^IaAaA^MaaA^MM* '^Aa'A'Ja' ffmm M HiTAlAlAMJliM^JiEfflMl nmm r:m :/^'r^AA' .,^r\nr^. :^M ^^r^mm^:mMmM ^t ':^AA^A/ Mmi^mmmda^ mm&i^^mm iai ■aaAA-A'O ^^^a^^^^l Mmm^i iTJiJiTil ''^-^t^H- ^^:-'!^r'?iW ^fH\t\fS^f\hf\fi^p^h^^^^ Z» >-x> C3» x^ymfw^MHSEBKi :!>•■■ a>:> > ^tbrarg of tlj^ S^us^itm COMPARATIYE ZOOLOGY, AT HARVARD COLLEGE, CAMBRIDGE. MASS. jFountieIr h^ pribate suiiscrfption, in 1861. No. 6 ^(oO ■ 'ys3»"> ^i^;^ ^^ ^-^li <. -^-^ ms>> 2> - 1^ >>a ''^-2}r»:> 5> j»j> >>3 .»;:> ^X3> ^ J>j T US ^^^-C^^ ^'^^^^^.? >3 > j>Ts^^.->,''^ i1 ^x> »a^Jim:5V. ^ > >2i©:3> Sjt >:>;j>. ^^^3*.a?<5^: -^.^5|> -;>::^l » v^> » .'.S> ^.^^,^ ^^^^^ 5o: o:»^ '2Ji>^>i>j&>. 5J»S>^' :^«>>1>^ ^ m>^. :^i^ ^f^::i^kfc ^^^3fe;-^^^^)i^ ^'^^^>'>3>^^23S3fc^ ^ ^J32» >:> ^>3 --^ ^^ — -s p" Ja)3>3 r^o. JO»>3>^I»^ ) ^^&M.. r,i-^>> :»~^r$ ^^^^^^^53 ^»> ILLUSTRATED CATALOGUE OF THE MUSEUM OF COMPARATIVE ZOOLOGY, AT HARVARD COLLEGE. No. IV. DEEP-SEA CORALS. BT L. F. DE POURTALES, ASSISTANT U. S. COAST SURVEY. Published by pcrniission of the Siipcriiitcndcjit of the U. S. Coast Survey. CAMBRIDGE: PRINTED FOR THE MUSEUM OF COMPARATIVE ZOOLOGY. 1871. University Press: Welch, Bigelow, & Co., Cambridge. ERRATA. Page 50, line 12, for Parasmilia prolifera read Ceratocyatlius prollfer. " after line 37, add Lepidopora cochleata .... 270 C. INTRODUCTION. rpHE deep-sea corals described in these pages were collected in the J_ years 1867, 1868, and 1869, during the expeditions made for the exploration of the Gulf vStream by the United States Coast Survey, B. Peirce, Superintendent, in the steamers Corwin and Bibb, Acting- Master Robert Piatt, U. S. N., commanding. It will not be superfluous to preface their description by a condensed statement of the successive steps taken by the Coast Survey in deep-sea researches, and by a short review of our present knowledge of the con- stitution of the bottom of the sea on the Atlantic coast of the United States. The subject has been already treated in a paper published by me in Petermann's Geogra'pliislie ^Ilttheilungen, but it may be well to repeat it here for the benefit of English readers. The first step of the investigation may be said to have been taken by the late Professor A. D. Bache, on his assuming the duties of super- intendent of the United States Coast Survey in 1844, when he ordered the preservation of the specimens brought up by the lead during the progress of the hydrographical operations. The object was twofold, — a practical one, to check the indications of the character of bottom laid down on the charts by the hydrographei's ; and a scientific one, the for- mation of a collection illustrating the character of the bottom and the distribution of the materials constitutini*; it alono; our coast. The greater number of the specimens were obtained at first inside of the hundred-fathom line, but the deeper waters of the Gulf Stream soon began to add collections of still greater interest, if of less extent. Simple contrivances were substituted for the usual tallow arming the lead, so as to bring up the sand or mud in sufficient and uncon- taminated quantities. Every specimen is placed in a small bottle, labelled with the date, geographical position, and depth, and preserved in cabinets arranged for the purpose in the Coast Survey Office in Washington. The microscopical examination of the collection was begun by the late Professor J. W. Bailey, of the Military Academy at West Point, and after his death by myself, in the intervals of other duties. NO. IV. 1 2 INTRODUCTION. The general results are exhibited on the map accompanying the paper in the Geograplmclie Mittheilungen^ and representing the bottom off our coast, from Cape Cod to the island of Cuba. The bottom is repre- sented of different colors, according to its principal constituents, Avhich are chiefly of four different kinds, — silicious sand, clay (the " mud " or " ooze " of the charts), Globigerina mud, and coral sand or mud. The silicious sand and the Globigerina mud have the greatest extent. Silicious sand is, with few exceptions, the prevailing material from the shore to the hundred-fathoms line. It extends thus, skirting the shore in a broad band narrowing down to a point as far south as Cape Florida, where it ceases altogether, and is replaced by the coral formation, but reappears again on the west coast of the peninsula near Cape Sable. Clay or mud occurs to a considerable extent only to the southward of the Vinejard Islands and eastern part of Long Island, in a region knoAvn to naviijcators as the Block Island Soundino-s, and also in a series of depressions off the entrance to New York, called the Mud- Holes. This mud, of a stiff consistenc}^ and dark gray color (called blue or green by the sailors), may be derived from the tertiary formations of which the bluffs off Gay Head, on the island of Martha's Vineyard, are the remainder. Outside of the hundred-fithom line we find the Globigerina mud forming the bottom of the Gulf Stream, whose western edge coincides nearly with that line of depth, and of the greater part of the ocean, of the Gulf of Mexico, and of the deep channels separating the Bahama Banks from each other and from the neighboring lands. The discovery of this important formation is generally attributed to the expedition sent out under the command of Lieutenant Berryman, U. S. N., to sound out the path of the Atlantic Telegraph Cable in 1855. In reality, however, the discovery was two years earlier. Lieu- tenants Craven and Mafht, U. S. N., having severally, whilst engaged in Gulf Stream explorations, in the service of the Coast Survey, obtained specimens of this formation. (See Proceed. Americ. Assoc, for Advanc. of Science, Cleveland meeting, 1853.) An interesting discovery was made by Professor Bailey in examin- ing these specimens, — the transformation, of Foraminifera into green- sand. The origin of the green-sand grains in geological formations had previously been recognized by Ehrenberg; but here we have evidently the process going on at the present day, particularly along a tract at the meetino; of the silicious and Globio-erina bottoms off the coast of Georgia and South Carolina, and also to a less extent in other places. (The localities are indicated by black dotting on the map quoted ; near the mouth of New York Bay the same sign indicates tertiary green-sand, the continuation of the New Jersey beds.) INTRODUCTION. 3 The specimens obtained by the lead did not suffice to give a true idea of the deep-sea fauna. The Foraminifera alone were numerous enough to give an idea of their distribution, which is approximately attempted by difterent signs and colors on subdivision B of the same map, representing the approaches to New York. To extend our knowledge, therefore. Professor B. Peirce, when he assumed the superintendence of the Coast Survey, after the death of Professor Baclie, ordered, at the instigation of Professor Agassiz, a more thorough exploration by means of the dredge.* A beginning was made in 1867, and the work continued during part of the working seasons of 1868 and 1869. The ground explored is the part of the course of the Gulf Stream known as the Straits of Florida. (See map, Plate VIII.) Through these the stream passes from the Gulf of Mexico into the Atlantic Ocean. They run at first from west to east between the coast of Cuba and the Florida Reef and Keys, then bend in a semicircular sweep, to take afterwards a course nearly due north. On the outside of the bend, the old Bahama Channel opens into the straits through its two mouths, the St. Nicholas Channel and the Santaren Channel, separated from each other by the Salt Key Bank. In the part of the channel running due north, it is bounded by the Bahama Banks on the east and the Florida Reef on the west. In transverse sections of the channel the greatest depth is nearest its southern or eastern shore, and in a longitudinal . section the depth diminishes in passing towards the north, finding its minimum in the narrowest part between Cape Florida and the Bemini Islands, after which it increases again. In a transverse section between Key West and Havana, the greatest depth is 853 fathoms ; between Sombrero Light and Elbow or Double-Headed Shot Key, on the Salt Key Bank, 500 fathoms ; between Carysfort Reef and Orange Key, on the Great Bahama Bank, 475 fathoms; and between Caj)e Florida and the Be- mini Islands, 370 fathoms. * I cannot resist the temptation to add a short historical note on that important instrument ot the modern naturalist, — the dredge. The Museum of Comparative Zoology possesses in its library, among the books formerly belonging to Professor de Koningk, the copy of O. F. Miiller's Zoologia Danica, used by Tilesius in the voyage of circumnavigation of Captain Krusenstern in the begin- ning of this century. Tilesius, among many interesting remarks written on tlie fly-leaf, mentions the dredge invented and used by Miiller, and reprpsented in the vignette of the title-page ; he purchased it on his passage at Copenhagen, in 1803, from Vahl, Miiller's collaborator, and used it now and then during the voyage, though, as he admits, not often, as it took several sailors to handle it, the rope alone weighing eighty pounds ! The expedition must have been decidedly short and weak-handed. On their return the dredge was deposited in the collection of the Imperial Acad- emy of Sciences of St. Petersburg. To O. F. Miiller belongs the honor of having invented the dredge, very nearly in ils more modern form, and of having inaugurated its use by naturalists in a most successful manner. His quaint remarks on the hopes and disappointments of the dredger, in the Preface of the Zoologia Danica, are worth reading. 4 INTRODUCTJON. In followinor a cross section from the emero;ed coral reef called the Florida Keys, the so-called Hawk Channel is first crossed, limited outside by the living coral reef. Its greatest depth is seldom more than six or seven fathoms, generally much less towards its northern extremity ; it is often interrupted by shoals and so-called heads of live coral, and its bottom consists of calcareous mud from decomposed corals and corallines. Next comes the reef, rising nearly to low- water mark, but by no means continuous. It extends from Cape Florida south and w^est to a short distance bej'ond Key West, and seems to be slowly increasing in that direction. Although the deep blue color of the water after passing the reef seems to indicate a very abrupt slope, there is in no part of it anything to compare with the sudden deepening on the edge of the coral reefs of the Pacific Ocean, or even of the Bahamas or the coast of Cuba. The distance from the reef to the hundred- fathom line is not less than three miles, and often as much as six. In this space the bottom consists of calcareous mud, and is not particularly rich in animal life. From ninety or a hundred fathoms to two hundred and fifty or three hundred, the bottom slopes rather gently in the shape of a rough rocky floor, without great inequalities ; this formation obtains its greatest breadth, of about eighteen miles, a little to the east of Sombrero Light, and tapers off to the west, where it ends in about the same longitude as the end of the reef; towards the east and north it approaches nearer the reef, and ends gradually between Carysfort Eeef and Cape Florida. This bottom, which is called "Pourtales' Plateau," in Professor Agassiz's report, is very rich in deep-sea corals, the greater number of those described in these pages having been dredged on this ground. Outside of the rocky bottom the Globigerina mud prevails and fills the trough of the channel."^ On the Cuba shore the bottom is rocky and the slope very abrupt, particularly for the first four or five hundred fathoms. Along the Salt Key and Bahama Banks, the slope is also exceedingly abrupt, but the underlying rock is often covered with mud. Each of these regions named has its peculiar coral fauna, as will be shown afterwards. The dredgings began with a few casts in 1867 off Sand Key, Florida, and off Chorrera, a small harbor three or four miles Avest of Havana, the landing-point of one of the Florida and Cuba telegraph cables. The next year lines of soundings and dredgings were run across the St. Nicholas and Santaren Channels, from the Salt Key Bank to * See Bulletin of the Museum of Comparative Zoology, Nos. 6, 7, and 13; also Petermann's Geog. Mittheilungen, 1870, Heft XI. INTRODUCTION. 5 Sombrero Light, and on four different lines normal to the reef between Sombrero Light and Sand Key, and extending from the shore across the different kinds of bottom. In 1869 a few casts were obtained on the Cuba coast and near the Salt Key Bank, and lines of soundings and dredgings ran to the westward of Tortugas, and at intervals of about ten miles between these islands and Sand Key, and between Sombrero and Cape Florida.* The localities of the dredgings are indicated on the map by small crosses. As an appendix, I have added a preliminary list of the shoal-water and reef corals of Florida; a fuller account of them will be given by Professor L. Agassiz. In the classification used, the system of Milne-Edwards and Haime has been followed generally ; some of their subfamilies, however, I have thought proper to elevate to the rank of families and to separate altoo-ether from their former associations. The affinities of the families to each other cannot be expressed by the order of arrangement in which they are placed, as a family or group may form the connecting link between more than two others. The larger groups are still more difficult to limit than the families, and have not heretofore been hap- pily formed ; as an instance, I w^ould mention the association of the Eupsammida? with the Madreporidse, based only on the porous con- dition of the coenenchyma, whilst the polyps of the two groups are entirely different from each other. As long as we are confined to the knowledge of the hard parts of the corals, their classification must remain incomplete. For a full study of the soft parts a long residence on the spot is indispensable, as the mvestigation requires much time and patience. Much confusion has been introduced by a hasty examination of imperfectly expanded polyps, in which the shape and arrangement of the tentacles, of the mouth and the folds surrounding it, etc. is often exceedingly different from their condition in a well-developed, healthy animal. I must express my thanks for aid received, in the shape of specimens for comparison, or the gift of publications not readily accessible, to Pro- fessor P. M. Duncan of King's College, Mr. W. S. Kent of the British Museum, Mr. J. G. Jeffreys of London, Professor A. E. Reuss of Vienna, Professor A. E. Verrill of Yale College, and lastly and principally to Professor Louis Agassiz, the instigator of these researches, the com- panion during part of the cruise of 1869, and the kind and obliging friend and adviser. * The succession of the reefs or keys wliich may be mentioned in the descriptions of the corals are from west to east : Tortugas, Rebecca Channel, Quicksands, Marquesas, Boca Grande, Sand Key, Samboes, American Shoal, Bahia Honda, Sombrero, Coflin's Patches, Tennessee Reef, Alli- gator Reef, Conch Reef, French Reef, Carysfort Reef, Pacific Reef. DEEP-SEA CORALS. Family TURBINOLID^ M.-Edw. & H. Caryophyllidce Dana (pars). Caryophyllidce Verrill. With some exceptions the family of TurbinoHdee, as defined by Milne- Edwards and Haime, seems to form a good natural group. The name is unfortunately chosen from one of the most aberrant o-enera com- posing it ; but it is not a sufficient reason for changing it, as long as the genus Turbinolia and a fcAV allied ones remain in the family. The subdivision of the family into subfamilies (Caryophyllinse and Turbinolinis) characterized by the presence or absence of pali is not natural, genera nearly allied being thus separated from each other, and associated with very dissimilar ones. It would be better to make sev- eral groups of equal value, gathered around the genera Caryophyllia, Turbinolia, Desmophyllum respectively, and perhaps a part of the Parasmilidse. CARYOPHYLLIA Stokes. Cnrynpliyllia Ch. Stokes. Zool. Jonrn., t. III. p. 481, 1828. Cyathina Ehrbg., Dana, M.-Edw., & II. Caryophyllia M.-Edw. & H. Hist. Nat. des Corall. Caryophyllia formosa Pourt. Caryophyllia formosa PouRT. Bull. Mus. Comp. Zool., No. 6. Plate I, fig. 16. More or less turbhiate, on a rather slender curved or straight stem. Costae equal, distinct only near the calicle ; the latter circular or su]> ovate, moderately deep. Columella formed of four to six twisted, 8 CARYOPHYLLIA. rather loose laminae. Septa in six complete systems and four cycles, thin, prominent, sharp, and rounded on the edge, sparsely granulated. Twelve pali, equal, large, iiexuous, with few and comparatively large granules. The young are elongated, and show much diversity in their devel- opment as regards the columella and pali ; when the columella is developed early, the pali are much delayed in their appearance, and vice versa. The differences between this species and C. cyatlms are not very striking ; our specimens are, however, always smaller, slenderer, and have the columella much less develoj^ed. None of my specimens attain the size of a full-grown Mediterranean specimen. Off Havana in 270 fathoms. Off Tortugas in GO to 68 fathoms. Caryophyllia Berteriana ? Duchass. Caryoplujllia Berteriana Duchassaing. Anim. rad. des Antilles, 1850. A single specimen obtained in 68 fathoms off Tortugas, with the preceding species, differs from it and from C. formosa by its very prom- inent primary and secondar}^ septa, thus resemljling the figure of Milne-Edwards and Haime. The systems of septa are, however, all complete and very regular, so that there are twelve pali, as in all other Caryophyllia. The descrij)tions of Duchassaing and Milne- Edwards and Haime appear to have been made from the same specimen, perhaps an exceptional one. Caryophyllia clavus Scacchi. , CaryophijUia clacus Scacchi. Notiz. int. alle conch, ed. a zoof. foss., etc , 1835. Cyalh'ma tnrblnata Piiilippi. Enum. Moll. Sic. 1836. Carijophyllia pseuilolurhinoUa Mich. Icon. Zooph. 1841. Ci/athina ci/athns Lkuckart. De Zooph. coral!., 1841. Ci/at/iina pseufloturbinolia M.