DUKE 
 
 UNIVERSITY 
 
 LIBRARY 
 
Digitized by the Internet Archive 
 
 in 2016 
 
 https://archive.org/details/brachiospongidae01beec 
 
MEMOIRS OF THE PEABODY MUSEUM OF YALE UNIVERSITY. 
 
 VOL. II, P'ART I. 
 
 BKACHIOSPONGIDBE: 
 
 0 
 
 % 
 
 A 
 
 MEMOIR 
 
 ON A 
 
 GROUP OF SILURIAN SPONGES. 
 
 WITH SIX PLATES. 
 
 BY 
 
 CHARLES EMERSON BEECHER. 
 
 NEW HAVEN, CONN.: 
 
 PBINTED FOE THIE MUSEUM. 
 
 1 88 9 . 
 
 0 HR f Ui 
 
 H 
 

 
 
 
 - 
 
 v> • ... 
 
 
 ■ . VS, '■ 
 
 , -■ r 
 
 ■ 
 
 . . ... . ' 
 
 ■ 
 
 - h V ■ ■ • -r ^ -v; : 
 
 
 
 WM 
 
 " ; -• ‘ ■ ■ " ; • tf • /.,;•* fe ■ .•*.**/ ''/> 
 
 - V .• • . ■, ■••• ■' " . >> ; Kg 
 
 ■ ", v- ' - 
 
 
 v • ■ ; , •• i i .* *v • •.< >! . \ •• *v» • 
 
 — - ;r.-> / - 
 
 ‘ r; ' - ■ ■ ■ ; .. - 
 
 . s 
 
 
 
 
 
 ; 
 
 - - , - _ . ■ : . 
 
 , . V •. . r , ■ : , •/' ■' • • • /»*• - 
 
 .• • ' . * • • • - • ff i« - 1 
 
 
 1 ' . 'j 
 
 
 
 
 ' - V- S 
 
 
 
 press, of Tuttle, Morehouse & taylor, New haven, conn. 
 
 ■v' • •* " 7 -■ . 
 
 
 - i, - 
 
 , 
 
 
 
 
 
 , 
 
 - 
 
 . 
 
 
 
 ■'-T'ih " 
 
 «'r'- 
 
 
 
 /■' 1 7 ' .■ \ ‘ , r _ 
 
 V.. 1 
 
 
 - 
 
 > t c v Y \ ... ^ ■ • ■ ■ 
 
 . 
 
 ■ - ' ■ : .• -'V'T.". ■" ' ; 
 
 
 ’ /r | 
 
 
 
 
 ,'j,;AVv^OO .in' 
 
 & M 
 
 - 
 
 J < : ' 
 
 •a ■■■■:: 
 
 gjEvV t 5 
 
 .. -■ i' 
 
 
 
 
 ■ v - 
 
 
 
 ■ 
 
 , • - ■ n ,■ s ’ 
 
 • • 
 
 ' ' . ■ '-H 7 
 
 
 , v ' v'j - • 
 
 , - 1 . 
 
 ' . *• r • . • ■. . ' '• v-.f 
 
 - 
 
 
 , ■■■• “ ■■ ■ V • ■' 
 
 ■ ;■ ■ - : ' " 1 " " 
 -• 1/ -••• > . 
 
 1 ' M 
 
 
 ' 
 
 v, •' • ■ ’ - . • 
 
 v" 
 
 ■ 1 ■ 
 
 ■ ■ -• . 
 
 
 
 
 
 . ■ 
 
 ' 
 
 ;1 ,■ ; ■ . 
 
 
 
 
MEMOIRS OF THE PEABODY MUSEUM OF YALE UNIVERSITY. 
 
 VOL. II, PART I. 
 
 BRACHIOSPONG-ID M: 
 
 MEMOIR 
 
 ON A 
 
 GROUP OF SILURIAN SPONGES. 
 
 WITH SIX PLATES. 
 
 CHARLES EMERSON BEECHER. 
 
 NEW HAVEN, CONN.: 
 
 PRINTED FOP?. THE RTTTSPHTIVE. 
 
 1 8 8 9 . 
 
BRACHIOSPONGID^E. 
 
 (With Six Plates.) 
 
 By Charles Emerson Beecher. 
 
 BRACHIOSPONGIA, Marsh. 
 
 These sponges were made known as early as 1838, and were the first 
 strictly American paleozoic sponges described. The systematic position 
 of Brachiospongia, however, has remained hitherto undetermined. Pro- 
 fessor Karl A. von Zittel placed the genus provisionally in the fatuity 
 Euretidse, Zittel, under the order Hexactinellida. This arrangement was 
 followed by Dr. Gr. J. Hinde, although he regarded the position of the 
 genus uncertain, as its structure was unknown. 
 
 The basis of our information of these variable, and often imperfectly 
 preserved fossils must rest largely upon the results of the study of more 
 recent forms, and the interpretation of the fossil must be in the same terms 
 used for the living organism. 
 
 The first systematic grouping of sponges, based upon reliable char- 
 acters, was worked out with great care, in Zittel’s memoirs, published 
 during the years 1876-78. All subsequent writers have accepted the 
 principles which he established, and have adopted his main subdivisions. 
 
 O ocho/’ 
 
4 
 
 BRACHIOSPON GIDaE. 
 
 The later investigations of Dr. G. J. Hincle have largely supplemented our 
 knowledge of the fossil forms, and with the crowning work by Dr. F. E. 
 Schulze, 1 on the recent species of the order Hexactinellida, the paleontolo- 
 gist can readily interpret his observations, and classify the various species 
 and genera which preserve the main features of their structure. 
 
 Histokical account of Beachiospongia. 
 
 1838. 
 
 G. Teoost. Description d’un Nouveau Genre de Fossiles. (Mem. Soc. 
 
 Ge'ol. France, Vol. Ill, pp. 95, 96 ; PI. XI, figs. 8, 9, 10.) 
 
 The new genus described in this paper was a cephalopod (Conotubu- 
 laria), which has since been considered as a synonym under the genus 
 Orthoceras. A new species of Asaphus (A. megalopthalmus ) is also indicated 
 and figured. The paper concludes with “ Description d’un fossile repre- 
 sent^ dans la pi. XI, fig. 8, 9, 10,” where an account is given of a fossil 
 which is at once recognized as Brachiospongia. The specimen has nine 
 arms, and is described as silicified. Dr. Troost suggests its affinities with 
 sponges, and states that he discovered the fossil in a limestone occurring 
 in Davidson County, Tennessee. Singularly, no name was proposed, and 
 apparently, no notice has been taken of the description and figures in any 
 publication, previous to Dr. G. J. Hinde’s Catalogue of the Fossil Sponges 
 in the British Museum, 1883. 
 
 It seems more than probable that the specimen in the possession 
 of Dr. Troost, in 1838, and described by him, is identical with the one 
 which was made the type of Brachiospongia roemerana , by Professor 
 0. C. Marsh, in 1867, and is now illustrated on Plate I of the present 
 memoir. The evidence is based upon the striking resemblance between 
 this example and the figures on Plate XI of Troost’s paper. The general 
 dimensions agree very closely, and in addition, the diameters and height 
 of the neck, and the length of the longest arm are the same. Besides 
 having a like number of arms, and agreeing in size, a more marked 
 
 1 Report on the Hexactinellida collected by H. M. S. Challenger during the years 
 1873-1876. Zoology, vol. xxi, 1887. 
 
RRACHIOSPONGIDiE. 
 
