be a distinct species. The outer lip is slightly broken, and hence the following measurements are only approximate. Length 24-6 millim.; breadth 16'6 millim. Callopoma lineatum, sp. nov. Plate YI, figure 2. Shell turreted; spire elevated; whorls six (?), convex. Upper whorls slightly depressed in front, marked by a few, strong, subnodu- lous ridges, alternating with finer revolving lines. Body whorl very convex, marked above by two strong tuberculosa ridges, and laterally and below by a few revolving lines, varying in size, as on the upper whorls. Whole surface marked by very fine and numerous longitudinal lines, rather broader than the spaces between them. Aperture not observed. Length (4 whorls) 15-8 millim.; breadtli 13-8 millim. * A graduating thesis presented at the Sheffield Scientific School, Julj^, 1869. Nelson on Tertiary Mollusca of Peru. 187 This beautiful species, although quite distiuct, closely resembles both Callopoma saxosnm Wood and Callojyoma fluctuosum Mawe. From C. saxosum it may be distinguished by having the whorls less flattened above ; lacking the row of tubercles at top of the body whorl; and in having much finer and smoother longitudinal lines. From C fluctuosum it may be distinguished by lacking the strong rows of tubercles near the base of the body whorl ; by having fewer revolving lines, and stronger and more distinct longitudinal ones. Callopoma, sp. ind. I refer to this genus a very large cast found with the preceding species. It gives the following approximate measurements : length 105 millim.; breadth 95 millim. Calliostoma nodnliferum, sp. nov. Plate VI, figure 1. Shell conical and elevated ; whorls six, moderately convex ; sutures very distinct. Surface of spire marked by a few nodulous or beaded lines, six to eight on each whorl, well elevated and about half the width of the spaces between them. Body whorl convex above, keeled below, marked by the beaded lines and intermingled finer nodulous ridges. Apertu.re subquadrangular ; outer lip sharp; columellar lip covered thickly by callus. Length (4 whorls) 8*8 millim.; breadth 10-9 millim. The marking of the body whorl is very peculiar and characteristic. The strong elevated lines bear on their summits a row of nodules resembling beads, while alternating with these lines there are finer ridges, also nodulous. This species is less elevated, has more distinct sutures and fewer striae than Calliostoma lima Phil., the nearest related species. Uvanilla, sp. ind. I refer very doubtfully to this genus a specimen too poor for identi- fication. It is mostly in the state of a cast and bears resemblance to this genus. External characters mostly wanting. Breadth 77*2 millim. Crepidnla, sp. ind. Genus represented by six casts. The generic relation was proven by breaking open one of the casts, when the transverse partition became apparent. 188 Nelson on Tertiary Mollusca of Peru. Crucibulum marine, sp- nov. Most of the specimens of this genus are also casts, but a fortunate break laid open the interior of one and showed the " cup " of a Crucib- ulum. The shell is oblong-oval, twice as long as high, and smooth externally, thus diifering from all known species of the West Coast. The cup is large, semi-lunar, and apparently strongly attached to the shell along the whole of the convex side. On the free margin the cup is depressed, with a shallow sinus similar to that in C spinosus Sby, The following are the approximate measurements: shell, length 24 raillim. ; height 11*6 )nillim ; cup, length 13'4; he?ght 8 millim. Vermetus, sp. ind. I refer to this genus, doubtfully, a mass of irregular tubes which may, perhaps, be those of a species of Serpula. In the size of the tubes and manner of growth it resembles somewhat the species now living on the West Coast, but no characters remain for identification. The size of the tubes varies from six to eight millimeters. Turritella plana, sp. nov. Shell elongated, turreted, with from 13 to 19 (?) nearly flat whorls, gradually tapering to a point. Whorls flat above, slightly convex below, marked by fine, equal revolving lines, 20 to 25 in the space of 5 millim. Sutures deeply impressed and broad. Two lower whorls much more convex than the upper ones ; revolving lines stronger and crossed by distinct lines of growth. I have not seen a perfect specimen of this very interesting species, and hence measurements and the number of whorls can only be given approximately. A specimen consisting of the 8 lower whorls gives the following