-Edw. & Haime. Ann Sci Nat., 3d Ser., t. XX., 184S. I have selected, from the numerous authors who have mentioned this coral, those who have given figures representing the type ap- proaching nearest my specimen. Not having the materials at hand to enable me to pronounce an opinion on Dr. Duncan's sweeping reduction of all the European species to a single one, and not wishing STENOCYATHUS. 9 to burden the nomenclature with a new name, I refer to C. clamis a number of specimens obtained near Tortugas, and also a few found w^ith C. forniosa on the coast of Cuba. The passage from one form to the other is not plainly marked, though a good deal of variety certainly exists. On the Florida side not a single specimen of the C. formosa was found. Caryophyllia cornuformis Pourt. CaryophylUa cornuformis Pourt. Bull. Mus. Comp. Zool., No. 7. Plate I, figs. 14 and 15. Corallum simple, conical, always regularly curved, distinctly but faintly costate. Calicle circular, rather shallow. Septa very httle exsert, thin, and somewhat wavy, in six systems and four cycles. Pali in front of secondary septa only, sometimes twisted. Columella of one or two twisted processes. Height about 6 mm., diameter of calicle 3 mm. OfF Sand Key in 23 7, 248, and 298 fathoms. Off Cojlma, Cuba, in 450 fathoms. All the specimens have the base broken and apparently decayed even when living, and were never seen attached. A specimen still living was agglutinated to the shell of a Xenophorus. I was at first tempted to refer this species to the genus Cera- tocyathus of Seguenza (an inconvenient name, by the way, very apt to be confounded with Ceratotrochus), but I have seen good reasons lately to remove altogether Seguenza's genus from the Caryophyllidae and place it next to Parasmilia (wdiich see). STENOCYATHUS Pourt. Corallum simple, free, very elongated, and of nearly equal diam- eter throughout ; a single crown of pali ; a columella of one or more twisted processes ; no epitheca. The single species known of this genus w\as placed by me in the genus Cc^nocyathus, having supposed the few specimens then in hand to be detached from clusters formed by budding. A larger supply has proved that they are always free. NO. IV. 2 10 STENOCYATHUS. Stenocyathus vermiformis Podrt. Coenocyathus vermiformis, Pourt. Bull. Mus. Comp. Zoo!., No. 7. Plate I, figs. 1 and 2; Plate III, figs. 11, 12, and 13. Corallum very elongated, cylindrical. Cost^ indicated only by lines of very flat tubercles. Calicle circular, shallow. Septa rather thick, flexiiose, not exsert, in six systems of tln-ee cycles. Pali thick, curled, in front of the secondary septa. Frequently one of the systems remains incomplete, and there are then but five pali. Columella of a single twisted process. The older parts of the corallum are nearly filled np by the thickening of the septa, but the process is never carried out to a total obliteration of the interseptal chambers, which can be traced in the shape of slender canals to the very base. The costal tubercles are hollow, and communicate through narrow canals with the interseptal chambers. These little cavities are no doubt homologous to the hollow roots of Rhizotrochus, Thecocyathus, and other genera of the family, but here their use is not apparent. (See Plate III., fig. 13, representing part of a horizontal section passing through these cavities.) A singularity is the frequent occurrence of specimens having a living and growing polyp at either end (Plate III., fig. 11). These specimens are generally somewhat curved, having apparently been lying in the mud with both ends turned up and projecting. A longitudinal section shows no interruption in the continuity of the interseptal chambers from one end to the other, so that the diges- tive cavity is probably common to both polyps. Height, 2.5 to 4 cm. ; diameter, 2 to 3 mm. <^)ff Boca Grande, in 125 fathoms. Off Key West, in 135 fathoms. Off Key West, in 138 fothoms. Off Sombrero, in 152 fathoms. Off Sand Key, in 154 fathoms. Off Tennessee Reef, in 1 74 fathoms. Off Bahia Honda, in 1 76 fathoms. Off Sombrero, in 183 fathoms. PARACYATHUS. H PARACYATHUS M.-Edw. & H. Paracyathus confertus Pourt. Paracyalhus confertus Pourt. Bull. Mus. Comp. Zool., No. 7. Plate VI., figs. 11, 12, and 13. Coralliim turbinate, pedicellate ; costae distinct to the base, not prominent, granular. Calicle oljlong, concave. Septa crowded, thin, entire, slightly exsert, in five cycles, but with considerable irregularity in some of the systems. Pali numerous, difficult to distinguish from the papillo3 of the columella. Like Caryophyllia, this genus presents so many points of variation that it is very difficult to define a species. I have retained the above name provisionally, not having been able to make direct comparisons with European specimens, with which I have little doubt it will ulti- mately be identified. Specimens of a Paracyathus from the Azores do not differ from mine. Off Tortiigas, in 36, 37, 60, and G8 fathoms. Oif Carysfort Reef, in 52, 60, and 63 fathoms. Off Conch Reef, in 60 and 7 7 fathoms. Off Alligator Reef, in 79 fathoms. Off The Samboes, in !>3 fathoms. Off Sand Key, in 125 fathoms. Off Double-Headed Shot Key, in 315 fathoms (dead). Paracyathus? foUicuhis Pourt. Paracyathus? folUculus Pourt. StephanophylUa folUculus Pourt. Bull. Mus. Comji. Zool., No. 7, Corallum free, without mark of adherence, broader at the base than at the calicle. Costae broad, granulated, flat, nearly meeting at the centre of the base ; the primaries continuous, the tertiaries united to the secondaries. Intercostal furrows narrow. Calicle cir- cular or subhexagonal, slightly concave. Septa in six complete sys- tems and three cycles, covered with large tubercles, not exsert (in the only specimen obtained the edge of the septa, and the summit of the pali and of the columella, appear to have been injured or abraded during the life of the animal). The primaries and secondaries meet in the centre with an indistinct columella, the tertiaries connect with 12 LEPTOCYATHUS. the secondaries about the middle of the radius. There are indis- tinct traces of paH. I have satisfied myself that the former determination of this coral was wrong ; it is probably a free Paracyathus, not very unlike P. can/ophf/llus M.-Edw. and H. Height 3 mm., diameter of calicle 2 mm. Off the Samboes, in 237 fathoms. LEPTOCYATHUS M.-Edw. & H. Leptocyathus Stimpsonii Pourt. Plate HI, figs. 1, 2, and 3. Corallum free, more or less hemispherical, without mark of ad- herence. Costa3 equal, continuous with the septa, without definite boundary, finely serrate and spinous, deepl}^ separated by very narrow furrows, so that the wall proper is invisible. The primary costie reach nearly to the centre, those of the fourth order embrace and unite with those of the third, and the resultant costaj of two half- systems unite below the secondaries. This arrangement is seldom regidar except in the young, as the base of most of the old ones has the appearance of having been broken and mended, thus introducing much confusion in the arrangement. Calicle circular ; fossa small and not deep. Septa subequal, finely denticulate, the teeth becoming larger towards the interior, crowded, in six systems and four cycles ; the arrangement of the different orders as in the costse. Pali in front of the secondary septa very distinct, those of the other orders scarcely distinguishable from the columellar processes, of which there are from six to twelve, crowded and irregularly prismatic. The young are almost spherical, the calicle being quite small ; they resemble a Melocactus deprived of its spines. The polyp is of a delicate pink or sometimes greenish color, with simple conical tentacles. Diameter 6 to 7 mm. ; height about 3 mm. Abundant in 60 fathoms off Conch Reef; one sj)ecimen in 160 fath- oms off Tennessee Reef. This coral is referred to the genus Leptocyathus with some doubt, as its denticulate septa would really remove it altogether from the lamily of Turbinolida). TIIECOCYATHUS. 13 THECOCYATHUS M.-Edw. & H. This genus, established by Mihie-Ed wards and Haime, contained heretofore only three species, fossils of the Upper Lias ; there is but one other genus of Turbinolidas as old as this one, the flimily having received its greatest development in the tertiary and present epochs. No representative of the genus Thecocyathus is known from any of the formations intermediate between the Lias and our epoch. Recent forms being found now present, therefore, a rare instance of the reappearance of a genus apparently extinct through a considerable succession of ages. The two species described below agree in their shape, which is that of a short cylinder. To maintain this shape, notwithstanding the increase of the diameter of the calicle, a peculiar arrangement is necessary. From every one of the interseptal chambers starts a hollow root, homologous to those of Rhizotrochus, but never detached, nor visible externally. These roots form a concentric circle round the original base of attachment. As the coral grows new roots are formed higher up, and a new circle concentric with the preceding is formed. In full-grown specimens five or six of these circles are produced. (See Plate V., figs. 3 and 4.) All the interseptal cham- bers emit these canals (which are filled up by extensions of the polyp body), thus differing from Rhizotrochus, in which only twelve chambers are thus provided ; in our JR. fragilis only once during the whole growth, in R. ijums apparently twice. A similar but less pronounced arrangement exists in Paracyathus, and probably in other corals. Li the Bulletin of the Museum of Comparative Zoology, No. 7, I had confounded two species under the name of T, cylmdracens . I have since recognized the differences, after examining a pretty large number of specimens. Thecocyathus cylindraceus Pourt. Thecocyathus cylindraceus Pouut. Bull. Mus. Comp. Zool., No. 7. Plate 11., figs. 14 and 15. Corallum attached by a broad base, short, cjdindrical. Costoe gen- erally visible through the epitheca, which reaches to, and sometimes over, the border of the circular calicle. Fossa shallow. Septa very 14 TROCIiOCYATHUS. finely dentate, slightly sinuous, grcanulated, not exsert, forming six systems of four cycles ; one of the systems sometimes incomplete. Pali thick, with sinuous surfaces, fronting all the septa but those of the fourth cycle ; those of the second largest. Columella thick, formed of a dense mass of papillose processes, arranged in rather indistinct rows coinciding with the pali. Height 13 to 19 mm.; diameter 9 mm. In 100 to 200 fxthoms off the Florida Reef; less common than the following species. Full-grown specimens rather rare. Thecocyathus laevigatus Pourt. Plate v., figs. 3 and 4. General form, costas, and epitheca not different from preceding species. Septa in four cycles and six incomplete systems, the fourth cycle remaining undeveloped in one of the systems and two half- systems. Septa smooth and as if enamelled, with very fine granules and entire l)order. Pali only before the septa of the third order, and seldom complete. Columella papillose, formed of four to ten pointed and smooth processes. Height 9 to 13 mm., diameter 6 mm. Rather abundant in 100 to 200 flithoms off the Florida Reef; also found living in 315 fathoms off Double-Headed Shot Ke^^s. TROCHOCYATHUS M.-Edw. & H. Trochocyathus ? coronatus Pourt. Platylrochus coronatus Pourt. Bull. Mus. Comp. Zobl., No. 6. Plate VI., fig. 16. Corallum free, cylindrical ; horizontal base with a knob in the middle. Primary and secondary costa3 having each a large tuber- cle at the junction of the vertical wall with the base. Calicle cir- cular. Septa in six complete systems and four cycles, meeting in tlie centre. Tertiary septa frequently but not regularly coalescing with DELTOCYATHUS. 15 the primary or secondary ones. Columella probably papillose (nearly destroyed in the specimen). Diameter 16 mm. ; height, including central knob, 10 mm. One dead and badly preserved specimen was brought up by the lead some years ago, by one of the United States Coast Survey parties, from 460 fathoms, in lat. 30° 41' N. and long. 77° 3' W. This specimen was described as a Pkitytrochus by a mere inadver- tence on my part. I place it now among the Trochocj^athi with a doubt, since the specimen has lost its columella and pali, if it had any. Its general appearance would place it near T. armahis M.-Edw. & H., differing from it in having only knobs instead of spines, and equally on the primary and secondary septa. DELTOCYATHUS M.-Edav. & H. Deltocyathus Agassizii Pourt. DeUocyaihus Agassizii Pourt, Bull. Mus. Comp. Zool., No. 6. Plate IL, figs. 1 to 5 ; and Plate V., figs. 9 and 10. Corallum discoidal, free at all ages. Wall nearly horizontal, or flat conical, sometimes with a nipple-shaped projection in the centre. Costae well marked, covered with spiny or smooth granules, generally subequal, but sometimes the primary costa3 much larger and pro- longed beyond the edge, forming a star. Septa in six complete systems and four cycles, covered with small spinous tubercles. Pali of the first, second, and third cycles projecting generally higher than the septa to wdiich they are soldered at the base. The pali of the first cycle short, those of the tliird joined to those of the second, but the point of junction not being exsert, the V, or delta, is not as apparent as in the fossil species. Columella papillose and small, rising from the primary and secondary j)ali which meet in the centre. (The colu- mella is represented a little too large in fig. 1, and particularly fig. 2, of Plate II.) This coral has been pronounced by Dr. Duncan identical with the fossil species D. italicus. They are certainly very closely allied ; but as far as I have been able to compare them, they are readily distin- guished by the costce, which in D. italicus are formed of large grains, and have thus a moniliform appearance, whilst all the Cuban and Floridan specimens, although in other respects very variable, have 16 DESMOPHYLLUM. the costne covered with very fine sharp graniihitions. The columelLT, of the recent species is also much less developed than in the fossil, and the pali more slender and delicate. Among the varieties, those with large and prolonged primary costsB (Plate II., figs. 4 and 5) appeared to present therewith good specific characters, until a large specimen was found which had enlarged and projecting primary costas until about half grown, and then sud- denly changed to the form with equal costae. One mutilated specimen has the primary costxe prolonged into horn-like appendages as long as one fourth the diameter of the calicle, and surmounted with a smaller prong, directed upwards. (Plate V., figs. 9 and 10.) The polyp of a large living specimen dredged in 115 fathoms, off the Tortugas, was whitish, with brown stripes between the pali and on the secondary septa. The tentacles were club-shaped, short, placed in the depressions between the septa and the pali. On the Florida side of the channel most of the specimens were immature. Until half grown the wall forms a regular cone of very large aperture (about 110°) ; at a later period it becomes flatter and saucer-shaped. One single specimen was obtained living, and was also the largest; 13.5 millimeters in diameter. Off Concli Reef, in 60 fathoms. Off Alligator Reef, in 113 fathoms. Off Tortugas, in 115 fathoms. Off Chorrera (Cuba), in 270 fathoms. Off Rebecca Channel, in 290 fathoms. Off Marquesas, in 296 fathoms. Off Boca Grande, in 327 fathoms. DESMOPHYLLUM Eiirbg. Desmophyllum Cailleti Duch. «fc Mich. Desmophyllum Cailleti Duch. & Micri. Suppl. Mem. Coral., 1864. Six fine living specimens of this coral were obtained in 315 fathoms, off the Double-Headed Shot Keys. They agree with the definition of the above species, but are more elongated and more contracted at the base than the fig-ure of those authors. Wall and septa are very thin and fragile, and the whole coral resembles Rhizotrochiis fragilis very much. The largest specimen is 23 mil- limeters in height, and 13 in diameter at the calicle. RHIZOTROCHUS. 17 Desmophyllum solidum Pourt. Plate v., figs. 5 and 6. Coralliim conical, peduncnlate, heavy. Septa very thick, finely granulated, in six complete systems and three cycles. The primaries and secondaries are nearly equal, somewhat exsert, the tertiaries are lower, but twice as thick ; all the septa are so nearly in contact that the interseptal chambers are almost obliter- ated, except near the middle of the calicle. CostaG also very thick, continuous w^ith the septa, not continued far from the border of the calicle. Height 11 millimeters, diameter eight. One specimen (in a semi- fossil condition) found with the preceding species in 315 fii thorns, off Double-Headed Shot Key. There are several species of this type with thick septa in the tertiary beds of Sicily, which have been described by Seguenza; none, however, appears to be identical with this. RHIZOTROCHUS M.-Edw. & H. Rhizotrochus fragilis Pourt. RMzolrochus fragilis Pourt. Bull. Mus. Comp. Zobl., No. 7. Plate lY., figs. 1 to 4. Corallum simple, pedunculate, straight or slightly curved, regu- larly conical. Calicle sul^elliptical, deep. Six complete systems of septa, four cycles. Septa very thin, not exsert, finely granu- lated ; those of the first and second order meeting in the centre, and united for about half their height. Costa3 not prominent, no epitheca. Hollow roots attached to the costae of the second order in pairs, at about one third or one half the height of the corallum, and continued down to the peduncle, where they slightly expand ; they are never detached from the wall. The whole corallum very thin and fragile. The young at first form a six-sided rounded cell, wider at the base of attachment than at the top. The six primary septa reach only half-way to the centre. At the next stage they unite in the centre, frequently leaving a small open space in their middle. Simultaneously or after the appearance of the secondary septa, a 18 TROCHOSMILIDJE. circular wall arises from the floor and separates the interseptal chamljers in two, the outer portion becoming the rootlike append- age. The cavity of the roots opens freely into the interseptal chambers (Plate XIY., fig. 3), and is occupied by a fibrous extension of the polyp. In old specimens the cavity becomes much reduced by the thickening of the walls. The color of the polyp is greenish, or pale brick-red. This coral was obtained in about forty different casts along the Florida reef, the least depth in which it was found being 49 fathoms, the greatest 324. The greatest abundance was between 100 and 200 fathoms. Height 2 to 2.8 cm., diameter of calicle 1.5 to 2 cm. TROCHOSMILID^. Trochosmiliacece M.