 5 
 
 resemblance may be traced, if the type of B. rmmerana is placed in a 
 slightly inclined position, which is the one naturally assumed on account 
 of the imperfection of the lower portion, and the arms then compared 
 seriatim in respect to direction, length, and size. It must be assumed, 
 however, that the lithograph was made without reversing the drawings, 
 and that the specimen appears reversed on the plate. The inference is 
 supported by the fact that the shadows on all the figures are manifestly 
 on the wrong side. 
 
 A comparison under these conditions shows that the widest divergence 
 between any two arms is adjacent to the longest and unbroken one. The 
 latter arm is oblique to the longitudinal diameter of the osculum, and the 
 three shortest broken arms are adjacent to it, and are situated on the side 
 opposite the greatest interbrachial space. The side view offers similar 
 points for comparison, if the specimen is placed in the proper position. It 
 should also be noted, that this individual shows a greater elevation of the 
 neck, and of the arms near their origin, than has been observed in any 
 other. 
 
 The history of the type specimen is somewhat obscure, but nevertheless 
 furnishes additional proof in support of its identity with the one described 
 by Troost. According to the best evidence at present attainable, it was in 
 the possession of Dr. Yandell in 1855, who at that time was custodian of a 
 portion, at least, of the Troost collection. Rev. H. C. Hovey obtained the 
 specimen in exchange, and subsequently transferred it to Professor 0. C. 
 Marsh, who, after describing it in 1867, presented it to the Museum of 
 Yale College. 
 
 1858. 
 
 D. D. Owen. Kentucky Geological Survey, Vol. II, p. 111. 
 
 In his report on the geology of Franklin County, Owen refers to the 
 principal external features of these sponges, and proposes for them the 
 specific name cligitata, which he places under the genus Scgphia, Oken. 
 The locality of their occurrence is described, and the geological horizon is 
 referred to a position near the middle of a detailed section of Lower 
 Silurian strata, at Frankfort, Ky. 
 
 
6 
 
 BRACHIOSPON GIDAE. 
 
 1862 . 
 
 R. Owen. Indiana Geological Survey, 1859-60, pp. 362, 363, fig. 1. 
 
 A very imperfect specimen is figured, accompanied by a brief descrip- 
 tion, in which the name is changed to Syphonia digitata. In a letter 
 written to Professor Marsh, in 1867, Professor Owen states that the altera- 
 tion in the name was through inadvertence. 
 
 1865 . 
 
 H. A. Ward. Eighteenth Annual Report New York State Cabinet of 
 Natural History, p. 29, fig. 4. 
 
 A wood-cut of an eleven-armed specimen is given, and referred to 
 “ Amorphozoum ? ” The same figure is reproduced by Professor Ward, 
 
 under the name “ Amorphospongia ,” in his “Catalogue of Casts of 
 
 Fossils,” published in 1866. This specimen was made the type of 
 Brachiospongia lyoni , by Professor Marsh in 1867. 
 
 1867 . 
 
 0. C. Marsh. American Journal of Science, 2d Series, Yol. XLIY, p. 88. 
 
 In this preliminary paper, the genus Brachiospongia is first proposed 
 and described, and the species B. roemerana and B. lyoni are named 
 provisionally. B. roemerana was proposed for a nine-armed specimen 
 which now appears to be the original type of Troost’s description and 
 figures. The name B. lyoni was given to a specimen having eleven rays, 
 then in the collection of S. S. Lyon. Photographs and lithographic figures 
 of the type specimen B. roemerana were made at this time, but copies of 
 the former alone were distributed. 
 
 1868 . 
 
 0. C. Marsh. Proceedings of the American Association for the Advance- 
 ment of Science, p. 160. 
 
 A paper was read at the sixteenth meeting of the Association, held at 
 Burlington, Vt., “ On some new Fossil Sponges from the Lower Silurian.” 
 In the published volume of the proceedings, the title only of this paper 
 appears. 
 
BRACHIOSPONGIDiE. 
 
 7 
 
 1874. 
 
 H. C. Hovey. Transactions Kansas Academy of Science, pp. 344, 345, 
 figs. 1, 2. Republished in the Scientific American, pp. 387, 388, figs. 
 1, 2, June, 1875. 
 
 An incomplete historical review is presented of the published notices 
 of Brachiospongia, beginning with the account by D. D. Owen, in 
 1858. A wood-cut of the type of B. rcemerana , made from one of the 
 photographs distributed by Professor Marsh, and a figure of a specimen 
 referred to B. lioveyi (Marsh), first appear in this publication. The name 
 B. lioveyi was proposed for an example with twelve arms, and was 
 announced in the paper read by Professor Marsh, before the American 
 Association. The specimen figured, which was not the type, is an abnor- 
 mal individual of B. digitata, in which the arms have been developed 
 spirally about the axis. A strict enumeration of the partially and fully 
 grown arms indicates fifteen as the number, but the neck is imperfect, and 
 there were probably more in the entire specimen. 
 
 Subsequent notices of these fossils are mainly of the character of 
 references, and include the following : 
 
 1877. 
 
 Joseph LeConte. Elements of Geology, p. 302, fig. 290. ( B . rcemerana.) 
 S. A. Miller. The American Palaeozoic Fossils, p. 42. ( B . digitata , 
 
 B. lyoni , B. rcemerana.) 
 
 1878. 
 
 K. A. Zittel. Handbuch der Paheontologie, Band 1, p. 173. ( Brack - 
 
 iospongia.) 
 
 1880. 
 
 F. Rcemer. Lethsea Geognostica, 1 Th., p. 319, fig. 61. (B. rcemerana.) 
 
 1883. 
 
 G. J. Hinde. Catalogue of the Fossil Sponges in the Geological Depart- 
 
 ment of the British Museum, p. 102. (B. digitata = B. rcemerana and 
 
 B. lyoni.) 
 
8 
 
 RRACHIOSPONGID^E. 
 
 Acknowledgments. 
 
 The abundance and richness of the material which has been placed at 
 my disposal is due to the long continued efforts of Professor 0. C. Marsh, 
 who for several years secured all the specimens found at the principal 
 locality. It was the original intention of Professor Marsh to publish a 
 memoir upon the genus Brachiospongia, as soon as the completeness of 
 the material should warrant it ; but the development of his researches in 
 the fossil vertebrate fields of the Western States and Territories soon 
 reached such proportions, and became so important, as to preclude the 
 carrying out of his original purpose. The publication of this memoir in 
 its present form is also due to the kind liberality of Professor Marsh. 
 
 I wish to express my obligation to Mr. Moritz Fischer, of the 
 Kentucky Geological Survey, for the privilege of examining and figuring 
 a fine twelve-armed specimen belonging to the State Museum, and, also, 
 for courtesies shown me during a recent visit to the region where these 
 fossils have been obtained. 
 
 The collections have been considerably enriched through the efforts 
 of Mr. E. C. Went, of Frankfort, Ky., in the discovery of a new locality 
 yielding a number of interesting examples occurring in situ, and furnishing 
 the first definite knowledge as to their true position in the geological series. 
 
 Rev. H. C. Hovey, of Bridgeport, Conn., has kindly given me an 
 opportunity to study several specimens in his cabinet, and has com- 
 municated important historical points. He first systematically collected 
 Brachiospongia, and observed the geographical distribution of the genus. 
 A number of the specimens in the Yale University Museum were obtained 
 through his researches. 
 
 Geological Position. 
 
 The precise horizon of the strata furnishing Brachiospongia has been 
 uncertain. Hitherto, all writers have agreed in referring them to the 
 Silurian, the particular beds being unknown. Rev. H. C. Hovey placed them 
 in the Birdseye limestone. 1 None of the specimens, however, present any 
 
 Trans. Kansas Acad. Sci., p. 344, 1874. 
 