measurements: length 117*4 millim.; breadth 34'6 mil- lim.; breadth of upper whorl 13*4 millim. A fragment belonging appa- rently to the same specimen gives for the length of the ui^per seven whorls 35 millim. The species may easily be distinguished from any with which it might otherwise be confounded, by its nearly flat whorls and equal, thickly crowded, revolving lines ; its impressed sutures ; and the con- vexity of the two lower whorls. Turritella snturalis, sp- nov. Shell turreted, whorls twelve to fifteen; upper ones regularly con- vex ; lower ones most convex about one-fourth from the bottom of the whorls ; marked by four to seven strong, sharp revolving lines, which are strongest on the lower whorls. Above and below the point of Nelson on Tertiary Mollusca of Peru. 189 greatest convexity the strong lines are supplemented by finer and more numerous ones. Sutures very deeply impressed. No perfect specimens found except young shells. A specimen with six whorls measures: length 76'90 millim. ; breadth 25 millim. ; breadth of upper whorl 12-6 millim. A young specimen measures: length .30'1 millim. ; breadth 10*6 millim. This species seems almost as variable as abundant. Some of the specimens resemble T. tigrina Kien., but may easily be distinguished from that species by the greater convexity of the whorls, and stronger revolving lines. On all mature specimens the finer striation of the lower part of each whorl is very characteristic, but in younger specimens the striations appear nearly iiniform from base to apex. Some few specimens show occasional fine lines, intermediate between the larger and stronger ones. The place of greatest convexity of the whorls varies in a few specimens, owing to a flattening of the whorls. Lines of growth very distinct on some specimens. Turritella bifastigata, sp. nov. Shell turreted, slender ; whorls twelve to sixteen, flat or slightly concave, except the body whorl, which is regularly convex ; whorls bordered on each side by a strong obtuse ridge. Intermediate spaces ornamented by fine raised, nearly equidistant, revolving lines, about ten in the space of five millimeters. Sutures small and narrow, or rendered indistinct by the development of the bordering ridges. Body whorl somewhat convex, except in young shells ; strongly wrinkled by the lines of growth, which, on well pre- served specimens, are sharp and acute. Base of this whorl marked by from seven to ten lines, nearly as strong as the ridges of the upper whorls. Aperture rounded; outer lip thm anl slightly produced below. A specimen consisting of the seven lower whorls gives the following measurements: length 61 millim.; breadth 19" 1 millim.; breadth of upper whorl 7 millim. Nine whorls from a younger specimen gives: length 39'05 millim. ; breadth 10*6 millim.; breadth of upper whorl 3*2 millim. This interesting species shows some resemblance both to T. plana Nelson and T. goniostoma Val. But T. plana is a much stronger shell, and lacks the bordering ridges, so characteristic of this species. T. goniostoma Val. has only one bordering ridge, viz., on the lower side of each whorl, while a central ridge gives to the whorl a slight convexity, which this species lacks. 190 Nelson on Tertiary 3Iollasca of Peru. Turritella, sp. ind. Shell elongated, turreted ; whorls broad and very concave ; sutures indistinct. Surface just above each suture marked by a very strong ridge. Intermediate surface marked by a few distinct concentric lines, five to seven on each whorl. If the characters just given be constant, this species is very distinct from any of tliose described above, and from any now living on the West Coast. It is perhaps most nearly related to 2\ bifastigata Nelson, but has the whorls more concave and lacks one of the bordering ridges. Only four specimens of this species were found, all so badly worn and covered by Bryozoa and Serpulse that it is impossible to give a more detailed description. Eight Avhorls measui*e : length 63*4 millim. ; breadth 19"4 millim. ; breadth (basal, along the ridge) 26*2 millim.; breadth of upper whorls 5 '6 millim, Aphera Peruana, sp. nov. Plate VI, figure 3. Comp. Cancellaria tessellata Sby., Proe. Zool. Soc. Lond., 1832; Kiener, Iconog., p. 32, pi. 9, fig. 4. Aphera tessellata Adams; Chenu, Manuel Conch, et pale., ii, p. 276. Shell elongated, sub-fusiform ; spire short, pointed, formed by five or six moderately convex whorls. Body whorl large, three-fourths the length of the shell, ventricose. Surface marked by nearly