-Edw. & H. For reasons elsewhere stated, I prefer to elevate this group to the rank of a familv, instead of leavino- it amouo; the Astranda^ as a mere section of the subfimily of Eusmilina^, the place assigned to it by Milne-Edwards and Haime. Its affinities are j)artly with the Astraiida? and partly with the Turbinolida?. We are not sufficiently acquainted with the value of the characters connecting it with either to say positively wdiich of them ought to have the preponderance. As a family, it appears homogeneous, but the order or suborder to which it ought to be attached can be left undetermined for the present. The principal character by which to separate the Trochosmilida? from the Turbinolida? is the presence in the former of dissepi- ments closing the interseptal chambers. In a large number of them they are of a very rudimentary character, and no more de- veloped than the}^ are occasionally in true Caryophyllia^ where I have observed similar structures secreted abnormally to exclude parasites. The mode of propagation observed in three of the species here described, namely, a gemmation from the interior of the calicle, recalls the peculiar intracalicinar gemmation of the Rugosa, and does not prevail among any of the families of corals of the present epoch. CERATOCYATHUS. 19 CERATOCYATHUS Seguenza. Having received, through the kindness of Mr. J. G. Jeffre^^s, some specimens of Cerafoci/athus ornahis Seguenza, from the dredgings of the Porcupine Expedition off the West Coast of Ireland, I have satisfied myself that the genus does not belong to the family of Car3^ophyllida3 at all, but to Trochosmilidie. If the European species from the deep-sea dredgings and the Sicilian tertiary beds does not show very distinct dissepiments (at least in the specimen of which Mr. Bicknell made a section for me), it is because the interspaces have been filled up, as a careful inspection shows plainly. The pali, which give to these corals so striking a resemblance to the typical Caryophylliae, are found, when traced to their origin, to be formed by lobes of the septa, and not by inde- pendent growths rising from the bottom of the calicle. Ceratocyathus prolifer Pourt. Plate III , figs. 8, 9, 10. Corallum free when adult, conical, curved, showing no trace of attachment. Costa? finely granulated. No trace of epitheca. The Avail rather thick and of compact texture. Calicle circular or slio'htlv elono-ated in the direction of the curvature ; fossa moder- ately deep. Sej)ta finely granulated in six systems and five cycles, but the fifth cycle generallj^ incomplete in some of the systems. The septa of the third order, and sometimes those of the fourth, have a paliform lobe, which in older specimens becomes merged in the columella. (This is the case in fig. 9.) Dissepiments rather massive, about eight or nine occurring in a chamber (Plate III., fig. 10). The base of the coral has a tendency to become filled up. Columella spongy with papillose convex surface ; in the young it is little developed, and sometimes lamellar, so that an isolated specimen would be liable to be classed as a Placosmilia. The propagation is by buds inside of the calicle of the parent, not unfrequently two or three growing out at the same time, but in such cases one of them outgrows and stunts the others. The parent is probably killed by the growth of the offspring, which fills the whole calicle before becoming free, but the wall being very strong 20 PARASMILIA. it is never broken as in ParasndUa Lymani. Out of twenty-four specimens obtained, nine show this mode of multiplying. Height 15 to 20 mm., diameter 12 mm. Found onl}^ in one locality, off French Reef; in 45 fathoms. ParasmiUa nutans Duch. & Mich, is perhaps nearly related to this species, but the description and figure are too vague for identifi- cation. PARASMILIA M.-Edw. & H. Parasmilia Lymani Poukt. Plate VI., figs. 8, 9, 10. Corallum more or less elongated, conical, nearly always attached to a fragment of the parent, otherwise free. Wall very thin. Costa3 small, huely granulated, subequal or alternately larger and smaller, equally developed from base to top. Calicle circular or slightly elongated or irregular. Fossa moderately deep. Five cycles of septa, and six very irregular systems, having often the appearance of a greater number ; the septa entire, moderately exsert, with few rather large and flat granules joined sometimes in oblique ridges. The internal edge of the septa is deeply divided into lobes forming false pali and sometimes a false columella (Plate VI., fig. 10). The paliform loljcs are formed by the penultimate septa ; whilst the twisted ones, simulating a columella, rise from the other cycles, except the last. Dissepiments rudimentary, membraniform, nearly horizontal, only two or three in number in an interseptal chamber. They are too faintly represented in fig. 1 0. The propagation is by buds on the septa in the interior of the calicle, in the parent, the walls of which are burst by the growth of the young ; the latter carrying a fragment attached to the base generally through life. The largest specimen (incomplete at the base) is 4 cm. high, 18 mm. in the greatest and 13 in the smaller diameter. OfT Elbow Reef, in 70 fathoms. Off Alliirator Reef, in 79 fathoms (dead specimen). Off the Samboes, in 96 fathoms. Off" Alligator R. ( f, in 110 fathoms. Off Conch Reef, in 117 fathoms. Off Alligator Reef, in 147 fathoms. CCELOSMILIA. 21 Farasmilia variegata Pourt. Plate I., fig. 13. Corallum simple, free, conical, often deformed ; the irregnlar shape resulting from the young being originally attached to fragments of the parent, which become gradually covered over and incorpo- rated into the base. No epitheca, costge distinct to the base, flat, finely granulated, those of the third order somewhat broader and more prominent than the others. Calicle oval, with moderately deep fossa. Septa in six complete systems of four cycles, consider- ably exsert and very unequal ; those of the first and second order being; closelv connected at the wall with those of the fourth and fifth respectively, and rising high above those of the third order, which are the lowest of all. Septa entire on their upper edge, but deeply laciniated in the interior of the calicle, which thus becomes filled with a false columella of curled and twisted septal processes. The primary and secondary septa are colored of a dark purple, which color continues some distance down the corresponding costie. Those of the fourth cycle are lighter colored, with a dark spot on the corresponding costae, a short distance below the border. Those of the third order are white, with the same dark spot on the costa. Dissepiments not observed, none in the upper three fourths of the corallum. The young bud out of the calicle of the parent, which is split in consequence. Height 15 to 16 mm. ; diameters of calicle 10 and 13 mm. OlF Conch Reef, in 60 fathoms. OfFCarysfort Reef, in 63 fathoms. Off Conch Reef, in 7 7 fathoms. CCELOSMILIA M.-Edw. & H. Coelosmilia fecunda PorRT. Plate L, fig. 12 ; Plate IIL, figs. 4 and 5 ; Plate VI., figs. 14 and 15. Corallum elongated, conical, irregularly bent, showing sometimes a succession of slight swellings and contractions, attached by a slightly thickened base. Costae finely granulated, rather obsolete near the base. Calicle circular or subelliptical, with deep fossa. Septa entire, granulated finely, not exsert, in six systems and four 22 OCULINID^. cycles, the fourth developed in part of the systems only. The inner edge of the septa sends out trahicules uniting together to form the rudimentary columella, which is, however, frequently absent. Dis- sepiments thin but complete, moderately abundant. Gemmation ex- ternal, at all heights, but generally near the calicle ; abundant, one specimen having as many as eleven buds ; some good-sized specimens, however, show none ; as, for instance, the one figured on Plate III. Height of largest specimen 4 cm., diameters of calicle 6 and 7 mm. West of Tortusras, in G8 fathoms. Off Havana, in 270 fathoms. Off Double-Headed Shot Key, in 315 fathoms (dead specimen). The generic affinities of this coral are a little doubtful. By its general aspect it reminds one of Cladocora, or rather Goniocora ; but according to the classification generally adopted, it is widely sepa- rated from it on account of the entire border of the septa. In its general features it approaches the genus Onchotrochus Duncan ; but this is described as destitute of endotheca, and therefore placed among the Turbinolians. Family OCULINIDiE Verrill. OcuUnacece M.-Edw. & H. By constituting the section of Stylasteraceae of Milne-Edwards and Haime into a separate family, the Oculinidne become restricted to the Oculinacea3 of these authors, and form thus a much more homogeneous and natural group. OCULINA M.-Edw. & H. Oculina Lamk. (pars). Oculina varicosa Lesueur. Oculina imrirn^a Lksueuu. Oculina varicoxa Dana. Oculina Petireri M.-Edw. & H. Oculina dislicha PouRT. Bull. Mus. Comp. Zobl., No. 7. Plate VI., figs. 3 and 4. I have become convinced that the specimens described by me as DIPLOHELIA. 23 Oculina disticha were fragments of the younger branches of OcxiUna varicosa Les. They were ol^tained Off Conch Reef, in 40 fathoms. Off American Shoal, in 43 fathoms. Off French Reef, in 50 fathoms. The description and figure of Lesueur are so imperfect, that they may apply to ahnost any species of the genus. Dana's description, however, really establishes the species, and is prior to Milne-Edwards and Haime's, who evidently referred to the same object. Why they should make Dana's species a synonyme of 0. difma is not very clear. , Oculina tenella Pourt. Plate v., figs. 11 and 12. Corallum branching irregularly, slender. Coenenchyma finely gran- ulated. Costse distinct only on the border of the calicles, which are circular, irregularly alternate, very prominent, about one diameter apnrt, and as wide or slightly wider than the branches. Fossa mod- erately deep. Septa finely serrated, slightly exsert in six systems and three cycles. Pali before the primary and secondary septa. Colu- mella papillose, more deeply sunk than the pali. Smaller branches about 2 mm. in diameter. Calicles 2 to 2.5 mm. Off Tortugas, in 36 fathoms. Oculina diffusa Lamk. Was obtained living as deej^ as 15 fiithoms. DIPLOHELIA M.-Edw. & H. Attention has lately been drawn by Dr. Duncan to the difficulty of distinguishing from each other the genera Diplohelia, Amphihelia, and Lophohelia of Milne-Edwards and Haime. Like others of the genera established by these authors and based on very few species, they become very perplexing as our collections increase and the distinc- tive characters are found not to be so precise as they were at first enounced. My opportunities for the study of this group have 24 DIPLOHELIA. not been very extensive, still they are probably sufficient to illus- trate and partly confirm Dr. Duncan's views. The following are the species examined, keeping the usual generic names : — 1. LophohcUa iwoUfera M.-Edw. & H., Florida and North European seas. 2. Lopliohelia cxigua Pourt., Florida. 3. AniphihcUa oculata M.-Edw. & H., North European seas. 4. Amphihdia Carolina Pourt., Cuba. 5. Amphihelia iufundibuUfera Lamk. sp.. East Indies ? 6. Amphihelia miocenica? Seg. (^. ramea Sars), Norway. 7. Diplohelia rarislella M.-Ed. & H., fossil, France. 8. Diplohelia profunda Pourt., Florida. The first and second have no columella, entire se^^ta, thin caenenchyma (but quite distinct) ; the third has also entire septa, a small columella, thick caenenchyma ; the fourth, entire septa, no columella, thick caenenchyma ; the fifth, serrate sej)ta, no columella, thin ciBuenchyma; the sixth, serrate septa (?), a columella, thin caenenchyma ; the seventh and eighth, serrate septa, small colu- mella, moderate caenenchyma. It is evident from this arrangement, that, giving to the above characters the weight usually attached to them, these seven species would be assigned to at least five distinct genera. The distinction on account of presence or absence of caenenchyma falls to the ground, since it is only a question of more or less. The same might be said of the presence or absence of columella, since it is almost absent in some of the corallites of Amphihelia oculala for instance, and, on the other hand, rudiments can be found in some corallites of Lophohelia prolifera. There remains, then, only the serrate or entire edge of the septa, which in some fixmilies of corals constitutes a character for their division into subfamilies. It is in many cases a very unsatisfactory character, particularly to the paleontologist, and has given rise to serious mistakes, an example of which will be shown in another part of this paper (AstrocjBnia). For the present, and considering the small number of species under comparison, it seems proper enough to combine the genera Lophohelia and Amphihelia into one, as there is really no permanent character to separate them. The genus Diplohelia might, however, be kept for those species in which the corallites never appear individualized, even when young, but form, as it were, part of the branch with the same diameter throughout. The character is not a very good one, perhaps, still there is something distinctive in it which strikes at first sight. If living specimens of Diplohelia profunda can be procured (mine are all more or less worn), some otjier characters may be detected. LOPHOHELIA. 25 This arrangement seems to me more natural than to combine Amphihelia and Diplohelia into one genns, and leaving out Lopho- helia, as Dr. Duncan has done. Diplohelia profunda Pourt. Diplohelia profunda Pourt. Bull. Miis. Comp. Zool., No. 6. Plate VI., figs. 6 and 7. Corallum branching, cylindrical, finely granulated or striated, par- ticularly on younger branches and near the calicles, where the granulations assume the shape of rudimentary costae. Calicles slightly projecting, very deep, irregularly alternate, and very dis- tant from each other. Septa in three cycles and six complete systems, subequal, not exsert, finely serrate and tuberculated, nearly meeting at the bottom* of the fossa. Columella formed of six or seven club-shaped styles, not very distinct from the septa. Height, 3 or 4 cm.; diameter of branches, 4 mm.; of calicles, 2.5 mm. This species resembles D. raristella, but has fewer, deeper, and lars^er calicles. It seems to differ also from D. Doderkiniana Seguenza by the same characters ; in fact, the difference between Seguenza's species and D. raristella is not very apparent. Off Havana, in 350 fatlioms. Off Bahia Ilon.la, Florida, in 324 fathoms. In lat. 28° 24' N., long. 79° 13' W., in 1050 fathoms. In all cases, dead and more or less worn specimens only were obtained. LOPHOHELIA. Lophohelia -\- Amphihelia M.-Ed\v. & H. Lophohelia prolifera M.-Edw. & H. Lophohelia affinis Pourt. Bull. jNIus. Comp. Zool., No. 7. Plate I., figs. 3, 4, 5. My species is identical with L. prolifera, as numerous comparisons with specimens from European seas have shown. Off Coffin's Patches Florida, in 195 fathoms. Off Double-Headed Shot Keys, in 315 fatLoms (dead fragments). 26 STYLOPHORIDiE. Lophohelia exigua Pourt. Plate I., figs. 6 and 7. Coralkmi branching; the corallites alternate, forming angles of 70°— 75° with their predecessors, proportionally shorter and more open than in L. iwoUfcra. Wall smooth or very finely granulated. Costce sharp and prominent, but not extending far downward. Cal- icle circular, well opened, very deep. Septa thin, slightly exsert, in six systems and four cycles, the last almost rudimentary. Height, 3 to 4 cm. ; height of single corallites, 4 to 5 mm. ; diam- eter of calicle, 4 to 6 mm. Off Tortiigas, in 36 and 79 fathoms. Off Pacific Reef, in "287 fathoms (dead fragments). The fragment from the last locality is doubtful, being much worn. Lophohelia Carolina Pourt. Corallum branching, with smooth and very fiiintly striated caenen- chyma, dense and heavy. Calicles regularly alternate, cost^ dis- tinct on the border of the calicles only. Calicles prominent, not very deep ; septa with entire border, not exsert. Six complete systems, three unequal cycles, the first nearly meeting in the centre ; no columella. The branch from which this description is taken is 18 cm. long, 8 mm. in diameter in the thickest part, calicles 4 mm. This species is named according to the desire of Mr. R. Arango of Havana, who sent it to the Museum. The depth from which it was obtained is not stated, but it came probably from a fisher- man's line or net, as many other interesting specimens in that gen- tleman's collection. Family STYLOPHORID^E Pourt. Stylopliorince M.-Edw. & H., 1857. PseudoculincB M.-P:d\v & IT. Ann. Sc. Nat, 3d Ser., A^ol. XIII., 1850. This group was jDlaced by Milne-Edwards and Haime at the end of the Oculinidae, as a subfamily intermediate between the Ocu- linida3 and the Astraeidge, more nearly allied to the former than to MADRACIS. 27 the latter. This rehitionship is based chiefly on the density of the ca?nenchyma, but this is a very variable character, and some genera left by the same authors among the AstraeidoB proper, such as Cy- phastroea for instance, have it just as compact as Stylophora, though less abundant. The reliance on this character alone has induced Milne-Edwards and Haime to establish the two genera Axohelia and Madracis, and to place the one among the Oculinidge proper and the other among the Stylophorinae, whilst in reality the two genera can- not be separated by any other characters. The group, on the whole, does not appear to be allied closely enouffh either to the Oculinida3 or the AstraeidiB to warrant their combination with the one or the other as a subfamilv. It is better for the present to leave it as a small intermediate family. MADEACIS PouRT. Madracis M.-Edw. & H. Axohelia M.-Edw. & H. A comparison of several representatives of the genus Madracis with an Axohelia from the West Indies, very closely allied to, if not identical with, Axohelia mijriader, M.