BRACHIOSPONGIDJE. 
 
 9 
 
 evidence of having- come from that horizon, and all the known localities 
 furnishing Brachiospongia are at some distance from exposures of the 
 Birdseye, and at a much higher level. A single example was found in a 
 loose piece of rock, by Mr. Moritz Fischer, at Benson Station, Franklin 
 County, which was apparently derived from strata belonging to the middle 
 of the Trenton series. So far as known, this is the lowest horizon at -which 
 these fossils have been observed. At the famous locality on Benson Creek, 
 about three miles above Benson Station, where nearly all the specimens 
 first discovered were obtained, no accurate determination of their position 
 could be made, as they were found loose in the debris on the hillsides, 
 and on the banks and bed of the creek. The same observations apply 
 to the locality near Bridgeport, Franklin County. At Cedar Run, in an 
 exposure discovered by Mr. E. C. Went, of Frankfort, Brachiospongia 
 is found in place, and its true horizon and associated fossils have been 
 determined. 
 
 The beds consist of a fine cherty nodular limestone lying above the 
 horizon of Orthis borealis, which has been considered as the upper member 
 of the Trenton. They are succeeded by the lower beds of the Hudson 
 group with their abundant and characteristic fossils. The adjacent lime- 
 stones are without beds of chert, and are coarser grained, so that the rocks 
 containing Brachiospongia are easily recognized. 
 
 With this information, the sponge horizon was traced over a consider- 
 able extent of territory, and connected with the other localities at which 
 Brachiospongia occurs. It was found that with the exception of the 
 single specimen from the Middle Trenton, at Benson Station, all the 
 localities in Franklin County were at the same horizon. Similar beds cap 
 the Trenton at Lexington, Ky. 
 
 Under the description of B. digitata , mention is made of a specimen 
 from the Middle Hudson of Spencer County, Ivy., which presents several 
 points of difference from the ordinary form, both in its general characters, 
 and condition of preservation. As a whole, however, this species has a 
 comparatively limited vertical range, and its occurrence is of geological 
 importance. 
 
 3 
 
10 
 
 BRACIIIOSPONGID^E. 
 
 On account of the faunal and lithological similarity between the 
 Trenton and Hudson groups in central Kentucky, the horizon of B. digitata, 
 taken in connection with the character of the rock and associated fossils, 
 is an important one, as furnishing a definite and easily recognizable bed 
 separating the two formations. The Utica Slate is absent in this region, 
 and without more precise information, it may be desirable to correlate 
 with it the Brachiosponge beds, in which case the latter would represent a 
 synchronous deposit, although physically different, from its sedimentation 
 in deeper waters, probably at some distance from the shore line. 
 
 Associated Fossils. 
 
 Under the microscope, the rock is seen to be composed of the finely 
 comminuted testaceous remains of Crustacea and mollusca, mingled with 
 some argillaceous and silicious material. 
 
 In general, the beds are nearly barren of fossils, and close scrutiny 
 reveals but a meager fauna. The species which have been observed are 
 given in the following list : 
 
 1. Brachiospongia digitata, Owen. Rare. 
 
 2. Strobilospongia aurita, Beecher. Rare. 
 
 3. ? Strobilospongia tuberosa, Beecher. Rare. 
 
 4. Root tufts (basalia) of sponges. Common. 
 
 5. Hindia parva, Ulrich. Common. 
 
 6. An undescribed genus and species of sponge. Rare. 
 
 7. Diplograptus putillus, Hall. Common. 
 
 8. Monticulipora (fragments). Rare. 
 
 9. Arabellites cornutus, Hinde. Common. 
 
 10. Arabellites sp. Common. 
 
 11. Eunicites sp. Common. 
 
 12. Cyphaspis (fragment). Rare. 
 
 13. Beyrichia chambersi, S. A. Miller. Rare. 
 
 14. Leperditia sp. Rare. 
 
 15. Leperditia sp. Rare. 
 
BRAOHIOSPONGIDvE. 
 
 1 1 
 
 16. Zygospira modesta, Say. Common. 
 
 17. Orthis.testu din aria, Dolman. Rare. 
 
 18. Tellinomya obliqua, Hall Rare. 
 
 19. Cleidophorus fabula, Hall. Rare. 
 
 20. Cyclora minuta, Hall. Common. 
 
 21. Microceras inornatum, Hall. Common. 
 
 22. Bellerophon sp. Rare. 
 
 23. Theca parviuscula, Hall. Common. 
 
 24. Conularia trentonensis, Hall. Rare. 
 
 It will be noticed that the only common fossils are silicious sponges, 
 graptolites, jaws of annelids, Zygospira modesta , and the minute forms of 
 gastropods, pteropods, and lamellibranchs. The Bellerophon, Conularia, 
 and Theca represent pelagic types, and the remaining forms, not micro- 
 scopic, are so sparsely distributed that they were probably transported 
 specimens, or at least did not flourish where found. 
 
 The Brachiospongia are in their natural position in the rock, and 
 evidently were buried without having been removed from their original 
 bed. The presence of the numerous masses of filamentous spicules, or 
 basalia, may be explained by their having originally grown in the soft 
 mud of the sea bottom, and thus were protected, while the exposed cups 
 of the sponges would be subject to dissolution and destruction unless 
 previously covered. 
 
 Condition of Preservation. 
 
 The spicules of the dermal and gastral surfaces are usually replaced 
 by calcite, and the parenchyma of the sponge is commonly filled with 
 silica Previous to the process of mineralization, the spicules, pores, and 
 interior canals, in many specimens, were enveloped in a thin film of 
 peroxide of iron, and, although the invasion of silica and calcite has 
 involved the entire space occupied by the organism, this film now serves 
 to differentiate the various structures, so that a satisfactory study may be 
 made of their leading features. Many of the more minute and delicate 
 spicular membranes are probably destroyed. 
 
12 
 
 BR ACIIIOSPON GIDaE. 
 
 It has been shown by Dr. G. J. Hincle (Cat. Foss. Sponges, p. 7), that 
 very few of the silicious sponges are preserved as fossils in their original 
 condition, and that a replacement of the spicules by crystalline calcite 
 is of very common occurrence. Even in a silicified fossil sjDonge, the 
 silex is not in its original state, but has been altered or replaced by silica 
 of a different condition. 
 
 The majority of the specimens of Brachiospongia in the collections 
 have been weathered out of the rock, and are represented as silicified 
 casts. Owing to the usual exemption of the dermalia from silicification, 
 the superficial structures are not preserved in weathered examples, nor in 
 specimens which have been freed with acids. The exposed surface of 
 these casts shows merety the extent of the process in the parenchyma, 
 and the inequalities in its distribution. In rare cases, the dermal skeleton 
 has been silicified, and its features obliterated, except the large hypo- 
 dermal pentacts, which are represented by numerous papillae on the 
 surface. 
 
 The first alteration of the original colloid silica of the sponge was, 
 apparently, to calcite, which also filled many of the interior canals, and, 
 with few exceptions, destroyed the axial canals of the spicules. The 
 subsequent invasion of the silica and partial solution of the calcite have, in 
 some instances, separated and enclosed the calc rhombs, and such specimens 
 when weathered now exhibit rhombic cavities in the silica. Ordinarily, 
 however, the calcite has been simply replaced by silex in the form of 
 chalcedony, except in some of the larger cavities which are incrusted with 
 quartz crystals. 
 