equal longitudinal and transverse ridges, which form strong raised cancella- tions, and are so arranged as to form blunt, obtuse granulations at the point of contact. Longitudinal lines finer, and much crowded near the outer lip. Aperture oblong-oval, narrow, half as long as the shell. Lips covered with callus, which is continuous above and below the aperture. Callus of columella lip strongly reflexed over the shell, much broader above than below, almost completely covering the umbilicus. Outer lip thick, and reflexed above, furnished within with a few rather strong teeth. Inner lip with two plaits near the center, the upper one being much the stronger. There is also a plait at top of the lij), small but quite distinct. Canal wanting. Aperture prolonged into a short, open sinus. Length 17*4 millim. ; length of spire 4*4 millim. ; breadth 10 millim. This species closely resembles Aphera tessellata Adams, but is dis- tinguished from that species by its less slender form, stronger cancel- lating ridges, by its shorter and more open apertui*e, and by the third fold at the top of the columellar lip. Nelson on Tertiary Mollusca of Peru. 191 Cancellaria triangularis, sp. nov. Plate VI, figure 10. Shell ovate, ventricose, spire elevated acuminate, composed of five or six whorls. Three upper, are regularly convex, and marked by prominent ribs and lines ; the remaining whorls are very angular, flattened and depressed above. Body whorl large, very triangular, nearly two-thirds the whole length of the shell, strongly depressed. Sutures distinct, but not prominent. Ribs strong, ten to twelve on each whorl, and well marked on the top of each whorl. Whorls of spire are marked just below the su- tures by two or three distinct but fine lines, and much depressed in front of them ; and marked laterally by tln-ee strong ridges, the upper one nodulous. Body whorl with the ribs strong above, gradually disappearing below, and with nine to eleven transverse, nearly equal lines, which form, with the ribs, quadrilateral cancellations, averaging 4' millim. by 1*8 millim. Aperture long and narrow ; outer lip thin. Columellar lij) covered by a thin callus, strongly reflexed over the whorl above, and having within two strong plaits, the upper one much the larger. Um- bilicus small, nearly covered by callus, surmounted by a prominent keel. Ganal short, nearly straight and open. Length 25*4 millim. ; length of spire 7 "6 millim. ; breadth 17 millim. Cancellaria spatiosa, sp. nov. Shell ovate, ventricose; spire short, elevated, acuminate; sutures distinct, especially the one separating the spire from the body whorl. Whorls seven, convex. Body whorls very convex and ventricose, three-fourths the length of the shell, broadest near the center of the shell and rising into more or less of a shoulder above the aperture. Surface of upper whorls not examined. Remaining surface smooth, except the markings of the lines of growth. When the outer surface is removed there is seen a series of strong transverse lines, about five or six in the space of 10 millim. Aper- ture semi-oval, nearly as long as the body whorl ; outer lip sharp, marked "wdthin by rather distant teeth, which extend well into the interior, but gradually thin out. Columellar lip covered by a strong thick callus, which spreads over the convex surface of the whorl, and over the umbilical region, rising within the aperture into three strong plaits, the upper being much larger than either of the others. Canal short, open, slightly reflexed, and surmounted by a prominent keel. Our largest specimen measures: length 65-4 millim.; length of spire 192 Nelson on Tertiary Mollusca of Peru. 15 millim. ; breadth 48-45 millim. Second specimen measures : length 61'2 millim. ; length of spire 12*2 millim. ; breadth 42*2o millim. Cancellaria Bradleyi, sp. nov. Plate VI, figures 8, 9. Shell thick, ovate ; spire turreted, elevated, and acuminate, com- posed of six convex whorls, slightly depressed above. Whorls sepa- rated by distinct sutures, and marked by from 13 to 15 strong, nearly equal ribs to each whorl, and four or five revolving elevations. Body whorl somewhat ventricose, convex ; ribs more distant and accompanied on some specimens by lines of growth. Aperture oblong- oval, prolonged into a short, open, and slightly reflexed canal. Outer lip thick and smooth. Columellar lip covered by callus, almost covering the umbilical region; furnished within the aperture with two strong folds, the upper much the largest. Umbilical