-Edw. & H., has convinced me that the two genera cannot be separated, and much less placed in different subfamilies. The only difference is in the c^enenchyma, more abundant and compact in the one than in the other; but this is a difference only in degree, for in both forms the caenenchyma becomes solid, but the interseptal chambers do not fill up com- pletely as in the true Oculinidge. The younger parts of Madracis decactis show the space between the calicles proper and the mural ridge occupied by large ^vesicular spaces, which afterwards become gradually filled up. I propose to retain the name Madracis for the two genera com- bined. Madracis asperula M.-Edw. & H. Plate VII., fig. 4. The specimens which we refer to this species are in general a little slenderer than the one figured in the Annales des Sciences 28 MADRACIS ASPERULA. Naturelles, but there are no sufficient grounds for separating thein specifically. The color is variable, sometimes light brown with l)lack or dark brown calicles, or purplish pink of different shades, with or without darker calicles. West of Tortugas, in 3G fathoms. Saint Nicholas Channel, off' Bahia de Cadiz, in 120 fathoms. Madracis decactis Verrill [Astrcca decactis Lyman, Proc. Bost. Soc. Nat. Hist., Vol. VL, 1857), Plate VII., figs. 1, 2, and 3, is found to a depth of 17 fathoms. It is generally thin and incrusting, but also rising in club-like masses. A specimen received from Mr. R. Arjingo, in Havana, forms thick branches bluntly expanded at the end, 6 cm. high, and 4 cm. in diameter.* Sb/lophora incrudans Duch. & Mich. I strongly suspect to be a young Madracis decactis, f and Heiima lamcl- losa Duch. & Mich., the full-grown form. % Slfjlopliora mirahilis Duch. & Mich, is a Madracis also, wdth mas- sive cfenenchyma. The Museum of Comparative Zoology has also received from Mr. Arango, in Havana, specimens of a coral, which agrees with the descrijDtion of Axohelia myriader, M.-Edw. & H. It differs from Madracis mirabilis chiefly hj its striated ca?nenchyma and larger calicles, which, instead of being prominent, are rather sunk below the surface. * The polyp is purplish brown, tentacles tipped with white ; disk emerald green, mouth yellow. The mural lines separating the polyps are tipped with white. The tentacles are in one circle, five of" tiiem large, prominent, and almost spherical when fully expanded, the white tip swelling up in that shape. Tiie other tentacles are in groups of thi'ce between the larger ones, the total number being thus twenty, or double the number of the se[)ta. There are thus five primary, five secondary, and ten tertiary tentacles. The disk, when fully expanded, projects in the shape of a cone, with the mouth at the apex. The water is kept in a whirling motion over the mouth by ciliary action. The animal bears handling and taking out of the water very well, expanding again after a few min- utes of rest. t Duchassaing et Mi(;helotti, Supplement au Memoire sur les Coralliaires, Plate IX., fig. 3, is evi- dently a magnified portion of Sli/laster chf/ans, represented on fig. 4. Fig. 2 may be intended to represent S/ij/ophnra iiicnistans, the nund)ers having been niis[)laced. X There is again here (Duchassaing et INIichelotti, Memoire sur les Coralliaires des Antilles) a confusion in the lunnbering of the figures, and in the figures themselves. One of the generic char- acters of Reussia (diifering in no way from those of Madracis) cons'sts in haA'ing always ten septa. Fig. 9 of Plate IX., purporiing to be a magnified portion of lleussin lamellosa, shows from fourteen to twenty-seven ! Fig. 8 agrees well with the large specimens of Madracis decactis from Havana. STYLINACEiE. 29 Family ASTR^IDiE M.-Edw. & H. (pars). As the typical Astrseiclce are not represented in the deep-sea fauna, it would lead too far to attempt a revision of the family or even sug- gestions towards it in this place. But one group of genera I have felt justified in separating from the Astraeidge, as I have remarked before, under the head of the Tro- chosmilidffi. It may remain for the present an open question if they ought to form a family by themselves, and if so, what place that family ought to occupy and what other corals might be added to it. Most of the Astraeidge described here will be found to be adven- titious, having been drifted from shallow water. STYLINACEiE M.-Edw. & H. Galaxea eburnea Poukt. Plate HI, figs. 6 and 7. Corallites cylindro-conical. Wall smooth, shining, concave between the very flat cost^e of the two first cycles ; no costae for the third cycle. Calicle regularly dodecagonal ; fossa small and deep. Septa smooth, very little exsert, rather thick, very regular, in six systems and three cycles. Dissepiments few and deeply seated. Height 2 cm., diameter 6 mm. This description applies to the only specimen obtained off Havana in 270 fathoms. Although fresh, it had no trace of the peritheca except discoloration for about half the height. Since then I have seen specimens from the collection of Mr. R. Arango in Havana, having four cycles of septa in some of the systems, costae for all the cycles, and cellular peritheca reaching to al)Out one third the height. They were also detached corallites. There is probably but one species, notwithstanding the differences indicated. 30 ASTRANGIACE^. CLADOCORACEiE M.-Edw. & H. CLADOCORA M.-Edw. & H. Cladocora (pais) Ehrbg. & Hemp. Cladocora debilis M.-Edw. & H. Specimens of a Cladocora somewhat different from C. arhuscula, common in shoal water, are referred to this species. They are slenderer, less branching, have a more shallow fossa, well-developed pali and columella, and more exsert septa. Off Sand Key, in 26 fathoms. Off French Reef, in 45 fathoms. Off Elbow Reef, in 49 fathoms. ' Off Carysfort Reef, in 63 fiithoms. Cladocora arhuscula, as far as my observations go, is fomid from low- w^ater mark to 13 fathoms. It is possible that Cladocora debilis may be merely a deeper water variety of this species. ASTRANGIACEJE M.-Edw. & H. PHYLLANGIA M.-Edw. & H. Sijnclepas Lymax. Proc. Best. Soc. Nat. Hist., VI. p. 274, 1857. Phyllangia americana IVI.-Edw. & tt. Syndepas Gouldii Lyman 1. c. Stellangia reptans? DucH. & Mich. Off Donble-Headed Shot Key (Elbow Key), in 315 fathoms. Dead specimens, rather worn, Avere dredged up with quite a number of other dead corals at this place, shoal-water and deep-water species being mixed together. The locality is very near the edge of the Salt Key Bank, at the foot of a very steep submarine slope, and washed by the edge of the Gulf Stream. COLANGIA. 31 ASTRANGIA M.-Edw. & H. Astrangia solitaria Verrill. Asti-angia solitaria Vkrrill. Bull. Mus. Conip. Zool., No. 3. CaryoplnjUia solitaria Lesueur. Jour. Phil. Avad. Nat. S(;., I., p. 180, PI. VIII., fig. 11, 1817 ; and Mem. (lu Mus., T. VI., p. 273, Plate XV., fig. 1. Astrangia phyllangioides f DuCH. & Mich. Dead specimens off Double-Heacled Shot Key, in 315 fathoms. The same remarks apply to this species as to Phijllangia americana. Both are rather common shoal-water species. COLANGIA PouRT. Corallum immersed in an expanded epithecal membrane, forming several successive stories, following the growth of the coral. Pri- mary and secondary septa entire ; those of the lower cycles dentic- ulate. Well-developed pali in front of the tertiaries. Columella lamellar or papillose, not much developed. The genus is chiefly dis- tinguished by the highly developed epitheca from Phyllangia, which has, none, or only a false one formed by a coralline growth. Colangia immersa Pourt. Corallum incrusting, in loose clusters, the corallites tending to develop in rows, immersed in an epithecal membrane, stretching out horizontally from the margin of the calicles, and forming anew when the corallites increase in height. No costee visible. Calicles round or somewhat elongated ; fossa shallow. Septa thin, smooth, with very few granules, in six systems, generally complete. Four cycles. Primaries exsert, with straight perpendicular inner edge. Secondaries not exsert, having, like the primaries, an entire edge. Tertiaries with rounded denticulations. Pali prominent, thin, rounded, in front of tertiary septa. Columella in shape of an irregular lamella with a few lateral papilla3, or sometimes simply papillary. Diameter of calicles, 7 mm. Like the other Astrangiacese, this is a shallow-water form, found on 32 STYLASTERID^. the reefs, though ratlier rare. Dead specmiens were found off Double- Headed Shot Key (Elbow Key) in 315 fathoms, with the preceding species. The association of three allied shoal-water forms transported to this deep-water locality is rather singular. Family STYLASTERID^ Gray. Stylasteracece M.-En\v. & II. (pars). This family was first established by Gray (Ann. & Mag. Nat. Hist., Vol. XIX., 1847), for the genus Sti/laster alone. M.-Edwards and Haime made of it a group or subfamily [agele), and placed it among the Ocu- linidge under the name of Stylasteracea3. They have, however, left out the genus Errina Gray altogether, placed Didicliopora among the genera of doubtful position, but included Axohelia, which is a Madracis. Otherwise the limits of the group are the same which we shall use. Professor Verrill first recognized the close affinity of Distichopora, Errina, and Stylaster (Bull. Mus. Comp. Zool., No. 3, 1864). In his "Notes on Radiata" (Trans. Conn. Acad., Vol. I, 1870) he adopted a suggestion of mine to make a distinct family of the StylasteridaB, which he places in his suborder of Oculinacea, both of us overlooking the fact that Gray had long before already established it. The association of the Stvlasteridae with the Oculinidge does not appear natural, and a closer examination of the structure seems to warrant their removal from that vicinity. The crenenchyma is found, on close examination, to be abundantly perforated ; this can be seen more distinctly in Allopora miniata (and best in dead and bleached speci- mens) than any other form that I have examined, but after proper preparation there is no difficulty in recognizing that structure in all the genera. The best way is to prepare a section and brush it over with ink or some other colored liquid, which will fill the fine canals and show them to pervade the whole ca^nenchyma, anastomos- ing among themselves and communicating with the surface and with the interior of the calicles. According to this character we should be warranted in placing the Stylasterida3 among the Perforata, were we to follow M.-Edwards and Haime's system implicitly ; but this section contains so heteroge- neous an assemblage of families, that it will most certainly have to be dismembered. In another place in this paper reasons will be given for separating the Eupsannnida3 from the Madreporidie, with which they have been closely connected. It would l)e perhaps imprudent, without further research, to associate the St>lasteridte Avith the STYLASTER1DJ2. 33 Eupsammida3 ; but besides the porous casnenchyma, there is another character connecting the two famihes, though in a somewhat remote degree : it is the tendency of the septa to unite by their inner edges and enclose, in the interseptal chamber thus formed, the septa of a higher order. This character, more or less developed in the Eupsam- midse, is carried to excess in the Stylasteridae, where sometimes the interseptal chambers are so far removed from the central part of the calicle as to lose all apparent connection with it. The subfamily of Turbinaringe M.-Edw. & H. has also consider- able affinities with the Stylasteridae, having like them all the septa alike. In StylasteridiB the calicle is invaded by the casnenchj^ma, and in a great measure obliterated as in the Oculinidas proper. This char- acter was relied upon, together with the fancied solid caenenchyma, for associating these forms ; in reality the filling up of the Stylasteridae is of a different kind ; the interseptal chambers, always narrow, are continued down and merge in the fine canals of the ctenenchyma, whilst the central fossa is obliterated by the growth of the columella. True dissepiments do not appear to exist. I have, unfortunately, never succeeded in oljtaining a satisfactory view of the living polyp of any of the corals of this family, nor am I aware that it has been observed by others. When it is, our views of the classification may be much modified. The so-called ampullar, peculiar to this fimily, were found in the fresh polyp to be filled with a yellow mass resembling the yolk of an egg. The genera forming this family are the following : Allopora, Sty- laster, Distichopora, Cryptohelia (= Endohelia ? ) Lepidopora, n. g., and Errina. Professor Verrill has also included Axohelia, which I consider identical with Madracis, and which belono-s to an entirely different family. Of these genera, Allopora and Stylaster are very closely allied, differing only by their mode of growth. Stenohelia Kent, a new genus established to receive Allopora madeirensis Johnston, and Stylaster complanatus Pourt., I believe very difficult to maintain, as the pas- sage from the Stylasters with circular calicle to those with an oval one is very gradual, as is also the other character derived from the obliteration of one of the lips of the calicle by the intrusion of the CJBuenchyma. Cyclopora Verrill has all the fundamental characters of a true Stylaster (the columella is well developed) ; it leads the way towards Distichopora by its interseptal chambers becoming more separated from the fossa, and the latter assuming a more elongated form ; still, the differences from Stylaster proper are scarcely sufficient to warrant a new generic name. Lepidopora, a new genus, forms the passage from Stylaster through Cryptohelia to Errina, which is the 5 34 STYLASTER. most aberrant genus of the family. The genus Endoheha is the only one of which I have not seen a representative. From the description it seems to be scarcely different from Cryj)tolielia, except in the extent of the lii^-like appendage. To recapitulate, the characters of the family can be expressed as follows : — Corallum generally branching, frequently colored, multiplying by gemmation. Calicles small and deep, septa all equal, united by their inner edges, and sometimes containing, in the closed interseptal chamber, rudimentary septa of a higher order; in some genera the septa are merged in the casnenchyma and the interseptal chambers are at some distance from the calicle. Columella styliform or globular, spongy, and hirsute. Ccenenchyma abundant, finely porous, and per- meated by fine canals. \ calicles regular . . . Sttjlaster. ( jjeniDiation alternate i i- i , i • ^ /-- , 7 ;• ^ -.v V i- i * ) f calicles labiate . . . LniptokeUa. (with distinct septan •^' I ( gemmation irregular Allopora. Stylasteridae^ f calicles confluent on I septa obliterated, interseptal chambers | e>lge of corallum . . Dislichopora. L distant from fossa. . i ^.^,,^j^^ j^^,j.^^^ _ _ _ LepKlopora. 1^ calicles witliout I'p . Errina. STYLASTER Gray. Stylaster erubescens Pourt, Stijiaster ertihescens Pourt. Bull. Mus. Comp. Zool., No. 7. Plate IV., ficrs. 10 and 11. Branching densely, flabellate, not coalescing, rising from a broad encrusting expansion ; younger branchlets slender, with rather close- set alternate calicles ; older branches much thickened, with calicles in irregular rows on one surface, interspersed with ampullae. Caenenchyma smooth. Calicles slightly prominent, 1.2 to 1.5 mm. in diameter, deep. Septa nine to twelve, most frequently eleven, equal, shaped like folds of the wall, joined with each other at a little distance below the edge of the calicle, thus forming pit-like interseptal chambers. Tertiarj^ (?) rudimentary septa in the shape of a hairy fringe, enclosed in these interseptal chambers. Columella deeply sunk, rounded, and hirsute. Color white, with a delicate pink blush when fresh. Dimensions : height and breadth of flabellum 10 cm., diameter of largest stems 1.5 cm. Rather common between 120 and 324 fiithoms off the Florida Reef, on rocky bottom. STYLASTER. 35 Stylaster filogranus Potjrt. Plate v., figs. 13 and 14. Corallum a dense and elegant flabellum, the branches settmg off at an angle of 35° to 40°. Sometmies a few aberrant branchlets out of the plane, ver}^ rarely coalescing ; the branchlets generally bending slightly out of each other's way. Main branches consider- ably thickened. Calicles comparatively large, regularly alternate on the terminal branchlets, about one diameter apart ; this position is preserved on the larger branches, but on the main stem the calicles become obsolete. Calicles transversely elongated, with prominent outer lip ; the inner one almost merged in the ca^nenchyma. Septa generally twelve, prolonged rather deeply in the fossa ; tertiary rudimentary septa scarcely visible. Columella styliform. hirsute, too deeply seated to be seen from outside. Color light pink, fading into white in the younger branchlets. The color diifused through the entire thickness. Spread and height of corallum, 10 to 15 cm.; thickest branches, 5 to 10 mm. in diameter ; calicles, 0.8 to 1 mm. West of Tortugas ; ? flxthoms (lal^el lost). Stylaster Duchassaingii Pourt. Stylaster elegans DuCH. & Mich. Plate VI., figs. 1 and 2. As stated by me in Bull. Mus. Comp. ZooL, No. 6, the name SL elegans was preoccupied by Yerrill (January, 1864) when Duchassaing and Michelotti published their Siqypltment cm Mtinoire, &c. (May, 1864). A Stylaster found near Tortugas in 43 fathoms agrees pretty well with those authors' description and figure, though the latter is not detailed enough for safe identification. The calicle in this species is quite small and is invaded in its upper half b}^ the ca?nencliyraa, thus forming a transition to the forms for which Mr. Kent has established his genus Stenohelia. The color of my specimens was white or pale pink, with bright pink ampullae. 