 In order to preserve and study the dermalia and gastralia, it is 
 necessarv to secure specimens more or less enveloped in the matrix, and 
 to uncover them with instruments. The limiting structures may be further 
 exposed by the use of caustic potash, which removes much of the argilla- 
 ceous limestone filling the pores. 
 
BRACHIOSPONGIDaE. 
 
 13 
 
 Systematic Position. 
 
 Class SPONGIA. 
 
 Order Hexactinellida, 0. Schmidt. 
 
 Suborder Lyssacina, Zittel. 
 
 Family Brachiospongid^e, nobis. 
 
 Dermalia forming a quadrate mesh, in which are immersed large 
 tree pentacts, or modified hexacts. Parenchyma thick, with large free 
 hexacts, and more or less cylindrical canals, the outer ends of which are 
 covered with the dermal mesh. 
 
 Including the genera, 
 
 Brachiospongia, Marsh. 
 
 Strobilospongia, nobis. 
 
 Genus Brachiospongia, Marsh, 1867. 
 
 (Am. Jour. Sci., 2d Ser., Vol. XLIY, p. 88.) 
 
 [Type Scyphia cligitata, D. D. Owen.] 
 
 Sponge in the form of a broad cup or vase, with a row of projecting 
 processes or arms around the periphery of the base, and into which the 
 gastral cavity is extended. Osculum large, not operculate. Afferent and 
 efferent canal system well developed. 
 
 Dermal skeleton composed of a fine network of spicules, in which 
 are immersed large pentacts with the proximal rays penetrating the 
 parenchyma. Gastralia large and free. The observed parenchymalia 
 consist of hexacts which are sometimes variously modified. Supported on 
 the sea bottom by the peripheral arms and broad base ; not anchored by 
 bundles of spicules. 
 
 4 
 
14 
 
 BRACHIOSPONGIDJE. 
 
 Strobilospongia, gen. nov. 
 
 (From the resemblance to the “Strobila” stage of Aurelia, and from the twisted base.) 
 
 [Type Strobilospongia tuberosa, nobis.] 
 
 Sponge cyathiform or globose, with more or less concentric rows of 
 lobes or lobed expansions, on the surface. Anchored to the sea bottom by 
 means of massive bundles of filamentous spicules (basalia) proceeding 
 from the interior of the base of the cup. The bundle of basalia is well 
 defined at its origin, and does not merge into the tissues of the sponge. 
 The spicular structure of the parenchyma, so far as observed, agrees with 
 Brachiospongia. Both the known species show numerous cruciform 
 spicules on the surface, but they are so obscured b} 7 silicification that 
 their detailed form and relations cannot be ascertained. Smaller spicules 
 and traces of a continuous dermal mesh are also imperfectly indicated. 
 
 Any systematic division of paleozoic sponges into orders and families 
 is necessarily incomplete and unsatisfactory. The structures are but 
 partially, and often not at all preserved, and, therefore, the data for a 
 settled classification are to a large degree wanting. It is, however, of 
 importance to give the geological distribution of the paleozoic Hexacti- 
 nellida, and this is done in the table on page 16. With it is introduced 
 a grouping of the genera, which has in part been advanced by previous 
 writers. 
 
 The family Euretidse, into which many of these forms were originally 
 placed, has since been restricted by Schulze and von Lendenfeld to genera 
 agreeing more closely with Eurete. The family Pollakkke is probabl} 7 not 
 a natural one, and it is now a convenient receptacle for genera of doubtful 
 affinities. 
 
 Etydnoceras (= Dictyophyton, Hall) and allied genera are well char- 
 acterized by transverse and longitudinal bundles of long ra} 7 ed spicules, 
 which divide the surface into squares generally arranged in fours, and by 
 the absence, so far as known, of a fused dictyodermal layer. The 
 
BRACHIOSPONGIDiE. 
 
 15 
 
 parenchyma is also comparatively thin. Professor James Hall has 
 proposed the family Dictyospongidse to include genera possessing these 
 features. The group appears to be a well-founded and natural one, and 
 belongs to the suborder Lyssacina. In addition, Astylospongia was 
 removed from the Hexactinellida, by Zittel, in 1884, 1 so that it is now 
 significant that no paleozoic sponges are known, which can with certainty 
 be referred to the suborder Dictyonina. 
 
 This fact may have an important bearing upon the phylogeny of the 
 suborders. As stated by Schulze, either the Dictyonine forms have been 
 derived from the Lyssacine by fusion of the spicules, or both suborders 
 have been derived from a primitive form by gradual and parallel division. 
 At present, all the known paleozoic hexactinellids appear to belong to the 
 Lyssacina. The suborder is represented by numerous and diverse species, 
 and seems to have reached its maximum development previous to the 
 appearance of true Dictyonina. 
 
 There are several genera of fossils in the earliest formations, which 
 have as yet yielded but scant evidences as to their proper relations. 
 Among these may be mentioned the genera Calathium, Billings ; Trachyum, 
 Billings ; Spirocyathus, Hinde ; and Trichospongia, Billings. 
 
 A summary of the distribution of the genera as presented in the 
 following table shows that the rocks of the Cambrian (Taconic) system 
 have furnished two genera which have not been found in any of the 
 succeeding formations. From the Silurian, twelve genera are recorded, 
 seven of which are peculiar to the system, while four pass into the 
 Devonian, and one of these continues into the Carboniferous. The 
 Devonian has afforded but one genus peculiar to it, and the other four 
 cited are found also in earlier or later formations. 
 
 1 Neues Jakrbuck fur Mineralogie, etc., 1884 , Bd. ii, pp. 75 - 80 , Taf. i, ii. 
 
16 
 
 BRAOHIOSPONGIDJE. 
 
 Distribution of Paleozoic Hexactinellida. 
 
 Suborder Lyssacina, Zittel. 
 
 Family Dictyospongidje, Hall. 
 
 Protospongia, Salter 
 
 Cyathophycus, Walcott 
 
 Hydnoceras, Conrad 
 
 Ectenodictya, Hall 
 
 Plectoderma, Hinde 
 
 Lyrodictya, Hall 
 
 Thamnodictya, Hall 
 
 Phragmodictya, Hall 
 
 Cleodictya, Hall 
 
 Physospongia, Hall 
 
 Uphantaenia, Vanuxem 
 
 Family Brachiospongid^e, Beecher. 
 
 Brachiospongia, Marsh 
 
 Strobilospongia, Beecher 
 
 Family Pollakid^e, Marshall. 
 
 Hyalostelia, Zittel 
 
 Holasterella, Carter 
 
 Leptomites, Walcott 
 
 Astroconia, Sottas 
 
 Amphispongia, Salter 
 
 Acanthinella, Hinde 
 
 Family Monakid.e, Marshall. 
 
 Astrseospongia, JRcemer 
 
 Family Receptaculitid^e, Hinde. 
 
 Receptaculites, HeFrance 
 
 Ischadites, Murchison 
 
 Acanthoconia, Hinde 
 
 Cambrian. 
 
 (Taconic.) 
 
 Silurian. 
 
 Devonian. 
 
 Carbonifer- 
 
 ous. 
 
 
 • 
 
 • 
 
 • 
 
 • 
 
 — 
 
 • 
 
 • 
 
 • 
 
 — 
 
 m m 
 
 — 
 
 • 
 
 • 
 
 • 
 
 — 
 
 • 
 
 — 
 
 • 
 
 • 
 
 • 
 
 
 • 
 
 — 
 
 • 
 
 • 
 
 • 
 
 — 
 
 • 
 
 • 
 
 • 
 
 — 
 
 • 
 
 • 
 
 • 
 
 — 
 
 • 
 
 • 
 
 • 
 
 — 
 
 • 
 
 • 
 
 
 • 
 
 • 
 
 — . 
 