ridge strong and rugose. Length 27*1 millim. ; length of spire 8*4 millim. ; breadth IB'VS millim. Cancellaria Larkinii, sp. nov. Plate VI, figure 7. A fifth species of Cancellaria has the spire elevated and turreted ; whorls slightly depressed above. Sutures deeply impressed. Body whorl ventricose, three-fourths the length of the shell ; ribs strong above, but absent over the base of the whorl ; transverse ridges strong and distinct. A row of strong, acute tubercles covers the center of each upper whorl, and the point of greatest convexity of the body whorl. Outer lip very thin, and furnished within with a few strong teeth. Columellar lip with two nearly equal plaits, and a third, quite indistinct one, below. Umbilicus small, covered by a deposit of callus. Umbilical keel very strong. Canal short, open, and slightly reflexed. Owing to the bad state of preservation of our specimens it is impossible to give exactly the measurements or number of whorls. Our most perfect specimen gives, for four whorls, these measurements : length 27 millim. ; breadth 18 millim. A much larger specimen measui'es (5 whorls) length 40-1 millim. ; breadth 23 millim. Strombus, sp. ind. This genus is represented by four specimens in the condition of casts, which bear strong resemblance to the young of S. Peruviamis Swain. Outline conical, reflexed below. The largest specimen has the outer JVelson on Tertiary/ Ilollusca of Peru. 193 lip produced above the top of the shell. These charactei's, together \\dth the general form, lead me thus to refer the specimens, although specific detei'mination is impossible. Myurella tuberosa, sp. nov. Shell turreted, slender and acuminate ; whorls eight to ten, depressed or slightly concave, except the body whorl ; sutures indistinct. Cinc- ture broad, elevated, with obtuse tubercles, not as wide as the spaces between them. Longitudinal ribs distinct. Whorls marked by from four to six nearly equal transverse ridges, which rise into strong tuber- cles over the ribs. Body whorl large, over one-third the length of the shell, depressed above, convex below, rising in the middle into more or less of a shoulder. Shoulder marked by two or three concentric ridges, cov- ered by tubercles much larger than those of the others. Base nearly destitute of tubercles, but with the concentric lines very dis- tinct. Whole surface, on well preserved specimens, marked by fine, minute, longitudinal lines. Aperture elongated-oval ; ou.ter lip sharp; columella jjlicated; canal well reflexed, with the keel only moderately elevated. Only three specimens of this species were found, all having the apex slightly broken. Seven whorls give the following measurements: length 25 '2 millira. ; breadth at shoulder 8'4 millim, ; breadth at upper whorl 1 '95 millim. Myurella, sp. ind., A. A badly worn and broken specimen apparently represents another species. Wlioi'ls convex. Cincture scarcely raised above the level of the whorls, mai-ked by rather small tubercles, and separated by deeply impressed sutures. Longitudinal ribs strong. Body whorl evenly convex and without a shoulder, concentric lining indistinct. Four whorls, giving the following measurements, show this to be a less slender species than M. tuherosa Nelson. Length 2 8 "45 millim. ; breadth 10 "4 millim. ; breadth at upper whorl 8 '4 millim. If the characters given above be constant, the specimen is quite distinct from the M. tuberosa, but it has not characters suiRcient for complete specific determination. Myurella, sp- iad., B. A single specimen difiers from the species described above in having only slightly convex whorls and indistinct sutures. Cincture elevated above the level of the whorL Longitudinal ribs strong ; transverse ridges broad. Three whorls measure: length 26*2 millim.; breadth 98 millim. ; breadth at upper whorl 7*45 millim. Trans. Connecticut Acad., Vol. II. 13 July. 1870. 