36 STYLASTER. Stylaster punctatus Pourt. Corallum subflabellate, slender; ccenencliyma smooth, finely punctate with little pores, looking under the lens like eggshell. These pores are present in all the family, as stated before, but they are more conspicu- ous in this species than in the others of the genus. Calicles alternate, subpedicellate, but very little larger than the pedicel, always confined to two sides of the branches, upper lip not obliterated, except in the axillae of the branchlets, where the outer lip also becomes appressed and scalelike. Fossa deep and narrow, columella deeply seated. Septa twelve, as usual in the genus. Color light purple. It differs from Si. I'oseus, which it resembles most, by its slenderer branches, smaller and less prominent calicles, more distant from each other, and strictly lateral ; deeper fossa, smaller and more concealed columella. One specimen in 9 fathoms off Orange Key, and another in 315 fathoms off Double-Headed Shot Key, a very unusual range. I sup- pose the shoaler locality to be the true habitat, as the cast in 315 fathoms produced a great mixture of shoal and deep-water species. The locality is at the foot of a steep slope washed by a strong current. I have also seen it in Mr. R. Arango's collection, in Havana, from an unknown depth. Stylaster complanatus Pourt. Stylaster complanatus Pouut. Bull. Mus. Coinp. Zool., No. 6. Stenohella conipkinata Kknt. Ann. & Mag. Nat. Hist, February, 1870. Plate II., figs. 16 and 17. Corallum branching, flabellate, not coalescing, slender. Calicles pedicellate, gemmating from the edge of the preceding calicle, gener- ally on alternate sides, so as to give a zigzag form t^ the branches, but sometimes two or even three corallites rise from the border of the preceding. They are directed towards one of the faces of the flabel- lum. Calicles compressed in the same plane, their diameters being 1 mm. and .7 mm.; on the larger branches the calicle becomes hidden by the plicated lip raised against the stem. Fossa moderately deep, columella styliform, and surrounded by rudimentary pali. Spiny ampulla?, chiefly confined to the rear side. Septa twelve, rather exsert, and forming a plicated or scalloped edge to the calicle. Tertiary rudimentary septa in the interseptal chambers. Color white ; specimens in Mr. R. Arango's collection, in Havana, delicate pink. In 270 fathoms off Havana ; not found on Florida coast. (For other species of this genus, see Appendix.) ALLOPORA. 37 CRYPTOHELTA M.-Eday. & H. Cryptohelia Peircei Podrt. Cryptohelia Peircii Poukt. Bull. Mu<. Comp. Zool., No. 0. Plate IL, figs. 18 and 19. Corallum arborescent and subflabellate, irregularly dichotomons, slender, finely striated, with minute pores in the striae. Calicles sub- pedicellate, of a larger diameter than the stem, all facing to one side. Septa twelve to sixteen, not extending far into the calicle. Columella not visible. The lower border of the disklike expansion of the calicle prolonged into a lip folded over so as to hide the fossa. Some of the calicles inflated and globular, perhaps from the presence of para- sites. Dimensions : height 2 cm. (broken) ; diameter of stem 2.5 mm., of branchlets 1 mm., of calicles 2 mm. Off Havana, in 270 fathoms. Off Sombrero Lighthouse, in 262 fathoms (one specimen, dead.) OfTBahia Honda, in 324 fathoms (livinjr). In 600 fathoms, lat. 31° 32' N., long. 78° 20' W., fragments brought up by the lead. ALLOPOEA EiiRBG. Allopora miniata Polrt. Allopora miniata Poukt. Bull. Mus. Comp. Zoiil., No. 7. Plate III, figs. 14, 15, and 16. Corallum branching, flabellate, the main trunk rather massive and flattened. Surface finely and sharply granular, porous. Branchlets thick and obtuse. Calicles irregularly but densely distributed on one surfiice of the branches, obsolete on the main trunk. Small ampulla? aljundant between the calicles on the younger branches. Calicles slightly prominent, fossa deep, columella large, spherical, hirsute, deep- ly immersed. Septa from seven to ten, generally eight. Interseptal chambers comparatively large, with very distinct hirsute tertiary septa enclosed, giving to the whole when magnified the appearance of a 38 DISTICHOPORA. calicle surromiclecl by a number of smaller ones, all provided with col- umella. The edge of the calicle and of the septa is crowded with small sharp points and abundantly perforated. Color brick-red. The corallum seems to have grown in a horizontal trailing manner, as the lower surface often shows signs of contact with foreign bodies. Dhnensions : length 10 to 15 cm., thickness 1 to 1.5 cm., diameter of calicles 1.2 to 1.5 mm. OfT' Teiniossoe Reef, in 124 fatlioms. Oil' Codin's Patches, in 195 fathoms. This species is the most massive of our deep-sea corals, the next to it in that respect being perhaps Sttjlmter eruhcscens. DISTICHOPORA Lamk. Distichopora sulcata Pourt. Disdchopora sulcata Pourt. Bull. Mus. Comp. Zobl., No. 6. Plate IV., fig. 14 ; and Plate VII., fig. 7. Corallum dendroid, much compressed, somewhat rugose. Calicles sometimes isolated, but generally confluent and forming a deep farrow. Interseptal pores round, from eight to ten, surrounding the isolated calicles. Columella deep seated, hirsute, styliform. Color whitish. Off Havana, in 270 Hithoms. In Mr. Arango's cabinet, in Havana, from the coast of Cuba, depth unknown. Fragments brought up by the lead from GOO fathoms, lat. 31° 32' N., long 78° 20' \V. Distichopora foliacea Pourt. Distichopora foUacca Pourt. Bull. Mus. Comp. Zool., No. 7. Plate IV., figs. 12 and 13. Corallum branching, flabellate, much compressed, finely striated and granulated. The calicles in a somewhat irregular row, a little on one side of the edge, which is sharp and finely serrated. One of the rows of lateral pores on the summits of the denticulations, the other not well defined, represented by scattered tuberculated pores. The plane towards which the calicles are situated is thickly studded with am- LEPIDOPORA. 39 pnllee, each of which has a small lateral opening. They are less numerous on the other surface. Columella deep seated, long, hirsute, in the shape of a pointed club. Color orange-pink. Height 4 to 5 cm., thickness 2 to 3 mm. This species differs from the preceding by its smaller calicles, not placed in a furrow, irregular lateral pores, and serrated edge. Off Key AVest, in 100, 135, and 154 fathoms. Oft' the Samboes, in 116, 135, and 147 fathoms. Off" Tennessee Reef, in 124 fathoms. Oft" Sombrero Light, in 152, 183, and 262 fathoms. Oft" Bahia Honda, in 176 fathoms. Oft" Coffin's Patches, in 195 fathoms. * ERRINA Geay. Errina carinata. Heh'opora carinata Pourt. Bull. Mus. Comp. Zool., No. 6. Plioboihnis carlnatus Pourt. Bull. Mus. Comp. Zool., No. 7. Plate VI., fig. 5. Corallum branching, very slender. Calicles confined to one side of branches, round, without lip or trace of septa. Perforated tubercles prominent, continued downward in sharp ridges. Columella slender, elongated, hirsute, too deeply seated to be seen without a section. Having but a very small specimen of this coral, the determination of it remained uncertain until a part was sacrificed for a section, when a view of the columella settled the question. The branchlets are only a millimeter and less in diameter. Off" Havana, in 270 fathoms. * There is in the Museum of Comparative Zoology a specimen of a third species of Distichopora of West Indian origin. It was received from St. Thomas, through Mr. Duchassaing, and appears to be undescribed. It has nearly CNlindrical branches, ft ittened more or less near the ends. Calicles in a very flat furrow, rather crowded. Lateral pores more developed in one of the series than in the other, transversely elongated, on top, or down the side of small tubercles, thus forming an approach to the structure of Errina. Smaller branches studded with ampulla;, scatteied om s on the main stem. Color white. I would propose for it the name o? Distichopora cervina. 40 LEPIDOPORA. I LEPIDOPORA PouRT. Corallum finely branching ; interseptal cliambers separate from cali- cles, in the shape of sUt tubercles ; lip of calicle projecting as a scale over the calicle. This genus combines the characters of Errina and Cryptohelia, having the raised lip of the latter and the separate interseptal cham- bers of the former. Lepidopora glabra Pourt. Errina glabra Pourt. Bull. Mus. Comp. Zool., No. 6. Plate VII., figs. 8 and 9. Corallum flabellate, not coalescing ; the older parts of the stem much thickened, faintly striated and granulated ; a row of very small perforated tubercles on each side of the branches. Calicles very small, circular; when terminal, surrounded by a few interseptal pores, which become separated from the calicles in the subsequent growth and form the lateral perforated tubercles. The calicles on one face of the branchlets hid by a scale-like projection ; calicles obsolete on larger branches. Columella styliform, hirsute, in a deep fossa. Ampulla? scarce and small, smooth. Color white. Height 5 to 7 cm., diameter of branchlets .7 to 1 mm., of main branches about 3 mm. Off Havana, in 270 fathoms. Lepidopora cochleata Polrt. Errina cochleata PoUHT. Biill. Mus. Comp. Zocil., No, G. Plate III, figs. 17, 18, and 19. Corallum branching, flabellate, very slender, finely granulated and striated, studded with echinulate ampullae. Calicles formed as in pre- ceding species, but distinctly bilabiate when terminal, and covered by a large spoon-shaped lip when lateral. Perforated tubercles more distinct than in L. glctbra!'^ Height 2 cm., diameter of branchlets ,.3 to .4 mm. Off Havana, in 270 fathoms; rarer than preceding. * The Museum of Comparative Zoolojry possesses remarkably fine specimens of a third species of Lepidopora, sent from Fayal by Miss Dabney. From wane of specimens of compirison, and rflyinfr on the very brief definition of Errina axpera Gray, thpy were labelled with that name, and cata- BALANOPIIYLLIA. 41 Family EUPSAMMIDiE M.-Edw. & H. EupsammincB M.-Edw. & H. Hist. Nat. des Coral. This group, a subftxmily of the Madreporiclce, according to Mihie- Edwards and Haime, ought to be entirely separated from that associa- tion and constituted as a family by itself, as the same authors had done in their earlier publications in the " Annales des Sciences Naturelles " and in the " British Fossil Corals." It is one of the most natural fam- ilies of the corals, and its affinities are much closer to the Turbinolid^ (the polyps being very similar) than to the Madreporidse, with which it is only allied by its perforated walls. BALANOPHYLLIA Searles Wood. Balanophyllia floridana Pourt. Balanophyllia Jloridana Pourt. Bull. Mus. Comp. Zool., No. 7. Plate IV., figs. 5 and 6. Corallum elongated, conical, straight, attached in younger stages and often through life ; but frequently also becoming free, by covering up the small object (shell or fragment of shell), after the manner of Hete- ropsammia, or by becoming accidentally detached from it. No epitheca, wall porous, costate to the foot. Calicle rather deep, elliptical. Septa entire, slightly exsert, finelj^ granulated, in six unequal systems of four cycles, with rudiments of the fifth in some of the systems. Columella flat and papillose. Polyp generally brick-red. Height 2 to 3 cm. Diameters of calicle 10 and 15 mm. Off Sand Key, in 26 fathoms. Off the Qu't'ksands, in 34 fathom?. Off Conch Reef, in 39 fathoms. Off French Reef, in 45 fathoms. Off Carysfort Reef, in 48 fathoms. Off Pacific Reef, in 49 fathoms. Off Carysfort Reef, in 63 fathoms. logued in the same way in Bull. Mus. Comp. Zool., No. 3. Having since seen the original specimens of Eirina aspera in the British Museum, 1 have been enabled to recognize the mistake. The Fayal specimens form large (labelliform expansions, of great elegance, resembling Sti/laster Jlnhellifnrmh. The species is very nearly allied to Lepldohelia glabra, from which it differs by its more crowded branchlets, larger and more numerous tubercles, not confined to the sides of the branches, and rough CEenenchyma. I would propose the name Lepidopora Dabneyi for this species, which I cannot find to be described. (See Plate VII., figs. 10 and 11.) The specimens are associated with Caryo- phyllia cyathus, growing from the thicker branches. The spread of the corallum is nearly a foot, NO. IV. 6 42 BALANOPIIYLLTA. The 3^oungest individuals observed have the shape of a truncated cone attached by the base. The wall is quite smooth, imperforate, and the se2:)ta, twelve in number, equal and not quite extending to the centre, where the rudiments oi the columella are already visible. At this stage none of the flimily characters are displayed. The dimensions are then about 2 mm. in diameter at the base, 1.3 at the calicle, and .8 in height. The next step of develojDment is the formation of costee on the upward prolongation of the wall. They first appear in the shape of sharp jooints grouj^ed about tlie origin of the septa. About the same time an opening appears on the border and rather outside of the calicle, opposite each of the secondary septa, which gradually widens inwards, dividing apparently the sejDtum in two. The two bor- ders of that opening become the tertiary septa, the secondary septum is gradually pushed inwards, and is replaced by a new one growing out on the same radius from the wall, and but loosely connected with the joined tertiaries and original secondary. In a vertical section of an adult specimen a row of perforations is seen to extend both on the old and the new part of the secondary septum at the lines of junction with the tertiaries. The interior part of the tertiary septum is now to all intents and j^urposes a palus, and the wdiole arrangement quite similar to that of Deltocyathus, for instance. At this stage the septa are rather exsert and jagged on the edge, and seen in profile the resem- blance with Stephanophyllia is quite striking, not taking the attached condition into consideration. As the growth proceeds, the point of junction of tertiaries and sec- ondaries moves farther into the calicle, until it rea(;hes the columella. At this period the older or internal part of the secondary septum has nearly or entirely disappeared, and the same process of growth goes on with the septa of the fourth cycle, which become joined to those of the third. One or two specimens show a coalescence of two individuals grow- ing in a crowded position. Although this species is generally destitute of epitheca, I hesitate in separating from it a form of which a number of specimens were ob- tained in 68 fathoms oft" the Tortugas. These were almost cylindrical, covered nearly to the calicle by a thin wrinkled epitheca, through which the costce can be distinguished; the calicle contracted. All except a few young ones were dead, and have all the appearance of having l:>een killed by being gradually covered with mud, as the calicle has mostly grown smaller and become stunted until definitively smothered. Hav- ing seen none of this form of a healtliy growth, I am inclined to believe that they belong to BalanopliijUia floridana^ modified by external circumstances. THECOPSAMMIA. 43 THECOPSAMMIA Pourt. Corallum simple, attached, without costae, or only rudimentary ones on the peduncle, covered with a complete epitheca. This genus is intermediate between Balanophyllia and Heteropsammia ; like the latter, it is destitute of distinct costiB, but has an epitheca like some of the Balanophyllige, more developed indeed than in any other repre- sentative of the famil3\ Thecopsammia tintinnabulum Pourt. Tliecopsammia tintinndbulum Pourt. Bull. Mus. Comp. Zool., No. 7. Plate I., figs. 9 and 10. Corallum subcylindrical, or almost hemispherical, with turl^inate base and small, abruptly constricted peduncle. Wall thick, very porous and vermiculated. Epitheca well developed, seldom rising quite to the border of the calicle, its tissue penetrating the mural pores and solidifying the wall. Calicle slightly elliptical, moderately deep. Septa in six unequal systems and four cycles, entire, thin, not exsert, with fine granulations ; those of the fourth and fifth order scarcely bent towards those of the third, and not connected with the latter and with each other. The septa of the first and second order connected with the columella. The two opposite systems on the longer side of the calicle always incomplete in one of their halves, and one or two of the other systems also incomplete in the same manner. The columella papillose and porous, sometimes sublamellose, and form- ing three distinct masses in the adult, the middle one being largest. Height 1.5 to 2 cm. Diameters 1.2 and 1.4 cm.; the average size is rather less. The living polyp is of a handsome pinkish-orange color ; the mouth oblong, and surrounded by rather numerous conical tentacles. Off Key West, in 120 fathoms. Off Key "West, in 1 25 fathoms. Off Key West, in 135 fathoms. Off Key West, in 138 fathoms. Off Marquesas, in 140 fathoms. Off Key West, in 154 fathoms. 44 THECOPSAMMTA. Thecopsammia socialis Pourt. Thecopsammia socialix Pourt. Bull. IVIus Com p. Zool., No. 7. Plate IT., figs. and 10. Corallum turbinate, rather long conical, with a thick, not much constricted peduncle. Wall and epitheca as in preceding species, but obscure traces of costa3 can sometimes be seen through the epitheca near the peduncle. Calicle elliptical, moderately deep. Septa entire, smooth, crowded, not exsert, thick near the wall. Six unequal systems and five cycles. The convergence of the septa of the fourth cycle to meet in front of the tertiaries generally plainly visible from outside. The septa of the fifth cycle are developed in old specimens only, and then in few of the systems ; but they soon become very large and encircle the two preceding cycles. The tertiaries re- main smallest of all. The columella is papillose and porous, but more compact than in the preceding species, and nearly always forms three distinct masses, of which the middle one is largest. This species often forms clusters, several individuals being united by their bases. Height about 2 cm., diameter of calicle about 13 mm. Off Coffin's Patches, in 195 fathoms. Off Sombrero Lighthouse, in 2G2 fathoms. This species, according to Dr. Duncan, has also been found by the "Porcupine " Expedition, in 345 and 363 fathoms, in the seas north of Scotland. The specimens were more developed than those from Florida, and rather variable, since Dr. Duncan distinguished three varieties. There is a considerable range of variation also in the Florida specimens. I have found in some cases, since the first descrip- tion was written, the traces of costa3 which had been noticed in some of the European specimens. The epitheca is particularly subject to variation, sometimes extendino; to the ed^e of the calicle, sometimes confined to the lowest parts of the peduncle. A specimen of the latter category shows distinct granular costa3 in the bare part of the peduncle. DENDROPHYLLIA. 