 • 
 
 
 
 ? 
 
 • 
 
 • 
 
 • 
 
 • 
 
 • 
 
 • 
 
 — 
 
 — 
 
 • 
 
 • 
 
 • 
 
 • 
 
 — 
 
 • 
 
 • 
 
 • 
 
 — 
 
 • 
 
 • 
 
 • 
 
 
 • 
 
 
 • 
 
 — 
 
 
 • 
 
 . 
 
 
 
 
 . 
 
 — 
 
 — 
 
 • 
 
 • 
 
 
 • 
 
 • 
 
 2 
 
 12 
 
 5 
 
 10 
 
BRACHIOS PONGID^E. 
 
 17 
 
 Variety and habitat of the Paleozoic Hexactinellida. 
 
 In a review of the paleozoic hexactinellid fauna, more than twenty 
 genera must be considered, all of which probably belong to the sub- 
 order Lyssacina. The diversity of form and degree of specialization in 
 these ancient sponges seem to be even greater than is exhibited among the 
 recent species. From such simple cup-shaped forms as Protospongia and 
 Cyathophycus, there is a regular gradation to the angular and nodose 
 Hydnoceras, with the closely allied genera Lyrodictya, Cleodictya, Physo- 
 spongia and others belonging to the family Dictyospongidee, including, 
 also, the wonderful basket disk of Uphantsenia. Hyalostelia represents 
 the enormous root spicules of an otherwise unknown form. Astrseospongia 
 has the form of a compact concavo-convex disk. Brachiospongia with its 
 fingered cup, and the elaborate Strobilospongia with its solid bundles of 
 anchor spicules, and, finally, the Receptaculites, with no recent repre- 
 sentatives either in form or structure, exhibit the wonderful develop- 
 ment of the Hexactinellida during paleozoic time. 
 
 Altogether, the facts point to the culmination of the Lyssacina at this 
 period. 
 
 The recent Hexactinellida are characteristic deep-sea types, ranging 
 from a depth of ninety-five fathoms to abyssal regions. In a study 
 of the paleozoic forms, it is interesting and important to determine, if 
 possible, the conditions under which these sponges flourished. Deep sea 
 investigation tends to show that the present abyssal oceanic areas and 
 continental masses are of great antiquity, and, consequently, that we can 
 have but limited areas and localized horizons representing deposition 
 below the 100-fathom line. Also, it has been shown that detrital sedimen- 
 tation almost ceases below 100 fathoms, and any organisms found in such 
 formations must have lived in areas of less depth. 
 
 The evidence furnished by observations on many of the paleozoic 
 hexactinellids shows that although occurring in groups of strata often con- 
 taining an abundant fauna, yet the particular beds, or layers, preserving 
 the sponges in their greatest perfection and abundance are comparatively 
 
 5 
 
18 
 
 BR ACHIOSPON GID AS. 
 
 barren of other organisms. In some cases, they must have thickly covered 
 the sea bottom. The Hydnoceras tuberosum from Steuben County, N. Y., 
 occurs in countless numbers, at several limited horizons, in a fine grained 
 and at times argillaceous sandstone. It is rarely associated with specimens 
 of crinoids and brachiopods. Hydnoceras prisma ticum occurs under similar 
 conditions. Cyathophycus reticulatus is found also in great numbers on the 
 slabs of Utica Slate, from Holland Patent, N. Y. In Bkachiospongia, 
 it is shown that the associated fossils include pelagic species and others 
 not incompatible with a deep water habitat. Barely are other organisms 
 associated with these sponges, although the conditions were evidently such 
 as would have preserved them. The occurrence, too, of these delicate lace 
 sponges in so great abundance and in an unbroken state shows that the 
 waters were tranquil, and that they were removed from active physical 
 changes. 
 
 It seems reasonable to suppose that the most characteristic and 
 permanent of the deep sea types were among the first to remove from the 
 littoral zone, and that they were the earliest to develop features adapted to 
 a deep sea existence. In many organisms, the retention of brilliant colors 
 and mimicral characters common among littoral forms shows that they have 
 migrated from the shores to a deep water habitat without developing into 
 perfect harmony with the environment, and that they are comparatively 
 recent additions to the deep water fauna. 
 
 On the contrary, many of the deep sea genera have lost their littoral 
 features, together with many morphological and chromatic specializations, 
 and are in accord with their surroundings. Hence, the species within the 
 present limits of these genera are strictly a development under deep water 
 conditions, and to-day have no littoral representatives. Only in remote 
 time, or ancient deposits, are related or ancestral forms found, exhibiting 
 a shallow water distribution. 
 
 The recent hexactinellid and lithistid sponges are among the most 
 extensive groups which are peculiar to the deeper zones. Their abundance 
 and great development in the Cretaceous are in accordance with other 
 faunal and lithological features, indicating the deep water origin of that 
 
BRACHIOSPON GID^E. 
 
 19 
 
 formation. From this, it is probable that, at an early period, these orders 
 flourished in the deeper regions of littoral deposits, and at greater depths. 
 
 While there is no strong- reason for considering- the deposits contain- 
 ing Brachiospongia as having been formed in very deep water, yet they 
 cannot well be considered as belonging to a littoral fauna. 
 
 Description of Species. 
 
 Brachiospongia digitata, Owen. 
 
 Plate I, figs. 1,2; PI. II, figs. 1-7 ; PI. Ill, figs. 1, 2 ; PI. IY, figs. 1-8. 
 
 Sponge broad cup-shaped, or short vasiform, witli a row of arms (from 
 eight to twelve observed) projecting outwards and downwards from the 
 periphery of the base. Osculum elliptical, with the diameters usually in 
 the ratio of two to three. Below the osculum, the walls of the cup, or 
 neck, are vertical, extending for a distance of from 25 to 40 mm., and 
 slightly expanding below to the origin of the arms. 
 
 Fig. I. 
 
 Figure 1 . — Brachiospongia digitata; base of the original specimen. One-half natural size. 
 
 Base of cup concave, usually with a strong conical or mammiform 
 projection near the center, which is the initial point of the sponge (fig. 1). 
 The gastral side of the base at the initial point shows several pits, of 
 which from three to five have been seen (PI. II, fig’. 7). 
 
20 
 
 BRACHIOSPON GIDJE. 
 
 The arms are distant from each other from one-fourth to one-fifth their 
 diameter at the base. Starting from the origin of the arms, they are nearly 
 circular in transverse section, and extend outward at right angles to the 
 sponge axis, or frequently upwards, and then are abruptly bent downwards, 
 flattened along their outer faces, and terminate in a compressed extremity 
 which is rarely bifid. When in a perfect condition, the arms are closed at 
 their distal ends. 
 
 The increase in the size of the arms is not altogether commensurate 
 with the growth of the sponge, and to compensate for the separation of the 
 arms, new ones are added from time to time. Plate I, figure 2, shows a 
 wide space on one side of the longest arm, and as this is about the 
 maximum size for nine-armed specimens, further growth would have 
 developed another arm at this point. A portion of a ten-armed specimen 
 is shown in Plate II, figure 3, in which a new arm has just begun to start 
 out from the cup, and now appears as a large node, or swelling, between 
 the widely separated and divergent adjacent arms. A still farther brachial 
 development is represented in figure 4 of the same plate. The specimen 
 has eight normal arms besides the central one in the figure, which is about 
 half grown. In the collections studied, comprising twenty-one nearly 
 entire specimens, the number of rays varies, with some exceptions, accord- 
 ing to the size of the individual, and is from eight to twelve. Further 
 research will probably extend these limits. 
 