194 Nelson on Tertiary Mollusca of Peru. Pleurotoma, sp. ind. I refer to this genus three specimens too imperfectly preserved for specific determination or measurement. They agree in form and details with this genus, but further sjjecimens will be necessary to settle the question accurately. Conns, sp. ind., A. Three species of Conus occur in this collection. The first resem- bles C. mahogani Rve,, and might at first sight be confounded with that species. But the two may easily be distinguished by the whorls of the spire. In C mahogani the spire is regularly conical, and the whorls have all an equal slope, while in this species the whorls are slight- ly turreted. The transverse lines of the body whorl are also slightly narrower and extend further up the side of the whorl. This species is a very abundant one, both as casts and well preserved specimens. Length 20 millim. ; length of spire 5-05 millim. ; breadth 8"95 millim. A larger specimen measures: length 36'2 millim.; breadth 16'2 millim. Conus, sp. ind., B. Our second species more closely resembles Conus purpurascens Brod., but has the spire more elevated than the average of that species ; whorls more depressed above, and the transverse strite less distinct or wholly wanting. Body whorl not examined. Length 7 3 "4 millim. ; breadth 39*2 millim. Conns, sp. ind., C. This species, represented by four specimens, is remarkable for the very short spire. The shell is nearly flat above, except the last three or four whorls, which at the summit rise into an acuminate spire. Sutures very distinct. Our largest specimen gives the following measurements : length V5 millim. ; length of spire 6 millim. ; breadth 47 "8 millim. Solarium sexlineare, sp. nov. Plate VI, figure 11. Shell circular, depressed ; whorls seven to eight, modei-ately convex, separated by distinctly marked sutures, ornamented by broad, sub- equal revolving lines. Body whorl large, two-thirds the heighth of the shell, marked with four revolving lines, of which that next the suture is the broadest, the remaining ones nearly equal in size. The line which forms the edge of the whorl is double the width of the Nelso'ii on Tertiary Mollusca of Peru. 19') others. Base marked by six revolving lines. Fii-st narrow, separated by deeply marked sutures. The next four form a series, narrowing toward the interior, or umbilical region. The last, forming the wall of the umbilicus, is broad and deeply notched Umbilicus widely open. Three specimens, only, of this species have been found, all slightly worn ; it is therefore impossible to state the superficial mark- ings of the upper whorls. The species, however, appears to have been notched transversely. Length 13 "8 millim. ; breadth 25 '2 millim. This species resembles S. grannlatum Lam., but that species has seven lines on the base of the body whorl, instead of six as in our species. Polinices subangulata, sp. nov. Plate VI, figures 4, 12, 13. Shell varies from obliquely oval to sixb-globular, moderately heavy and ventricose ; spire short and pointed ; whorls from six to seven, convex ; body whorl large, nearly seven-eighths the length of the shell, convex, slightly produced anteriorly, broadest about one-fourth from top. From this point the whorl slopes, becoming very much flat- tened and presenting a marked angular appearance. Surface marked by distinct but irregular lines of growth. Sutures quite indistinct, except when the epidenuis is slightly worn ofi". Apertui-e semi-lunar, half as "svide as long, broadest a little below the middle. Outer lip sharp and thin. Columellar lip covered by a very thick callus, which rises into a more or less prominent ridge at the broadest part of the shell. Umbilicus small ; in most specimens reduced to a mere chink by the callus, which is prolonged below. Young, medium sized, and full grown specimens give the following measurements : First, Length, 12-6 millim. Breadth, 9--4 millim. Second, " 28-2 " 22-2 Third, " 47-4 " 39-2 This is the most common species in the collection. In manner of growth it resembles P. uher Val. sp., and is as variable as that species. Young specimens of the two might easily be confounded. The young are obliquely-oval ; by growth the body whorl becomes ventricose, and the flattening of the upper part becomes more distinct and prominent. The umbilicus also varies greatly. In some specimens it is open and almost circular in outline, while in others it is almost completely closed by a thick covering of callus. All full grown specimens, hence, may easily be distinguished from any species with which they might be confounded, by the short spire, the flattening or angularity of the body whorl, and the small umbilicus. 