45 DENDROPHYLLTA Blainv. Dendrophyllia cornucopia Pouut. Plate v., figs. 7 and 8. Corallum elongated, conical, generally curved, budding irregularly at various heights and on all sides. The younger corallites, remaining much smaller than the parent stock, implanted at right angles, after- wards curved upwards, but irregularly. Costoe distinct, but not promi- nent, finely porous. Sometimes a very rudimentary epitheca. Calicle elliptical, fi)ssa rather deep. Septa thin, finely granulated, serrated in- side of calicle, entire above, in six systems and four c^^cles in all the sys- tems, the fifth in some of the quarter-systems adjoining the primaries. Columella large, densely spongy, elongated, and slightly convex. Height of largest specimen 10 cm. Greater diameter of calicle 18 mm., lesser 16 mm. The younger ones budding out are not more than 6 mm. in diameter in any of the specimens. A specimen has thirty-two buds, the largest has eighteen, smaller ones have none. The polj^p is of a deep maroon color, darkest on the disk, lighter on the wall. Off Key West, in 120 and 125 fathoms. A specimen of Dendrophyllia, rather incomplete, in the collection of Mr. R. Arango, in Havana, appears to be D. cornigera of which the Museum of Comparative Zoology possesses numerous specimens from Fayal. It differs from our species by the large size of the buds, Avhicli rapidly equal the parent stock. They are also usually confined to two sides of the main stock. Dendrophyllia cyathoides Pouut. Plate I., figs. 8 and 9. Corallum broadly attached, elongated, branching irregularly and apparently only a few times, generally widened at the calicle. CostoB distinct to the base, in j)f^i't veiled by a rudimentary epitheca, but prominent and serrate near the calicle, and continuous with the septa, which are thin, granulate, rounded, finely serrate, more exsert than in any other species of the genus, in six systems and four cycles, mostly 46 FUNGIA. quite regular. Columella elongated, convex, composed of transverse or irregular twisted lamellae. Height 2 to 3 cm. Diameter of calicle about 1 cm. Oflf Havana, in 270 fathoms. Family FUNGID^ Dana. Tliis family has few representatives in the Atlantic basin, and most of them belong to the subfamily of Lophoserinae M.-Edw. & H. In the littoral and reef zone all the species are compound ; from deep water the dredge has brought up three small, simple sj)ecies, the first simple Fungidie found in these seas. Fungia symmetrica Pourt. Plate VII., figs. 5 and 6, Corallum circular, plano-convex. Wall perfectly plane, with very small convex umbilicus in the middle, very little j^erforated. CostiB distinct to the centre ; corresponding to the septa, finely spinous and granulated, particularly towards the border, subequal, primaries and secondaries slightly larger. Septa subequal, spinous, larger ones some- what lobed, in six regular systems and four complete cycles. The septa of the fourth cycle connected by their inner edge with those of the. third, and the latter with those of the second ; the points of connec- tion sometimes expanded into a kind of membrane. The primary septa reach the centre without lateral connection. Columella rudimentarv, sometimes covered with a membranous expansion, through wdiich some of the spines project. The synapticules are large, rather deeply seated, and correspond to each other in the contiguous chambers, so as to form from four to six more or less reg-ular concentric circles. Diameter of larger specimen 1 cm., of smaller 7 mm. Off Cojima, Cuba, in 450 fathoms. Off Carysfort Reef, in 350 fathoms. This coral resembles Cijcloseris hexagonaUs very much, but the per- forations of the wall and the echinulated costaB separate it apparently from that genus. I am, however, inclined to doubt the generic value of those characters, as the younger of the two specimens shows them in so small a degree that they would generally escape notice. If found in a sea where larger Fungise were common, these specimens would naturally be considered as the young of one of them. DIASERIS. 47 Diaseris pusilla Pourt. Diaseris pusilla Pouut. Bull. Mus. Comp. Zool., No. 7. Plate II., figs. 6, 7, and 8. Corallum subelliptical, very fragile. Wall flat or slightly concave, imperforate, very thin, finely costate. Costas thin, alternately large and small, finely dentate. The base shows the traces of lobes joined to- gether, often very imperfectly. Septa strongly dentate, laciniate, and perforate, marked with strong ridges and furrows, connected with each other by synapticula near the base (see Plate II., fig. 7). Five cycles of septa and six unequal systems, one or two being generally incomplete. The primary septa more lobed and much higher than the others ; those of the lower cycles tending to unite with those of the cycles preceding them. Fossa oblong, well marked. Columella rudi- mentary in the shape of a narrow ridge. Color of the polyp dark brown, tentacles simple, conical, rather long. Diameter of largest specimen 16 mm. Off Sand Key, in 119 fathoms. Off Sand Key, in 120 fathoms. Off Sand Key, in 135 fathoms. Off Alligator Reef, in 14 7 fathoms. Off Tennessee Reef, in 174 fatlioms. Off Alligator Reef, in 15(j fathoms. • . Off Alligator Reef, in 189 fathoms. Diaseris crispa Pourt. Plate v., figs. 1 and 2. Corallum irregularly circular, formed of loosely joined lobes. Wall very thin, horizontal, indistinctly costate, but deeply furrowed in the junction of the lobes. A few minut scatteered spines and gran- ules, particularly at the edge of the lobes. Septa subequal, not lacini- ate, serrate ; the teeth equal and flattened transversely to the septa, thus giving to the surfjice a densely crispate appearance. Systems unequal on account of the irregular mode of growth, four cycles and seldom rudiments of a fifth. Primary and secondary septa equal, those of the two next cycles slightly smaller. All the septa but those of the fourth cycle reach the centre, the latter connect with the tertiaries about half-way. Fossa moderately concave, no columella, the wall in the centre sometimes eroded. 48 MYCEDIUM. Dimensions slightly smaller than those of Diaseris pusilla. Both species are generally found together in some ahunclance, but perfect specimens are rare, on account of their great brittleness. Many speci- mens have evidently been broken and mended as the one figured in Plate Y., fig. 2. Such deformities in other groups of corals have been described as genera, as, for instance, Ecmesus by Phihppi, and Ilemicy- athus by Seguenza. MYCEDIUM Oken. Mycedium fragile Dana. This S23ecies was obtained in considerable quantities in the two localities named below. The greater number of the specimens were dead, but an interesting series of young ones of all ages were obtained (see Bulletin Mus. Comp. Zool., No. 13, by Professor Agassiz). It is not uncommon in less depth, though not often found on the reef. In looking over the materials in the Museum, for the determination of this coral, it has become very apparent that a revision of the group is very urgent, and that probably a large number of sjDCcies, and even some genera, are based on very insufficient characters and ought to be united. In fact, several of the specimens obtained by me in the same dredgeful. if seen separately from different localities, would unhesitat- ingly be pronounced different species ; whilst being put on guard by the association, a closer examination shows that the characters vary considerably, even in different parts of the same corallum. Such are the costal striie, which may be subequal, or alternately large and small ; the calicles, which may be flat or raised, isolated or in rows, columella pres- ent or absent, etc. In this connection it may be remarked that Mycedium fragile, which Milne-Edwards and Haime think may be closely allied to Agaricia iin- daia, has really no resemblance to that species at all. The latter is found on the coast of Cuba (cabinet of Mr. R. Arango, in Havana), but not in Florida, as far as I am aware. OffTortugap, in 3G and 43 fathoms. RUGOSA. 49 Order RUGOSA M.-Edw. & H. The section of Rugosa was established bj Mihie-Eclwards and Haime for a hxrge number of fossil corals, all of an older age than the carbonif- erous formation, and particularly abundant in the palaeozoic ages. Their chief characteristic, according to those authors, is the development of the septa proceeding from four primary ones, whilst in all of the living corals the primar}^ number is six. In addition, the chambers are closed inferiorlj by an abundant endotheca, often assuming the shape of floors ; but this character is not found in all the fomilies of the section, though certainly in the greater number. Mr. R. Ludwig has shown (H. v. Meyer's Pakeoidoc/raphica, Vols. X. and XIV.) that the tetrameral arrangement claimed for the Rugosa is only apparent, there being origi- nally six primary septa ; but that the further development in each system is asymmetrical, and that two of the systems remain generally undeveloped. I had, before having knowledge of Ludwig's researches, come substan- tially to the same conclusions b}' the examination of LopJioph/Uum pw- Uferum Edw. & H.,from the carboniferous formation, a form very suitable for that study. As I find my mode of explaining the development slightly different from Ludwig's, it may not be too much out of place to mention it here in a few words. When the youngest stage of the coral is examined by cutting through the tip of the conical Lop)hoph>/Uinii pniliferum, six primary septa and six interseptal chambers are found, pLaced symmetrically on two sides of a vertical plane, and unequally developed. Tliis piano would pass through .wo of the primary septa, of which one may be called anterior and the other posterior, for convenience merely. The two lateral systems (which would be divided by a plane at right angles to the antero-posterior plane) are at the earliest stages larger than the others. In progress of growth the, anterior primary system is checked in its development, and the two adjacent chambers partly coalesce to form the so-called septal fossa. The posterior primary septum is fully developed, and joins the columella (when there is one). The systems between this and the two adjacent primaries do not Ijecome subdivided again, so that these three septa remain close together ; according to Ludwig, they subdivide to a small extent in some genera. Thus, of the six primary interseptal chambers, the two posterior ones remain undivided. The four other ones are each divided by a secondary septum joining the primary ; of the two half-systems thus formed in each, the anterior one is again divided by a tertiary septum joining the secondary ; next the anterior quarter-systems are divided, and so on. The four lateral NO. IV. 7 50 RUGOSA. primary septa thus assume gradually the shape of unsymmetrically dichotomous branches.* The following notation may render the idea more plainly. The primary septa can be represented thus : — Si Si Si Si Si Si The upper and lower being in the longer axis, the lower S being the one in the septal fossa. The next step of development will be : — Si Si Si Si And further : — S, Si ^^ Si S2 §2 S3 S3 Si Si S2 S2 S3 _ S3 01 Still continuing with Milne-Edwards and Haime's system of notation, we would have, after a few more steps : — Si Si Si S3 S4 <^2 "^2 S3 S4 Sg Se Sio Sio Si Si S2 S2 S3 S3 S4 s« . __ S4 ^10 s^ Sio * A somewhat similar unsymmetrical development of the systems has been observed in some of the Medusa; ; for instance, Willla, Prohoscidactijla. See Agassiz, Contr. Nat. Hist. U. S., Vol. IV. ; and A. Agassiz, 111. Cat. Mus. Comp. Zobl., No. II. HAPLOPHYLLIA. 51 For the sake of symmetry, the number of septa has been made equal in all the systems in these diagrams, which is seldom the case. Sometimes a small rudimentary septum is seen in each half of the septal fossa on the sides of S- Numerous illustrative figures will be found in the plates accompany- ing Ludwig's paper.* Mr. A. Kunth ( Zcitschr. dcr Deidsch. Geol. Ges., XXI., Heft 3 ) has also examined the law of growth of the Rugosa, chiefly by the considera- tion of the successive development of the costae. He still adheres to the tetrameral primary division, from want of having examined indi- viduals of very young age, and hence great simplicity. The papers of Romer and of Lindstrom I have had no opportunity to examine. In the Bulletin of the Museum of Comparative Zoology, No. 7, I referred to the Rugosa, a new genus and species, though with some doubt. I hoped to be able to procure more numerous specimens, par- ticularly as the sing-le one on which the determination rested was not quite normally developed. In this I have been disappointed, so that the position of the genus in the system must still remain provisional. The chief ground for placing it among the Rugosa is its general appearance, its smooth ungranulated septa, not resolvable into the regular hexameral arrangement, but on the other hand difficult also to refer to the modified hexameral system just exposed. HAPLOPHYLLIA Pourt. HaplopliijUia Pouut. Bull i\Ius. Comp. Zobl., No. 7. Corallum simple, fixed by a broad base, covered with a thick epith- eca ; columella styliform, strong, very thick at the base. Interseptal chambers deep, uninterrupted by tabula) or dissepiments, but filling up solid at the bottom. The nearest known form to Haplophyllia which I have been able to find is Calophjllum prof nudum Germ. (Geinitz) [Astroci/athus Gemitzii Ludw.),t which, however, lacks a columella. It is fossil in the Dyas. * There are several points of resemblance in the mode of division of the calicle in the Pocillo- porae and in the Rugosa, whicli would warrant a closer examination ; such are the bilateral sym- metry combined with a more or less m irked antero-posterior asymmetry, the shallow calicles, the tabula;, etc. The order of succession of the septa in the Pocillopora; is not known. f Ludwig has formed new generic names for nearly ail the fossil corals, to adapt them to his classification; such a proceeding can only introduce confusion, particuhirly as bis arrangement is quite artificial. A table of double entry is no doubt very convenient, but unfortunately the organ- ized beings were not created on such a plan. 52 ANTIPATHARIA. Haplophyllia paradoxa Pourt. Plate ir., figs. 11, 12, and 13. Corallum siibcvlmclrical, short, fixed by a broad base ; epitheca thick, wrinkled, reaching above the border of the caUcle, and forinmg around the latter several concentric circles, as if formed of several layers. Calicle circular, fossa deep. Septa smooth, Avithout granulations or perforations, not reaching the border of the calicle, with smooth en- amelled surface, like the other internal parts of the calicle. Columella formed of two smooth conical processes, very thick at the base and tending to fill up the chambers. Eight septa Larger and connecting with the columella, alternating with smaller ones, which touch the columella at a much lower level. A further cycle is indicated by mere rudiments in some of the chambers. In the specimen there are irreg- ularities in two of the sj^stems or half-systems, one of which is closed by a horizontal plate, probably to exclude a parasitic intruder. Polyp scarlet, greenish in alcohol, with about sixteen rather long tentacles, bluntly tuberculated at the tip. Height 1.5 cm., diameter 1.2 cm. Oif Bahia Honda, Florida, in 324 fathoms. Suborder ANTIPATHARIA M.-Edw. & Haime. This suborder, Avith one single family, the Antipathida?, constitutes a very natural and homogenous group, if we remove from it the genus Gerardia of Lacaze-Duthiers, which has no other relationship with the other genera than the property of secreting a horny polypidom. The polyps of a Gerardia differ in no particular from those of most of the ZoanthidiB in the arrangement, number, or shape of the tentacles ; they even agree with that group in the habit of incrusting the derm with small foreign bodies. Hence it would l)e quite proper to ])lace Gerardia Lamarcki, Lacaze-Duthiers, the only species known, among the Zoanthidje, as type of a su1)family. The subdivision of the Antipathidix) into genera hy Blainville, Gray, and Milne-Edwards is based entirely on the solid parts. The few species of which I have had the opportunit}'^ of examining the polyps, all l)elong to the genus Antipathes proper, according to the charac- teristics heretofore adopted. It has seemed to me, however, that two distinct types of polyps could be distinguished, the one Avell circum- ANTIPATEIES. 