 In the following table, the grouping is based upon the number of 
 arms. It shows the variation in size, and enumerates the specimens in 
 each group. 
 
 Number of 
 
 Diameter of smallest and largest specimens, 
 
 * Number of 
 
 arms. 
 
 in inches. 
 
 specimens. 
 
 8 
 
 9± 
 
 1 
 
 9 
 
 34-64 
 
 4 
 
 10 
 
 3f— 9 
 
 9 
 
 11 
 
 5f-ll 
 
 4 
 
 12 
 
 6-10i 
 
 3 
 
 One of the individuals here indicated as having ten arms is 5^ inches 
 in diameter, and shows in addition to the ten fully developed rays, three 
 
BRACHIOSPONGIDJE. 
 
 21 
 
 nodes, and two branched arms, all spirally arranged, beginning at the 
 initial point, and extending to the oscnlum of the sponge. A ten-rayed 
 example, having a diameter of 6| inches, presents ten normal arms, and 
 one node representing an undeveloped arm (PL II, fig. 3). 
 
 The accompanying wood-cut, figure 2, illustrates an abnormal speci- 
 men, and shows very clearly that, for specific distinctions, little dependence 
 can be placed upon the number and character of the arms. On one side 
 of the specimen, they are normally developed, while, on the other, they 
 are bifurcate and covered with processes and large rounded protuberances. 
 
 Fig. 2. 
 
 Figure 2 . — Brachiospongia cligitata ; abnormal specimen. One-half natural size. 
 
 From what is shown regarding the growth and increase of the arms, 
 it is evident that this feature is not of sufficient importance to be used 
 as a basis for specific separation. With but two or three specimens at 
 hand representing the extremes of variation, these differences seem to be 
 important. Having numerous individuals, however, representing different 
 stages and conditions of growth, we are able to connect the extremes, and, 
 at the same time, show the wide extent of this mutable and mirable species. 
 
 The gastral cavity is very large, occupying the whole interior of the 
 cup, and extending nearly to the distal extremities of the arms. 
 
 6 
 
22 
 
 BRACHIOSPONGIDiE. 
 
 Some specimens which were freed from the rock without the use of 
 acids show that the extremities of several of the arms, and the basal boss, 
 were worn through to the gastral cavity before they were imbedded. But 
 the present imperfect condition of many of the weathered specimens is due 
 principally to breakage, and to the non-silicification of portions of the 
 sponge walls. 
 
 Dimensions . — The smallest example observed (PI. II, figs. 5, 6, 7) has 
 nine arms, and measures 45 mm. in height. This is somewhat less than 
 the original height, as the neck is imperfect, and the extremities of the 
 arms broken off. The greatest diameter is 80 mm., and the arms have a 
 diameter of about 16 mm. at their bases. Two of the largest individuals, 
 preserving their full extent, measure about 275 and 330 mm. respectively, 
 in diameter. 
 
 A restoration from fragments in the collection, based upon the propor- 
 tions presented by entire specimens, indicates that the largest individuals 
 of this species reached a diameter of about 400 mm. 
 
 Brachiospongia is among the largest of the fossil Hexactinellida, and 
 ranks in this respect with Uplian taenia and Cleodictya, although they are 
 all somewhat exceeded by the recent genera Pheronema and Poliopogon. 
 
 It is interesting to note that in the cavities of Brachiospongia and in 
 the surrounding rock are numerous small spherical bodies which suggest 
 the statement made by Schulze, 1 that in the cavities of Poliopogon amadou, 
 were small spherical sponges about 3 mm. in diameter, which he considered 
 as the young of that species. In the present instance, however, these 
 bodies vary in size from 1.5 to 10 mm., and show a structure which is 
 characteristic of Hindia, and the species has been named Hindia parva , by 
 Mr. E. O. Ulrich. 
 
 A specimen of Brachiospongia (Plate III, fig. 1), found by W. M. 
 Linney in the northern part of Spencer County, Kentucky, in strata of the 
 Middle Hudson series, offers some points of difference with those from 
 Franklin County. It is preserved in mudstone, and the parenchyma of the 
 sponge has been replaced by calcite. The specimen measures 235 mm. 
 
 1 Rep. on the Hexactinellida, p. 257. 
 
BRACHIOSPON GID^E. 
 
 23 
 
 in diameter, and has eight arms, whicli are constricted at their origin, 
 directed outwards and downwards at an angle of forty-five degrees, and 
 are not geniculated as in typical B. digitata. The osculum is subcircular, 
 and the neck is campanulate below. The cup, or body, of the sponge is 
 comparatively small. The base is flat, and without the initial projection 
 usually present. Were it not for the great range of variation shown in the 
 specimens from Franklin County, it would seem that this form represented 
 a distinct species. The differences are probably due to changed physical 
 conditions. 
 
 Spicular Skeleton. — Rod-like spicules were detected in the matrix 
 adjacent to the dermalia, in some of the sections made for microscopic 
 study. They were without apparent regularity, and probably correspond to 
 the uncinates or raphides in the dermal membranes of many recent forms. 
 As their relations are uncertain, they have been omitted in the diagram- 
 matic restoration given in Plate IV, figure 8. 
 
 The surface of the sponge when well preserved presents a minutely 
 papillose appearance to the unassisted eye. The papillse are of unequal 
 size, irregularly arranged, and distant 1 mm., or less. When magnified, 
 they are seen to be produced by an elevation of the dermal membrane, 
 and by five strong conical nodes occupying the center. It is shown that 
 the latter belong to large free hypodermal pentacts. 
 
 The dermal membrane (PI. IV, fig. 1) is composed of an exceedingly 
 fine and irregular quadrate mesh of four-rayed spicules, apparently fused, 
 with their centers distant from each other from .1 to .2 mm. Occasionally, 
 larger spicules occur, which seem, with the expansion of the surface due 
 to the growth of the sponge, to develop eventually into the large pentacts 
 of the hypoderm. 
 
 The hypodermalia are less regularly arranged than the dermalia, and 
 the rays overlap and cross each other in various directions. The proximal 
 rays have been observed to penetrate the parenchyma to the depth of 2.5 
 mm., and the tangential rays reach a length of more than 1 mm. 
 
 The parenchyma varies in thickness with the size of the specimen, 
 and in different portions of the same individual. It is thinnest near the 
 
24 
 
 BRACHIOSPONGIDiE. 
 
 extremities of the arms and at the edge of the cup. A specimen 80 mm. in 
 greatest diameter has a parenchymal thickness of 3 mm. near the ends of 
 the arms and edge of the cup, and reaches a thickness of nearly 5 mm. 
 near the origin of the arms. The walls in the central portion of an arm 
 52 mm. in diameter are 10 mm. in thickness laterally, and somewhat 
 thinner on the distal and proximal sides. A large arm nearly 60 mm. in 
 transverse diameter shows a thickness of 13 mm. 
 
 The accompanying wood-cuts (figures 3 and 4) represent actual 
 sections through two arms, and show the thickness of the parenchyma, the 
 irregular canals, and the diameter of the gastral cavity, in each. The 
 lower edge is the flattened outer face of the arm. 
 
 Fig. 3. Fig. 4. 
 
 Figure 3 . — Brachiospongia digitata ; section across middle of arm. 
 
 Figure 4. — B. digitata; section across lower third of arm, from larger specimen. 
 Both figures are natural size. 
 