196 Nelson o?i Tertiary Mollusca of Peru. Malea, sp. ind. I refer, very doubtfolly, to this genus three casts, which resemble somewhat the young of 31. ringens Sby. Further specimens are necessary to settle their relations accurately. Argobuccinum Zorritense, sp. nov. Plate VII, figures 1, 2. Shell slender, ventricose ; spire elevated, conical ; whorls about seven, moderately convex, and depressed above. Sutures distinct, but not deeply impressed. Surface marked by strong, flattened revolving ribs, varying in width. Spaces between the ribs well marked, as wide or wider than the ribs (except on the body whorl), smooth, or ornamented with fine revolving lines. Upper ribs of each whorl somewhat nodulous, forming a more or less distinct shoulder. Body whorl large, more than half the length of the shell ; ribs wider than the spaces between them ; upper ribs forming a dis- tinct shoulder, depressed above, and forming a strong angulation with the rest of the shell ; lines of growth strong, giving to the whorl somewhat of a cancellate appearance. Aperture oblong, regularly ovate, and broadest just above the center, one-third as long as the shell. Outer lip sharp and having within numerous teeth, extending well into the interior of the shell, nearly equidistant, about one- fourth as wide as the spaces between them, and ten in the space of 5 millim. Columellar lip covered thinly by callus, which is thickened below into a distinct ridge. Umbilicus wanting. Umbilical keel strong and rugose. Canal open, short and reflexed. A large speci. men measures: length 51*2 millim.; breadth 29 millim. A smaller specimen gives the following measurements: length 35*4 millim.- length of spire 18 millim. ; breadth 19 "2 millim. This species, one of the finest of the whole collection, is very abund- ant, especially in the condition of casts. One cast measures : length 59 millim., by breadth 30 millim. On all mature specimens the nodulous character of the top of each whorl is very characteristic. On the body whorl these nodules rise into obtuse tubercles, about ten or twelve to the whorl. In mature specimens, also, the lower whorl is produced in front, having its greatest width near the central line of the whorl, and causing the aperture, when viewed obliquely, to appear somewhat quadrilateral. Young specimens differ in lacking the teeth of the outer lip, and the tubercles of the body and adjacent whorls. JVelso7i on Tertiary Mollasca of Peru. 197 Mitra, sp. ind. Three specimens have been found, which I refer to the same species, and to this genus. The spire is very elongated. Sutures distinct, whorls moderately convex. Body whorl slightly depressed and angulated above. Outer lip sharp and thin. Columellar lip covered by callus and furnished with four strong plaits. The two upper are nearly equal in size and much larger than the lower ones. The speci- mens are so badly worn and broken that it is impossible to give any characters except those mentioned above. Our largest specimen, of five whorls, gives as measurements : length 9 8 '2 raillim. ; breadth 34-2 millim. Marginella incrassata, sp. nov. Plate VI, figures 5, 6. Shell large, conical, ovate, two-thirds aS wide as long, thick. Spire rather short and acuminate. Sutures indistinct. Body whorl regularly conical, very convex, broadest one-fourth from top, forming a well rounded shoulder, and tapering rapidly from this point to end of spire. Aperture linear and narrow. Outer lip with the margin thick and broad. Columellar lip with four nearly equal, well developed plaits; the two upper more widely separated than the lower ones.* Measurements as follows : Young, Length, 20-60'nm Length of spire, 2 -60™™ Breadth 10-40mm Medium, 23-05 2-65 14-0 Mature, 27-8 3-0] 18-6 This large and fine species may easily be distinguished from any now living on that coast by its proportionate measurements, by its thicker outer lip, great prominence of the top of the body whorl, and the short spire. Oliva, sp. ind., A. This genus is represented by a specimen slightly resembling 0. palpaster Mke., but the body whorl is less regularly convex ; pro- portionally broader near the top of the whorl and hence more coni- cal. Our specimen gives the following measurements: length 37*4 millim. ; length of spire 5*3 millim, ; breadth 19'1 millim, Oliva, sp. ind., B. A badly worn specimen differs from the preceding in having a shorter spire, and the body whorl proportionally broader. N"o other charac- ters observed. Length 41-2 millim. ; length of spire 3*4 millim. ; breadth 24 millim. * The upper plait is not represented in the figure. 198 Nelson on Tertiary Mollusca of Peru. Cuma alternata, sp. nov. Plate VII, figures 3, 4. Shell slender, fusiform ; spire elevated, turreted and pointed ; whorls six or seven, convex, separated by Avell-marked sutures and orna" mented by a series of rather prominent ridges, about eight to each whorl. Ridges rise in the middle of each whoii