53 scribed, flower-shaped, symmetrically radiate, with long tentacles ; the other so elongated longitudinally that the radiate shape is quite indis- tinct, the six tentacles being disposed in pairs at some distance from each other. Unfortunately the number of species examined is too small to allow of basing any attempt at classification on those charac- ters, which furthermore seem to bear no connection with the general shape of the corallum. ANTIPATHES Pallas. Antipathes tetrasticha Pouut. Anlipathes tetrasticha Pouut. Bull. I\Ius. Comp. Zoo!., No. 7. Corallum with simple stem, pinnate ; branchlets alternate and gen- erally double, two branchlets starting from the same point at an acute ano-le, thus forming; four rows, two on each side of the main stem. In some specimens few of the branchlets are double, in others nearly all. The branchlets are moderately beset with short, somewhat blunt, tri- angular spines. No swellings, as in A. JiundUs and A. filix. Polyps small, very much elongated, with the tentacles short and blunt, placed in pairs, so that in a branchlet or pinnule the lower side appears fringed with tentacles placed in pairs, and it requires some attention to distinguish the single polyps by the position of the mouth. They are all of one size on the pinnules, but there is generally a some- what larger polyp on the main stem between the successive pinnules. Total height 7 cm., length of pinnules 3 to 4 cm. Off Sind Key, in 120 and 125 fathoms. Ducliassaing and Michelotti have described very briefly, under the name of A. americana, a pinnate species with simple stem, but the pin- nules are described as dichotomizing rather frequently, which is never the case in ours. • Antipathes dissecta Ducn. & Mich. The very scanty description given of this species in Supplement au Memoire sur les Coralliaires, etc., leaves some doubt as to the correct identification of my specimens, of wdiicli it is therefore not superfluous to give a fuller description. Corallum large, irregularly branching into long, crooked arms, of more or less elliptical section. Ultimate branchlets or pinnules few in number, alternate and very loosely pinnate, without swellings. (Du- 54 ANTIPATHES. cliassaing and Micliellotti had evidently specimens with dried polyps before them, which give to the branchlets the appearance described by them.) Spines confined only to the smaller branches and branchlets, few, small, and rounded, thus having under the microscope a nipple- shape appearance in profile, though flattened longitudinally. Polyp large, resembling in their proportions those of A. arborea, as represented by Dana. They are all on one side of the branchlets (not the upper or lower), and generally the row is composed of alternately large and small polyps. Tlie mouth surmounts a tubercle and is elon- gated transversely to the branch, as has been observed by Lacaze- Duthiers in other species of the genus. The tentacles are elon- gated, conical, and do not appear susceptible of much contraction, but can be curled up over the mouth. They are never found contracted into small knobs, as in some species hereafter described. The largest specimen obtained is one meter high, the main stem being 2 cm. in diameter. Pinnules 3 or 4 cm. long. Of!" Allijiator Reef, in 110 fathoms. OtF Siivi Key, in 125 fathoms. Off Coffin's Patches, in 1D.5 fathoms. Off Bahia Honda, in 324 fathoms. Antipathes filix Pourt. Antipalhesjilix Pourt. Bull. Mus. Comp. Zool., Nj. 6. Main stem simple, pinnate ; the pinnules setting off nearly at right angles, rather short, alternate, covered with spines or short, stiff hairs, and showing a succession of slight swellings and contractions. Soft parts not observed. Height about 7 cm. Every specimen obtained served as support to the tube of an An- nelid. * Off Havana, in 270 fathoms. Antipathes humilis Pourt. Antipathes humilis Pourt. Bull. Mns. Comp. Zool., No. 6. Corallum flabellate, irregularly dichotomous, spreading more laterally than vertically, hirsute, the spines being slender and longer than the diameter of the branchlets, which are alternately swollen and con- * The specimens having been sent to Professor Ehlers with the eollection of Annelids, I have been prevented from comparing the polyps with those of the other species since obtained. ACTINID^. 55 tractecl. Polyps all on the same side of the flabellum, on the swellings of the branchlets, aljout one diameter apart. Tentacles very short, when contracted forming small knobs shorter than the spines, and placed in two rows, the mouth appearing between the middle pair. The spines surrounding the polyp are larger than those in other parts of the branches ; the largest appear inside the polyp, in the spaces between the tentacles. Mouth surmounting a tubercle surrounded by about twelve papillae in a close circle ; a second circle of papilla on the peristome just inside of the tentacles ; similar ones are scattered on the whole surface of the sarcosome. Height 10 to 12 cm., spread 12 to 14 cm. Off Havana, in 270 fathoms. Fragments of another subflabellate Antipathes were brought up from 68 fathoms, west of Tortugas, different from A. JmmiUs. There is not enough of it, however, to give a full description. ""■•^ Antipathes lenta Pourt. Mode of branching unknown. Pinnules very long and slender, like thin horse-hair, not in a regular pinnate arrangement ; spines in num- ber intermediate between A. larix and A. siihpinnata, as figured by Lacaze-Duthiers, but somewhat longer, and straighter than in either. Polyps of the same type than those of A. dissccfa, but very much smaller, and appear more distant from each other ; the longitudinal and transverse diameters are more disproportionate than in the preceding species, and the tentacles show more the tendency of arranging them- selves in two parallel rows. The alternance of large and small polyps, though not very regular, is quite noticeable, and the disproportion of size of the two kinds veiy great. Only a few branchlets of this species were obtained, some of them 10 to 12 centimeters long, without much diminution of diameter. Off Carysfort Reef, in 35 fathoms. Off' Tortugas, in 3 7 fathoms. Family ACTINIDiE M.-Edw. & Haime. This family is represented by five species in our deep waters, namely, two Actiniae, two Palythoa3, and one Ilyanthus. The almost insuperable difficulty of determining satisfactorily animals of this flimily from much contracted alcoholic specimens will prevent anything more than a simple notice, without attempt at naming them. 56 ACALEPILE The larger Actinia, milk-white, with orange disc and tentacles, is found sparingly between 120 and 195 fathoms. The smaller appears to live unattached in sand, in about 90 fathoms : it is of brownish color, and disk-shaped when contracted. A Palythoa not very different from P. occUata Lamx. from the Florida coast is quite common, ranghig from 35 to 287 fathoms, attached in considerable numbers to foreign bodies. A large black Palythoa appears to live solitary in 42 fathoms. One specimen of an Uyanthus was found in 124 fathoms, off Tor- tugas. A disk-shaped Actinian covered with sand was obtained in 85 fathoms off Tennessee Reef, and in 91 and 93 fathoms off Bahia Honda and off the Samboes. It probably belongs to the new genus Ammodiscus Carpenter, dredged off the coast of Portugal. It is not destitute of ten- tacles in my sjDScimens. Order ACALEPIIiE. Suborder HYDROIDE^ Johnston. MILLEPORID^E Agass. MUleporldce M-Edw. & H. Since Professor Ao;assiz's first announcement of the affinities of the Millepores with the hydroid polyps, it does not appear that new obser- vations have been made on the subject, although twenty years have elapsed since. The polyps of Millepora are very difficult to observe, not only on account of tlieir very small size, but also because they re- quire peculiar care in collecting, being killed by the shortest contact with air. If they have finally expanded, after much patient waiting on the part of the observer, the least shock to the vessel, or attempt to bring the specimen in a convenient position, is sufficient to make the polyps contract instantaneously and often finally. I have succeeded but once in having a good view of one of the larger polyps of 3Iiilc- pora cdcicornis, in company with Professor Agassiz. It differed from tlie figure in the " Contributions to the Natural History of the United States," Vol. HI., Plate XV., fig. 6, in being comparatively shorter, and having larger tentacles or rather tentacular masses studded with lasso cells, five in number instead of four. The mouth was not seen very PLIOBOTHRUS. 57 distinctly, but appeared to be a transverse slit in the middle of the disk. It remained expanded but a short time. It is almost unnecessary to add, after these remarks, that the deep-sea Milleporidaj afforded no opportunity to observe their polyps. PLIOBOTHRUS Pourt. Corallum branching ; tissue more compact than in Millejiora ; larger pores scarcer, smooth, without any rudiments of septa ; smaller pores tubulated, ca^nenchyma finely porous, the pores linear and arranged in rows. Pliobothrus symmetricus Pouut. Pliohollirus symmetricus PoURT. Bull. Mus. Conip. Zodl , No. 7. Plate IV., figs. 7 and 8. Corallum ramose, rising from an incrusting base and a short trunk, branching into a more or less regular flabellum. Branches cylindrical, flattened, and slightly expanded at the tip. The tendency in branching is towards considerable symmetry between the two halves of the flabel- lum. Three kinds of pores ; very small, linear, over the whole ca3nen- chyma ; larger tubulated, with very minute aperture when unbroken, and larger round or oval ones scattered irregularly. Internal structure coarsely porous ; the larger pores expanding at the bottom into a flat cavity communicating laterally with other canals. Tabulae very thick and not numerous ; the interposed chambers soon filling up. Occa- sional round cavities are found in the centre of the branch, filled with a yolk-like substance contained in a membrane. Color gray. Height 5 to 7 cm. Diameter of branches about 5 mm. Diameter of larger pores .4 to .5 mm. OfF American Shoal, in 98 fatlioms. Off Alligator Reef, in 118 fathoms. OlT Sand Key, in 123 fath(jms. OHTthe Samboes, in 12.5 fatlioms. Off Sand Key, in 13-5 fitlioms. Off Sand Key, in 143 fathoms. Off Sand Key, in 154 fathoms. It has also been found, according to Dr. Duncan, in from 500 to 600 fathoms, in the cold area to the northward of the British Islands by the " Porcupine " Expedition. NO. IV. 8 58 PLIOBOTHRUS. Pliobothrus tubulatus Pourt. Heliopora iubulata Pourt. Bull. Mus. Comp. Zobl., No. 6. Plate IV., fig. 9. Brandies not much cliviclecl, cylindrical ; the tubulated pores very prominent and slender, giving to the coral a hirsute appearance. The internal structure appears to be of the same pattern as in the jDreced- ing species. Height, incomplete, 2.5 cm. Diameter of branches about 3 mm. Off Havana, in 270 fathoms. GEOGRAnilCAL DISTRIBUTION. We are as yet too little acquainted with the deep-sea fauna of other parts of the world to be able to trace the distribution of our corals. The only parts of the sea bottom well exj)lored up to the present time, besides the Florida Straits, are along the coasts of Europe from Norway to the Straits of Gibraltar and the Mediterranean. The comparison of the Echinoderms of the tAvo regions has revealed quite a number of species common to both (see Bull. Mus. Com. Zool., Nos. 9 to 12). But this is not the case with the corals ; Dr. P. M. Duncan, in his paper on the Madreporaria dredged up in the expedition of H. M. S. " Porcu- pine," enumerates four species common to the northern seas of Europe and the Florida Straits, namely, DiploheUa profunda, Lopholielia prolifera^ Thecopsammia socialis and PUobothrus sfptimetricus. Lopholielia prolifera is found also in the Mediterranean. To these can be added CaryopUijllia claims, which has about the same range as the last. Madracis asperula and Cladocora debUis are found in Florida and Madeira, but they cannot properly be called deep-sea corals. The observations on the coast of Cuba are so few, that no definitive conclusions can be drawn as to the difterences between the faunae of the two sides of the straits, but I have found no reason yet for doubting the statement made in a former paper, that certain species are found only on the Cuban side ; for instance, Sti/lastcr complanatm, Dldichopora mdcaUi, Errina carinata, Lepidopora (jlahra, Deudroph/jllia cyathoides, and Pliohothnis iubidatns. Examples from other classes can also be found, which appear never to have crossed the Gulf Stream and the straits : such is the Pentacrinus. TABLE OF DISTRIBUTION ACCORDING TO DEPTH IN FATHOMS. Lit 10 9. 20 26. 40 36. 45 43. 36 36 36 43 60 68 68 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 etc. 450 460 1050 600 600 450 Carj'ophyllia clavus Caryophyllia forniosa Caryophjllia cornuforniis Stenocyathus verniilbniiis Trochoeyathus coronatus Thecocyathus cylinilraceus Thecocyathus Ijtvi^atus Leptocyathus Stiuipsonii Paracyathus confertus Paracyathus follieulus Beltocyathus Agassizii Parasinilia prolifci-a Parasmilia Lyuianii Parasinilia variegata Caelosmilia fecuuda Oculiua varicosa Oculina tenella Diplohelia profunda Lopliohelia prolifera Lophohelia exigua Madracis asperula Cladooora debilis {Phyllangia ainerioana) (Colaugia iumiersa) (Astrangia soUtaria) Stylaster erubesccns Stylaster filogramis Stylaster Duohassaingii Stylaster punctatus Stylaster complanatus Allopora miniata Cryptohelia Peircei Distichopora sulcata Distichopora foliacea Errina carina ta Lepidopora glabra Balanophyllia floridana Thecopsaniniia tintinnabulum Thecopsaniniia socialis Dendrophyllia cornucopia Dendrophyllia cyathoides Fungia symmetrica Diaseris pusilla Diaseris crispa Mycedium fragile Haplophyllia paradoxa Pliobothrus symmotricus Pliobothrus tubulatus 1 .270 .270 100 1 174 234 125 (315) (315) 327 (315) (315) (315) (315) (315) 324 (315) 350 100 160 237 125 60 70 60. 68 .50 147 .77 195 287 .63 9 .120 120. 26. 43 .63 124 195 270 C 270 270 C 262 270 C 270 262 270 C 270 C 100 324 350 ion 154 18f) 195 1 .126 98. 120. 119 119 IS't 154 N. B. — The letter C added to the figures indicating the depth means that the species was found exclusively on the coast of Cuba. A parenthesis indicates that the .specimens were probably carried to the locality by currents. Names in parentheses are littoral species. BATHYMETRICAL DISTRIBUTIOTi . The bathymetrical distribution is exhibited on the accompanying table. From it we can readily separate the species found in depths less than 100 fathoms, which inhabit the region of debris between the reef and the rocky plateau beginning at the above depth. They are the following : — Parasmilia prolifera. Parasmilia variegata. Oculina varicosa (dead fragments only). Oculina tenella. Cladocora debilis. Stylaster Duchassaingii. Balanophyllia floridana. Mycedium fragile (found also on the reef). A few others extend from the same region into the next or rocky one ; but as the limit is not a sharp one, it is difficult to say if they really belong to the one or the other, or were gathered by the dredge from patches of rock in the muddy bottom, or muddy patches in the rocky bottom. They are : — Caryophyllia clavus. Caryophyllia formosa (decidedly on rock on Cuban coast). Leptocyathus Stimpsonii. Deltocyathus Agassizii. Parasmilia Lymanii. Cselosmilia fecunda. Lophohelia exigua. Madracis asperula. Stylaster punctatus (see remarks in description). Pliobothrus svmmetricus. The characteristic species of the rocky bottom, at greater depths than 100 fathoms, are : — Stenocyathus vermiformis. Thecocyathus cylindraceus. 62 BATHYMETRICAL DISTRIBUTION. Tliecocyatlius la3vigatus. Lopbohelia prolifera. Stylaster erubescens. Allopora miniata. Distichopora foliacea. Tliecopsammia tintinnabulum. Tbecopsammia socialis. DenclropbylUa cornucopia. Dendropbyllia cyatboides. Diaseris pusilla. Diaseris crispa. Pliobotbrus svmmetriciis. Pliobotbrus tubulatus. In tbe Globigerina mud we find only tbe following species : — Caryopbyllia cornuformis (living). Paracyatbus folliculus (dead). Diplobelia profunda (dead). Fungia symmetrica (dead). i Haplopbyllia paradoxa (living). On tbe wbole tbe differences between tbese different regions are not very great, at least as far as family and generic cbaracters go. But tbe difterences from tbe nearest reef fauna are, as migbt be expected, very striking. No Astra3ida3, no Madreporida) (excluding tbe Eupsam- midie), tbe most abundant families in species and in individuals on tbe West Indian reefs, extend to any considerable deptb. Tbe flmiilies bav- ing apparently tbe greatest range in deptb are tbe Oculinidae, the Stylasteridie, and tbe Milleporidae. Simple corals wliicb form sucli a large proportion of tbe deep-sea fauna are not represented at all in tbe Florid ian reef fauna ; some species are described from tbe West Indies, but witbout indications of deptb. Tbe specimens of soundings brougbt up by tbe lead from tbe Gulf Stream fartber nortb contain frequently small fragments of corals, of wbicb tbe following liave been determined : — Cryptobelia Peircei, lat. 31° 32' N., long. 78° 20' W., 600 fatboms. Disticbopora foliacea, lat. 31° 32' N., long. 78° 20' W., 600 fatboms. Stylaster erubescens, lat. 31° 32' N., long. 78^ 20' W., 600 fatboms. Lepidopora glabra, lat. 31° 32' N., long. 78° 20' W., 600 fatboms. Diplobelia profunda, lat. 30° li' N., long. 79° 35i' AV., 530 fatboms. Caryopbyllia formosa, lat. 30° li' N., long. 79° 35^' 530 fatboms. Stylaster erubescens, lat. 29° 48"' N., long. 79° 17'"w., 450 fatboms. Lepidopora glabra, lat. 29° 48' N., long. 79° 17' W., 450 fiitboms. Diplobelia profunda, lat. 28° 53^' N., long. 79° 33f ' W., 410 ftxtboms. Tbese localities are all iu or near tbe axis of tbe Gulf Stream, be- BATHYMETRICAL DISTRIBUTION. 63 tween the latitudes of Cape Canaveral and Savannah. The specimens are all broken and worn and associated with Foraminifera remarkably clean and free from mud. The indications are that a current sweeps over the bottom in a direction from south to north : in other words, the Gulf Stream extends to the bottom at least as far north as the highest latitude mentioned, and is not underlaid by a cold arctic current running in opposite direction, as has sometimes been assumed to account for the low temperature at the bottom. We know that the corals enumerated above (with the exception of Diplohelia in-dfunda, which I have never found alive) live on the coast of Cuba and Florida in depths not less than 100 fathoms, and are thus out of reach of transportation by superficial agencies. A possible nearer habitat would be the northern extremity of the Bahama Banks, between latitudes 27° and 28°, and perhaps somewhat farther north on the extension of the banks which is known to exist on the eastern border of the Gulf Stream ; yet even in this case the trans- portation would be in the same direction. We have positive evidence of the depth to which the current extends in the straits, in the experi- ment of Assistant H. Mitchell, United States Coast Survey, whose conclusion is, " that the Gulf Stream has a nearlij uniform velocity and constant course for a depth of six hundred fathoms, although its temperature varies in this depth 40° Fahrenheit." (See Report of the Superintendent United States Coast Survey for 1867, pp. 176-179.) REEF CORALS. The corals constituting the Florida Reefs, or growing in the shoaler waters of that region, belong to the West Indian Fauna, without exception. They have been made a special object of study by Professor Agassiz for many years, and this part of the catalogue was to have been written by him. But as the multiplicity of other labors has thus far prevented him from putting this project in execution, and as a long absence from the Museum will still further protract the delay, it has appeared preferable to add this list, with remarks from notes made by me in the field and in the Museum, so as to have as complete a view of the coral fauna of Florida as possible. Professor Agassiz's future work will bear chiefly on the structure of the polyp as well as of the hard parts, and on the geological part taken by the different corals in the formation of the reefs and of the peninsula of Florida. Family OCULINIDiE Verrill. OCULINA M.