 The parenchymalia seem to be all hexacts, exhibiting great variations 
 in size and in the development of the rays. Many of them are of about 
 the size of the hypodermalia, but much larger occur. Their principal 
 axes are usually parallel or vertical to the adjacent surface of the sponge, 
 and frequently one or more rays are greatly developed, and seem to bear 
 spines or lateral processes. In one broken section, a single ray may be 
 traced for a distance of 10 mm. Other weathered specimens show many 
 similar elongate rays traversing the parenchyma. In rare instances, the 
 axial canals are preserved, and appear to have been about one-third the 
 diameter of the rays. 
 
 No evidence is shown on the exterior of the sponge of the presence of 
 the interior canals or afferent pores, but on the gastral surface of the walls 
 
BRACHIOSPON GIDiE. 
 
 25 
 
 are numerous oscula between the rays of the spicules, often radially 
 arranged, and with larger circular depressed areas 2 to 3 mm. in diameter 
 marking the terminations of the principal efferent canals. The interior, 
 represented in horizontal section, in figures 3 and 4 of Plate IV, exhibits 
 irregular cavities filled with crystalline quartz, and cylindrical vertical 
 canals having a diameter of from 1 to 2 mm. The gastral ends of the 
 large canals are covered with a membrane in which no spicules can be 
 detected. It is perforated by several efferent apertures, or oscula. Also, 
 the walls of the entire canal system appear to have had a limiting mem- 
 brane, probably spicular originally, but now preserved as rusty markings 
 in the rock. 
 
 The gastral membrane does not seem to be as well preserved as the 
 dermal, nor are the spicules so well differentiated. It is covered with a 
 layer of pentacts which are irregularly disposed, and not fused into a 
 connected mesh. The largest pentacts measure about 2 mm. across, and 
 interspersed with them are numerous smaller spicules. The proximal rays 
 of the larger pentacts extend into the parenchyma, as represented in 
 Plate IV, figure 7. 
 
 Summary of spicular ancl internal characters . — Dermalia constituting a 
 continuous, minute, irregular, quadrate mesh of four-rayed spicules. Hypo- 
 dermalia consisting of larger immersed pentacts with an elevated exsert 
 node at the base of each tangential ray, and a similar central node repre- 
 senting the atrophied distal ray ; proximal ray penetrating the parenchyma. 
 
 Parenchymalia comprising large and small hexacts with modified 
 rays, which are sometimes spiniferous. Gastral membrane with numerous, 
 irregularly arranged, large and small pentacts and tetracts, which are appa- 
 rently free ; proximal rays long-, penetrating the tissues. Dermal pores 
 not defined ; gastral pores irregularly circular or oval, sometimes disposed 
 in a radial arrangement. Gastral surface marked by circular depressions 
 covered by a perforated membrane, and marking the terminations of the 
 main interior efferent canals. Parenchyma penetrated by numerous 
 irregular canals and by vertical cylindrical canals. 
 
 7 
 
26 
 
 BRACHIOSPONGID^E. 
 
 Geographical Distribution . — The specimen found by Dr. Troost is 
 recorded as from Davidson County, Tennessee. Others have been collected 
 in Mercer and Spencer Counties, Kentucky. The most prolific localities 
 known are in Franklin County, Ky., on Benson Creek and its tributaries, 
 northwest of Frankfort, and on Cedar Run, two and one-half miles south. 
 
 Strobilospongia tuberosa, sp. nov. 
 
 Plate V, fig. 5 ; Plate VI, figs. 3-7. 
 
 The form of the sponge is somewhat conical, flattened, and deeply 
 indented longitudinally on opposite sides, from the base to the osculum. 
 Otherwise, the surface is covered with slightly pendent, solid, tuberose 
 extensions of the parenchyma. ~ Summit flattened. 
 
 Osculum irregular, sinuous ; margin thin, sometimes convoluted. 
 Base broad, giving a truncate appearance to the cup ; penetrated by a 
 large mass of basalia, or root tuft of anchor spicules, extending into the 
 interior nearly one-half the height of the parenchymal portion of the 
 sponge. In one example, the bundle of anchoring spicules, where it 
 emerges from the base, has a diameter of more than one-third that of the 
 body of the sponge above. 
 
 The principal specimen (PI. VI, figs. 3, 4, 5) has a height without the 
 root tufts of nearly 80 mm., and the greatest diameter across the base is 
 110 mm. It is remarkable as preserving a bud which is still attached to 
 the parent mass, and nearly equals it in size. The root of the bud is a 
 little smaller, and the lobation of the surface is considerably sharper and 
 more distinct. A smaller individual 70 mm. high is very irregularly formed. 
 One side is lobed as in the preceding example, and the other is marked by 
 two deep irregular pits. A vertical section was made through this example 
 revealing the extent of the mass of basalia and of the gastral cavity. 
 It is represented in Plate VI, figure 6. 
 
 The specimens are so thoroughly silicified that the spicular structure 
 has been to a large degree obliterated. The cruciform ends of large 
 hexactinellid spicules can be seen on some portions of the surface, resem- 
 bling those in similarly preserved specimens of Brachiospongia digitata. 
 
BRACIIIOSPONGIDiE. 
 
 27 
 
 Smaller dermal spicules and traces of a spicular mesh can likewise be 
 detected. The parenchymal canals are nearly vertical to the surface, and 
 measure .6 mm. in diameter. The basalia are closely aggregated, and 
 usually form a twisted rope-like bundle. The separate spicules are very 
 slender, not measuring more than .02 mm. in diameter. The axial canals 
 are occasionally preserved, and are represented in Plate VI, figure 2. 
 
 The source of the two specimens of the species here described is quite 
 uncertain. They were received accompanied by labels referring them to 
 the Hudson group, from Turner’s Station, in northern Kentucky. The 
 condition of preservation and enclosing matrix bear much resemblance 
 to some specimens of Brachiospongia, whose location is known. The 
 presence, too, of another species of the same genus, and of so many 
 similar masses of basalia in the Brachiosponge bed, serve to point to this 
 horizon as the probable source of the species. 
 
 On Plate V, figs. 2, 3, 4, three masses of basalia are shown, represent- 
 ing their general features. The microscopic structure is delineated in 
 Plate V, fig. 5, a, b, and Plate VI, figs. 1, 2. Ropes of spicules, both larger 
 and smaller than those illustrated, were collected. One example has a 
 diameter of 110 mm., and others measure that number of millimeters in 
 length. These numerous root tufts indicate imperfectly the richness of 
 the original sponge fauna represented in these sediments. In an area of 
 not more than two square rods, on Cedar Run, about forty specimens 
 were obtained on the surface of the ground. They are also common near 
 Bridgeport, Franklin County, and at several places on Benson Creek. 
 
 Similar masses of basalia have been called Tricholites, but it is obvious 
 that several slightly related forms of sponges may have anchoring bundles 
 of spicules, which are undistinguishable from each other, and that specific, 
 or even generic characters are seldom impressed on the root tufts. The 
 genera Leptomites, Hyalostelia, and Acestra, represent varieties of basalia 
 for which the sponge proper has not been discovered. Recently, root 
 spicules were described for the genus Protospongia, by Sir Wm. Dawson 
 and Dr. Geo. J. Hinde, 1 but they do not form a massive bundle. 
 
 1 Canadian Record of Science, Montreal : April, 1S88. 
 
28 
 
 BRAC H10SP0N GID^E. 
 
 Distribution . — The two specimens of 8. tuberosa bear the label 
 “ Turner’s Station, Ky.” Numerous masses of basalia have been found 
 in and above the Brachiosponge bed on Benson Creek Cedar Run, at 
 Frankfort, Franklin County, and at Lexington, Fayette County, Kentucky. 
 Under the name Tricholites typicalis, Mr. E. 0. Ulrich has catalogued 
 similar bundles of spicules from the tops of the hills at Cincinnati, Ohio. 
 
 Steobilospongia aueita, sp. nov. 
 
 Plate V, tig. 1. 
 
 The only example of this species known is a compressed cup-shaped 
 specimen found associated with Beachiospongia, on Benson Creek. The 
 summit and base are both imperfect, and the special features of these por- 
 tions cannot be described. There is a flattened, depressed area down the 
 center of the sides, which is without ornamentation. The specimen pre- 
 serves five concentric imbricating rows of auriculate pendent extensions of 
 the parenchyma, some of which are so closely oppressed to the cup that 
 the underlying tissues are excavated for their reception. In general, these 
 expansions appear to be solid, but several show that the gastral cavity 
 extends into them for a short distance. 
 
 The microscopic structure is no better preserved than in the pre- 
 ceding species, and the spicular skeletons of both agree in all the features 
 which can be observed. 
 
 The specimen lias a height of 136 mm., and a maximum diameter 
 through the center of 138 mm. The latter dimension is somewhat greater 
 than normal, on account of the compressed condition of the sponge. It is 
 completely silicified, and preserved in a condition similar to the majority 
 of the specimens of Beachiospongia. 
 
 As compared with S. tuberosa , this species reaches a greater size and 
 is more robust, the parenchymal expansions are much larger, pointed 
 directly downwards, and more regularly arranged in concentric rows. 
 
 Distribution . — From the Brachiosponge bed on Benson Creek, Franklin 
 County, Kentucky. 
 
PLATE I. 
 
PLATE I. 
 
 Brachiospongia digitata. 
 
 Page 19. 
 
 Fig. 1 . — Original specimen ; lateral view. Natural size. 
 
 Fig. 2. — The same ; seen from above. Natural size. Davidson County , Term. 
 Original in Yale University Museum. 
 
F. Berger, del. 
 
 L. Schierholz, lith. 
 
 BRACHIOSPONGIDAl . 
 
 PLATE I 
 

PLATE II. 
 
PLATE II. 
 
 Brachiospongia digitata. 
 
 Page 19. 
 
 Fig. 1.— Side view ; showing full extent of arms. 
 
 Fig. 2. — The same ; showing section through cup, and openings into gastral cavities of 
 arms. Natural size,. 
 
 Fig. 3. — Side view of imperfect specimen ; showing initial point in growth of one arm, 
 which here appears as a swelling, or node. One-half natural size. 
 
 Fig. 4. — Another specimen ; exhibiting one arm apparently about half grown. One-half 
 
 natural size. 
 
 Figs. 5, 6, 7. — Lateral, bottom, and top views of smallest specimen observed. Natural 
 size. 
 
 Brachiosponge bed. Franklin County. Ky. 
 
 Originals in Yale University Museum. 
 
BRACHIOSFONGIDAI 
 
 PLATE II 
 
 3 
 
i v j,A r rK hi. 
 
 9 
 
PLA TJE III. 
 
 Brachiospostgia digitata. 
 
 Page 19. 
 
 Fig. 1 . — Top view of eight-armed specimen mentioned on page 22. The parenchyma has 
 been partially exfoliated, and the cast of the interior is exposed in several 
 places. One-half natural size. Middle Hudson. Spencer County , Ky. 
 
 Fig. 2. — Lower side of large twelve-armed specimen, in the collection at the State 
 Museum, Frankfort, Ky. One-half natural size. Braehiosponge bed. 
 Franklin County, Ky. 
 
 Original of figure 1 in Yale University Museum. 
 
2#(S 
 
 BPACHIOSPONGIDAI 
 
 PLATE III 
 
 F. Berger, del. 
 
 L. Schiertiolz, lith. 
 
PLATE IV. 
 
PLATE IV. 
 
 Brachiospongia digitata. 
 
 Page 23. 
 
 Fig. 1 . — Outer surface of sponge ; showing dermal spicular mesh and exsert nodes on 
 rays of hypodermal pentacts. X 20. 
 
 Fig. 2. — Section representing large hypodermalia after removal of dermal mesh, x 20. 
 
 Fig. 3. — Horizontal section of interior of parenchyma. X 20. 
 
 Fig. 4. — Horizontal natural section from broken specimen, x 4. 
 
 Fig. 5. — Subgastral skeleton ; showing large spicules and ends of efferent canal system. 
 
 The circular depressions represent ends of vertical cylindrical canals. X 10. 
 
 Fig. 6. — Inner, or gastral, surface. X 20. 
 
 Fig. 1 . — Vertical section through parenchyma with small portion of adhering matrix on 
 dermal side, x 20. 
 
 Fig. 8. — Restored vertical section, combining characters represented in the preceding 
 figures, which were drawn from actual specimens. X 10. 
 
 Originals in Yale University Museum. 
 
BRACHIOSPONGID/B . 
 
 PLATE IV. 
 
 C.E. Beecher, del. 
 
 L. Schierhoiz, lim. 
 
PLATE V. 
 
 10 
 
PLATE V. 
 
 Strobilospongia aurita. 
 
 Page 28. 
 
 Fig. 1 . — Lateral view of type specimen. Natural size. Brachiosponge bed. Franklin 
 County , Iiy. 
 
 Basalia. 
 
 Page 27. 
 
 Figs. 2, 3, 4. — Three masses of anchor spicules presenting characteristic features. 
 Natural size. Brachiosponge bed. Franklin County , Ky. 
 
 Fig. 5. — a. Longitudinal section of small portion ; showing size and disposition of 
 basalia. X 25. 
 
 b. Transverse section from same specimen, x 25. 
 
 c. Section of small fragment of parenchyma of Strobilospongia tuber osa ; 
 
 showing two imperfectly preserved hexactinellid spicules, x 25. 
 
 Originals in Yale University Museum. 
 

 BRACHIOSPONGIDiR, . 
 
 PLATE V 
 
 F. Berger, del 
 
 L. Schier’aolz, lith. 
 

 
PLATE V I . 
 
PLATE VI. 
 
 B ASALIA. 
 
 Page 27. 
 
 Fig. 1. — Longitudinal section of fragment. X 7o. 
 
 Fig. 2. — Transverse section ; showing diameter of spicules and axial canals. X 70. 
 Brachiosponge bed. Franklin County , Ky. 
 
 Strobilospongia TUBEROSA. 
 
 Page 26. 
 
 Fig. 3. — Lateral view ; showing parent sponge at the right, and attached bud on the left 
 with its separate bundle of basalia. Natural size. 
 
 Fig. 4. — Basal view of same specimen. 
 
 Fig. 5.— Side view of bud ; showing strong lobation of surface. 
 
 Fig. 6. — Small example, cut longitudinally through the center ; exhibiting extent of 
 bundle of basalia, gastral cavity, and variable thickness of parenchyma. 
 
 Natural size. 
 
 Fig. 7. — The same ; showing external form of sponge, and narrow and sinuous aperture 
 at summit. 
 
 Both specimens of this species are recorded as from the Hudson group at 
 Turner’s Station, Ky. 
 
 Originals in Yale University Museum. 
 
B R AC HI O S P O NG IDTL 
 
 PLATE VL 
 
 F. Berger, del. 
 
 L. Schierholz, lith. 
 
 X70 
 
 6 
 
 X70 
 
Date Due 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Form 335— 15M— 7-36— S