-Edw. & H. Oculina Lamk. (pars). Ocidinu EilHBG. (pai^). Matr-epora Oken. Oculina Dana (pars.) Oculina diffusa Lamk. Non Oculina varicom Lesueuu, Daxa. Non Oculina jxillcns Ehkbg., Dana. Very common. Low-water mark to fifteen fathoms. NO. IV. 9 66 OCULINA. Oculina varicosa Lesueur. Madrepora mammillarts Ellis ? Oculina varicosa Dana, OcuJinn Peticeri M.-Edw. & 11. As pointed out by Mr. Verrill (Bull. Mus. Comp. Zool., No. 3), this species is entirely different from 0. diffusa, of which M.-Edwards and Haime had supposed it to be a synonyme. Lesueur's description and figure are quite insufficient to determine if he really had this species before him ; it has been, in point of fact, first described by Dana. It has not yet been found in Florida in a fully developed state, to my knowledo;e. I have referred to it some branchlets which I had at first described as Oculina disticha (see Deep-Sea Corals, p. 22), dredged off the Eeef. The Museum possesses fine specimens from Bermuda; it has also from the same locality specimens of Oculina pallens, Elirbg., Dana, which is certainly not synonymous with O. diffusa, as supposed by M.- Edwards and Haime. Oculina robusta Pourt. Corallum sparsely branching ; trunk rather massive (3 to 4 cm. in diameter). Calicles numerous, about 3.5 mm. in diameter and about the same distance apart, with prominent border, forming more or less regular spiral lines. Three cycles ; primary and secondary septa about equal, tertiaries but slightly smaller. Pali small, pointed, not very difierent from the columellar papillae, which are very numerous. Costal striae well marked, much curved, broad and fiat, delicately granidated. One specimen (No. 857) in the Museum was received from Dr. Holder at Tortugas, Florida, to whom the Museum is also indebted for a fine series of corals from that vicinit}^ I am informed by Mr. A. Agassiz that there is another fine specimen in the New York Free Academy, from the same locality. Oculina arhuscida and Oculina implicala, Ag. Ms. (Verrill, Bull. Mus- Comp. Zool. No. 3) are not found in the reef region. The former is common on the coasts of North and South Carolina, and its northern limit is probably Cape Hatteras. Tlie second is only known from specimens thrown up on the beaches of North Carolina, and has a semi- fossil appearance. EUSMILIA. 67 ^ Family STYLOPHORIDiE Pourt. Pseurloculince M.-Edw. & H. Ann. Sci. Nat, 3« Ser., Vol. XIII., 1850. SlylophorincB M.-Eow. & II. Hist. Nat. Coral., 1857. MADRACIS M.-Edw. & H. Madracis decactis Verrill. Axtrea decactis Lymax. Proc. Bost. Soc. Nat. Hist., Vol. VI., 1857. Found in depths ranging from 6 to 17 fix thorns. Not abundant. (See remarks in Deep-Sea Corals.) Family ASTR^ID^ Dana. Subfamily EUSMILIN^ M.-Edw. & H. EUSMILIA M.-Edw. & H. Eusmilia fastigiata Edw. & H. Seha, PI. 109, fi;r. i ; PI. 110, fig. 3 ; and PI. 116, fig. 4 ? Madrepora fasligiata Pallas. Fig. 53. Non Madrepora fin^tifi'mt a Ellis & Sol. PI. 33 ? * Non Madrepora capllnta Esper. PI. 82 (copied from Ellis) ? Non Carynphi/llia fustii/lafa Lamx. PI. 33? Mailrepora fastigiata Espkr. PI. 8 ? and var. /3, PI. 8 A ? Eupht/Uia aspera Dana. PI. 9, fig. 7. Non Musaa fastif/iata Dana. Based on Ellis's figure. Eusmilia aspera Ducn. & Mich. (They qaote Dana's figure.) This species is well characterized by its finely serrated costae, which do not extend far down from the calicle. The septa of the cycle pre- ceding the last have well-formed but very thin pali or paliform lobes, which are not found in the next species. Rare. The Museum has only one small specimen found living at Looe Key, Florida Reef, and a larger one from Cumana. * I have not seen any worn specimens with the septa forming such points as represented. The species cannot be made out satisfactorily. 68 PECTINIA. Eusmilia alticostata M.-Edw. & H. Several specimens in the Mnseum from Florida agree better with the description of this species than with the description and fignre of Eusmilia asperd M.-Edw. & H. (not Euph>/Uia aspera Dana), of which I have no typical specimens for comparison. As suggested by M.-Edwards and Haime, this may be only a variety. The columella in om- specimens is very little developed ; the costae thin, foliaceous, forming a succession of lobes, extending far down, very indistinctly or not at all serrate. The calicles are generally considerably smaller than in Eusmilia fasii- giata. The differences thus appear to be more considerable than the short description of Milne-Edwards and Haime would suggest. Florida Reef, rare ; Agassiz, Pourtales. DICHOCCENIA M.-Edw. & H. Dichocoenia Stokesi INI.-Edw. & H. D. Slokem M.-Edw. & II. Ann. S •. Ntit, 3"^ Ser., t. X., p. 307, PI. 7, fi^;. 3, 1849. The Museum has but three specimens of Dichocoenia from Florida, which present considerable differences among each other, particularly with regard to the intercalicinar surface, which in the larger spechnen is entirely occupied by thick granular costtie, the calicles being much more crowded than in Milne-Edwards and Haime's figure. In the others, which are young specimens, these spaces are coarsely granular, with less distinct costte. There is also considerable diversity in the development of the pali. They are all referred provisorily to the above species. A specimen from Hayti resembles the figure more closely. It is quite difficult to make use of the descriptions and figures of the various species established by Duchassaing and Michelotti, which are mostly based on very variable characters. PECTINIA Oken. Pectinia maeandrites M.-Edw. & H. Very rare. When examined with a magnifier, the septa are seen to be distinctly though very finely serrate. ASTROCiENIA. 69 DENDROGYEA EiirbCx. Dendrogyra cylindrus Ehrbg. Very rare ; only found occasionally thrown up on the beaches. Fine specimens in the Museum received from Mr. Duchassaing, of Guada- loupe, show very distinctly the mode of growth of the prongs arising from large hemispherical masses. Subfamily ASTR^INJE M.-Edw, & H. ASTROC.ENIA M.-Edw. & H. This genus is known only from the oolitic, cretaceous, and tertiary formations. It was placed by Milne-Edwards and Haime among the StylinaceaB, on account of its supposed entire septa. Reuss, who had better preserved specimens, has shown that they are denticulated. I have found this character strongly marked in some specimens of A. reticulata from Gosau, particularly in some calicles placed in a cavity where they had been protected from abrasion. Heller has referred to this genus a small coral discovered by him in the Adriatic. From the description and figure, however, it would seem to be a Madracis, as it has ten equal septa. Astrocaenia pectinata Pourt. Corallum small ; incrusting. Calicles crowded ; wall forming a sharp ridge ; costa) spinous, gen-erally not confluent. Fossa deep. Septa in two cycles, not very different from each other ; six complete systems ; the lamella of the septa rudimentary, but extending into long spines, which join the columella in the lower parts. Columella rather stout, conical, hollow, free to a considerable depth. Dissepiments not numer- ous. Growth by lateral gemmation. Diameter of calicles 1.5 mm. On base of a Madrcpora iirolifera, in Mus. Comp. Zool. 70 ISOPHYLLIA. LITHOPHYLLIA M.-Edw. & H. Under this name the young of several other genera have probably been described, such as Mussa, Isophyllia, Symphyllia, etc. The ques- tion can only be decided by series of specimens in different states of development. I leave the following species under that genus for want of that test, but should hesitate much before giving them a definite place in the system. Lithophyllia cubensis M.-Edw. & H. L. argenome ? Ducii. & Mich. L. cylindrica? Ducii. & Mich. Two specimens obtained at Tortugas show no appearance of division into lobes, though 3 to 3.5 cm. in diameter; another smaller specimen comes from Bermuda. The sej^ta and their spines are covered laterally with small spiniform tubercles, which give them a hirsute appearance. The spines of the younger septa are irregular, branching, sometimes coalescing and perforated. My specimens were found growing on dead branches of Madrcpora cervicornis. Lithophyllia dubia Ducii. & Micii. A fine specimen of this supposed species agrees wdth the figures of Duchassaing and Michelotti. I suspect it to be the young form of an IsophylHa or Mussa (perhaps lltissa cardans). My specimen shows a beo-innino; of division into lobes. It was obtained on the reef at Looe Kev. ISOPHYLLIA M.-Edw. & H. Isophyllia multilamella Pourt. Litliophjllia multilamella Ducn. & Mich, (young specimen). The differences between Isophyllia and SymphyUia are so slight as to lead to continual confusion. I prefer to merge them under the above name. The septa in this species are very closely set, there being from 13 to ISOPHYLLIA. 71 15 in a centimeter. The ridges are sometimes slightly flattened on top and marked generally with a narrow but distinct furrow. Columella w^ell developed, papillose. Though not sure of the identification, as indeed will be mostly the case in any attempt to identify the numerous species of Symphyllia too imperfectly characterized by Duchassaing and Michelotti, I prefer not to introduce a new name, as there are, no doubt, many superfluous ones in the number. The Museum of Comparative Zoology has a sjiecimen from Florida and one from Bermuda. Isophyllia dipsacea Ag. Ms. Sj/mph/llia f dipftacea Edw & H. Mussa dipsacea Dana. Non SijmphijUia guadulpensis Edw. & H. Rather common, quite a number of fine specimens in the Museum of ComjDarative Zoology. Isophyllia guadulpensis Poijrt. SympliyUia guadulpensis M.-Edw. & H. Isophyllia rigida Vehrill. Non Astrcea rigida Dana. The specimens labelled by Mr. Yerrill all agree in having lamellar costa3 when young, much more marked than in any other species. The other characters are rather vague. The diagram of the calicles of Astrcea rigida given by Dana showing a hexao-onal limitation of the calicles, so different from the irreorular, partly meandriform shape of those of the species in question, seems to preclude the identification as made by Verrill. The imperfect descrip- tion seems to call for a new. one. Corallum flat-hemispherical ; epitheca rudimentax}^ or none ; cost^e sul^lammellar, particularly in the joung, spinulous ; calicles rarely simple, mostly in sinuous series with two or three centres. Ridges rounded, with very small furrow on top. Septa closel}^ set, rather equal, 10 to 12 to a centimeter, sparselj^ granuhir on their faces, with rather strong triangular teeth ; columella papillose. The young divide quite regularly into six lobes, corresponding to the systems, the primary septa enlarging and forming the partitions between the lobes. 72 MANICINA. MYCETOPHYLLIA M.-Edw. & H. Mycetophyllia Lamarckana Edav. & H. Eatlier rare. I have found it living on the reef at Looe Key, and dredged it np from five ftithoms in the channel at Key West. Mycetophyllia Danaana Edw. & H. Several worn specimens in the Museum are referred to this species ; also a young one obtained living by myself Rare. MUSSA Oken. Mussa angulosa Oken. Mussa carduus Dana. Eare. Worn specimens. MANICINA Ehrbg. Manicina areolata Eiirijg. Ver}^ common from about three feet below low-water mark to five or six fithoms, chiefly in sheltered places. It is particularly {il)undant on ' sandy or muddy flats, partly overgrown wdth sea-grasses (Thalassia), in- side the reef. The young when developing normally assume a very elongated form with wavy sides, which soon form six lobes, two of them forming the ends; but the primary or secondary septa are not enlarged, as in Isophyllia, to form the partitions between the lobes. Although a great diversity prevails in the adults, with regard to the general shape, I can find no specific diflcrences among the large number of specimens examined. I have not seen authentic specimens of the other species enumerated by the authors as found in the West Indies; but from the descriptions I can see no good reason for admitting more than one West Indian species. M.EANDRIXA. 73 M^ANDRINA Lamk. Pallas had already pointed out the difference between the meandri- form corals with entire septa (Pectinia) and those with denticulate septa (Miieandrina) ; the former he called Madrepora lahyriniliica, the latter Madrepora mwandritcs. After him, however, they Avere placed in the same genus by most of the authors, until Oken finally separated them. Milne-Edwards and Haime subdivided the true Moeandrinse into several genera, though not always very successfully. Thus Manicina and Mseandrina are said to differ by the presence of paliform lobes in the former and their absence in the latter, whilst in reality they are fully as much developed in most of the MaBandrinae as in Manicina. These genera differ only by their mode of growth. The genus Coeloria, of iliQ same authors, has been separated from Mseandrina by exceedingly vague characters ; it is not difficult to find parts of the same polypidom which could be made to fit one or the other of these genera. No doubt much of the confusion existing in this group is due to a considerable range of variation in the same species, and to the large proportion of worn specimens which have served as types for description. I can recognize with tolerable certainty but three species on the coast of Florida. McGandrina labyrinthiformis Oken (Maeandra). ]\Iadrepnra JahiirlnthifonnU (pars) L'NX. Madrepora Inhi/rinthira Em.fs & Sol. PI. 4G, figs 3 and 4. Madrepora ma^aridrllcs (pirs) Espk.r. PI. 4 A. J\hranjUia pseudolurhinoUa 8 25 Caryophyllia solilaria 31 79 . 52 Caryophyllidce 7 53 Ceratocyathus .... . 19 . 53 Ceratocyathus prolifer . 19 54 Cladocora .... 30 78 . 54 Cladocora arbuscula 78 55 Cladocora brevis , 78 76,80 Cladocora debilis . . . . 30 76 Cladocora multicaulis • • 78 28, 67 Cladocora parvistella 78 81 Cladocoracese .... 30 78 . 71 Cocloria lahyrinthica . . 73 81 Cceloria labyrinlhijormis • 73 . 76 Cceloria sirigosa 74 29, 67 Cuelosmilia .... • • 21 . 69 Cadosmilia fecunda 21 31, 79 Cienocyathus verinifarmis . • • 10 . 80 Colangia .... .31, 80 80 Colangia immersa 31 80 , 80 Colpophyllia .... 74 31, 79 Colpophyllia gyrosa . • • 74 31, 79 Cryptohelia 37 30, 79 Cryptohelia Peircei . . -^ Z'l . 79 Cyathina 7 77, 80 Cyathina cyathus . . 8 . 77 76 . 75 Cyathina pseudolurhinoUa 8 Deep-sea corals , , 7 69 Deltocyathus . . . . . 15 . 69 Deltocyathus Agassizii . . 15 27 Dendrogyra . 69 . 28 Dendrogyra cylindrus . 69 Uondropliyllia . . . . • 45 . 41 Dendrophyllia cornucopia . 45 41 Dendrophyllia cyathoidcs . 45 . 60 Desmopiiylluni .... • 16 88 ALPHABETICAL INDEX. Desmophyllum Cailloti . 16 Desmopbyllum soliduin . 17 Diaseris crispa 47 Diaseris pusilla .... . 47 Dichoccenia .... 68 Dichocoenia Stokesi . . 68 Diplohelia .... 23 Diplohelia profunda . . 25 Diploria .... 75 Diploria cerebriformis . 75 Distichopora .... 38 Distichopora cervina . 39 Distichopora foliacea 38 Distichopora sulcata . 38 Errina . 39 Errina carinata 39 Errina cocJdeala . 40 Errina (glabra 40 EiipJiijllia asj)era . 67 Eupsammidaj 41 Eupmnimince .... . 41 Eusuiilia .... 67 Eusiuilia alticostata . . 68 Eusmilia aspera • 67 Eusmilia fastigiata . 67 Eusmilina; .... 67 Explanaria annularis . 77 Favia . 75 Favia ananas 75 Favia rotulosa .... . 75 Faviacea} .... 75 Fissicella ..... . 46 Fungia symmetrica 46 Fungida3 ..... 46, 80 Galaxea eburnea . 29 Geographical distribution . 58 Haplophyllia .... . 51 Haplophyllia paradoxa . 52 Hdiastrcea .... . 76 Heliastrcca annularis 77 Heliaslrcea cavernosa . 76 Hdiastrcea conferta 76 Heliastrcca stellulata . . 77 Heliopora carinata 39 Hydroidai .... 56, 85 Isophyllia .... . 40 Isophyllia multilamella . 70 Isophyllia dipsacea . . 71 Isophyllia guadulpensis . 71 Isophyllia rifjida . 71 Lepidopora .... . 71 Lepidopora cochleata . 40 Lepidopora Dabneyi . 41 Lepidopora glabra Leptocyathus . Leptocyathus Stimpsonii Lithophyllia Lithophyllia cubenjiis LithopliyUia an/enome Lilliop)hi/llia cijUndrica . Lithophyllia dubia . Litliopliyllid multilamella Lophohelia LopJiolielia affinis . Lophohelia Carolina . Lophohelia exigua Lophohelia prolifera LophoseriniE . Madracis . Madracis aspcrula Madracis decactis Madrepora . Madrepora agaricites Madrepora alces . Madrepora ananas . Madrepora annularis Madrepora astroites . Madrepora capitata Madrepora cavernosa Madrepora cervicornis . Madrepora cornuta . Madrepora faveolata Madrepora fastiijiata Madrepora JiaheUum Madrepora galaxea . Madrepora lahyrin th ica Madrepora labyrinth if ormis Madrepora maminillaris Madrepora mceandrites Madrepora jialmata Madrepora pleiades . Madrejjora prolifera Madrepora radiata . Madrejjora rotulosa Madrepora stellulata Ma drep or a Thomas ia n a MadreporidsB . Ma^andrina . Ma?andrina clivosa . Mceandrina grandilobuta Mceandrina labyrinth ica Mffiandrina labvrinthiformis Mceandrina mam mosa Mceandrina sinuosa Maeandrina strigc^ . '' 5, -,■■< V^ii^Jm" >VH'- Vl^^-^ ■,yd^^',>- .-^^ 18. ^^. •.^*'. r ^^P A '-y;:' ■^^M fe:^>^ ^tr r ^4»v J'.Hceaeronswne iromnaL, Prill t. ty H.'WLongf ellow. Deep Sea Ccrak PI. II. ^1%. ^ ^Miii :^'..vri>?:?^V<5>^^ 12. S . ^#^^' ^ir,!l' -A^ 14. /I^t!*^"^/*-;,J>!fr.,K ,1^^:'^^^ -^ * .■■ >^ 'v.'**r: #'"" 10. "P. ;i setter on stone fromnat Print 'by H 7,'Z"LoTigfellcfw. DeeD Sea Corals. ?!. ..',SV^?:XA, t ' i4 If: . ** KRcetter on stone from nat -'.on.tffellow. Deep 'Jea Curals Pi.lA;. F Rcetter on ^*or,s from na K ■ ' I ,rr !i ef'^' inv; Deep Sea Corals. Pl.V. iff ^if^^fT^M^Sv' f^f ^-, if/'?; € *■.. PRoetter on scone {roranaL Prmt/byH.W. Longfellow. Deep Sea Corals. Pl.VI. . -I retisr on sane from ;i a ' Erint.by 11 .WI- ongf sji uw Deep Sea Corals. ^l.YIl. SiS^^fej^ P.K.ce;teroristo)ie from nn'. Printliy H.W.Lonpfellow. ■Tahr^i^ 1870, "afelao C.DTM .ICSTLIS ITJITHES ^*C" CC < 4SC :<. c < . tit r c JCC < c s. :S _: Sfzc c cr ,>v c CC < vC < ■ ..< c ^ccc Cv c rf cc<:;^x-i <;ccr'Cr C CC Ct vc" c CC <<:•.. c C f C C r r ^c.c-- -' -^' CC C' V c<:-- ex cccx' CCc-. *r x^ " ^sc*-*^ ' cc ec ^^deufC- <^ C ^C^SCC C C C^ ; ^ L «:.<'<-.'■ « < = c* vc.'v^*_v.^.ji.^j*£^^CC^ V ^: "^^ *^' *^ *^^ < rrce c-dc: