ON THE NATURE OF EDESTUS AND RELATED GENERA,WITH DESCRIPTIONS OF ONE NEW GENUS ANDTHREE NEW SPECIES. By Oliver P. Hay,Of Washington, District of Columbia. 1. DESCRIPTIONS OF THREE NEW SPECIES OF EDESTUSEDESTUS CRENULATUS, new species.The type and only known specimen of this species belongs to theU. S. National Museum (Cat. No. 6050), and was found in a col-lection of fossils purchased from Mr. G. Hambach, of St. Louis,Missouri. No record regarding the origin of the fossil accompaniedit, but there can be no doubt that the specimen had been obtainedfrom the Coal Measures not far from St. Louis, probably from someof the coal mines of western Illinois. No species of the genus ismentioned in Mr. Hambach 's Preliminary Catalogue of the FossilsOccurring in Missouri.''The specimen (pi. 12, fig. 1) is almost as complete as it was onthe death of the animal, only the apices of some of the teeth andsome of the denticles being broken off and missing. The speciesresembles most that known as Edestus heinrichU,^ but numerous dif-ferences may be observed.The total length of the fossil is 207 mm.; the greatest height is58 mm. ; but to the latter measurement should be added about 2 mm.for the missing apex of the second tooth, counting from the right.The height of the shaft alone is 46 mm., the greatest thickness, be-neath the first tooth, 28 mm. As will be seen from fig. 1, pi. 12, the ? Geological Survey Missouri, Bull. No. 1, 1890, pp. 60-85.* Doctor Eastman (Bull. Mus. Comp. Zool., vol. 39, 1902, p. 05) points out thatNevpberry's specific name heinrichsii was improperly formed. Inasmuch asNewberry expressly says that the species was named for Mr. Heinrich. we mayassume that the foi'm heinrichsii was a lapsus calami, and on that groundadopt the form heinrHchii. Newberry himself used this form in 1879 (Geo-logical Survey Indiana, p. 347), although later he used the original spelling.Proceedings U. S. National Museum, Vol. 37?No. 1699. 43 44 PROCEEDINGS OF THE KATIOXAL MUSEUM. vol. 37.tooth-bearing border is arched, while the opposite border is slightlyconcave in each direction from near the middle of the length. Atransverse section (text fig. 1) taken between thefirst and second teeth, connting from the right,shows that the lower border is here broadlyrounded, while a section (text fig. 2) across theliinder half of the fossil shows that there thelower border is acute. Beginning just in frontof the first tooth a sharj:) and narrow grooveruns along the midline of the specimen, widen-ing at the front end, then contracting and run-ning backward on the lower side (text fig. 1, 7)to near the middle of the length. , ?' Along the upper border of the fossil are sixFig. 1. ? Edestus cren- & i iuLATus, XI. TYPE, compresscd teeth, the anterior and posterior.VERTICAL SECTION BE- jjordcrs of whlch are furnished with denticles.TWEEN THE FIRST ANDSECOND TEETH. 2. Of tliesc dcnticlcs there are 8 or 9 on the anteriorSECOND sheath; 2\ Q(\frQ and perhaps 1 or 2 fewer on the hinderENAMELED PROCESS OF '^ ^ iSECOND tooth; 3, edge. Each of these denticles is minutely cren-In^r^ ?L/"'"'' "lated. Each tooth is covered by a layer oftooth ; 4, sheath of ? "^FOURTH tooth ; 5, enamel, and at the base of the tooth a processSHEATH OF FIFTH ^f ^j^j^ euamcl ruus forward beneath the hindertooth ; G, UPPER BOR-DER OF SHEATH OF third of tlic ucxt tooth inSIXTH tooth; 7, j^^|^.?-,(.p ^he followingGROOVE along LOWER _ _ . *^BORDER OP ANTERIOR figurcs giYQ tlic dimcnsiousHALF OF SHAFT. ^f ^J^^ ^^^^^^ rj^j^^ ^^^^^^^ "^taken from the point where the tooth joins theone in front to its contact Avith the next behind.The height is from the apex to the lower borderof the enamel, along the axis. Tooth, 1 ; lengthof base, 26 mm.; height, 19? mm.: tooth, 2;length of base, 27 mm.; height, 19? mm.: tooth,8; length of base, 27.5 mm.; height, 19 mm.:tooth, 4; length of base. 28 nun.; height, 19? fig. s.-edestus cren-1 ^ ?, ^ . ^ ULATUS, X 1. TYPE.mm. : tooth, 5 ; length of base, 30 mm. ; height, vertical section be-19? mm.: tooth, 6; length of base, 30.5 mm.; ^ween third and... ' <=> FOURTH teeth. 3,lieigut, lo? mm. hinder end of thirdThe thickness of most of the teeth at the lower ^ooth ; 4, sheath , J. , . . mi r. -Til ?''' FOURTH tooth; 4^end or the axis is 8 mm. 1 he first one is slightly enameled process ofless ; the last one is only 6 mm. thick. The angle fourth tooth ; r..between the anterior and posterior borders of tooth ; 6. upper bor-any tooth is very close to 90?. It will be observed ^i=? '^^ sheath ofthat the hinder tooth descends to the lowerborder of the shaft. The surfaces of all the teeth are smooth. Anumber of cracks in the enamel pass from the base of each tooth tothe apex, and these seem to follow slight ridges. NO. 1699. ON EDE8TU8 AND RELATED GENERA?HAY. 45The body of the fossil is made up of a shaft of vasodentine, denseand rough on the surface and without enamel. As has been shown byother observers, in the case of other species of the genus, this shaftconsists of trough-shaped processes, one of which runs forward fromeach tooth, and which supports and partly incloses the process of thepreceding tooth and is supported and partly inclosed by the succeed-ing one. The photograph (pi. 12, fig. 1) shows the grooves limitingeach of these processes above and below. It will be seen that measur-ing each process, or sheath, from the hinder end of the tooth which itsupports to its anterior end, the first one is 84 mm. long, the second105 mm., the third 125 mm., the fourth 138 mm., the fifth 149 mm.,the sixth 158 mm. This means that the anterior end of each sheathreceded from the end of the shaft by considerably less than thelength of the tooth to which it belonged.This species appears to differ from E. heinrichii in various j^articu-lars, some of which will be mentioned.1. The form of the shaft is different in the two species. In E.cremdatus the greatest height is in front of the middle of the lengthand under the second tooth; in E. heinriclin it is behind the middleand under the fourth tooth. It is possible that in the original of thefigures of the last of these" the last-formed tooth is missing, butallowance for this would put the greatest height at the middle of theshaft. The shaft of E. crenidatus is relatively pointed in front; thatof E. heinrichii is deep and truncated. If in the figured specimen ofE. heinrichii the last-formed tooth is missing, a portion of its enamelought to show under the last one present. If no tooth is missing, theshaft terminated in quite a different manner from that of E. crenu-lafvs. Furthermore, in case the last-formed tooth of that specimenof E. heinrichii is missing there would have been present 9 teeth ; inthe type of E. crenidatus there are only 6. The type of E. heinrichiiis a considerably larger specimen (280 mm. long) than that of E.crenulatus. It is difficult to see how the latter could become modifiedso as to resemble the former.2. It will be noted the last sheath of E. crenidatus extends muchfarther forAvard than does that of E. heinrichii; also, it covers rela-tively less of the depth of the shaft.3. In E. crenulatus a tongue-like process of the unenameled surfaceof the shaft runs backward between each tooth and the pointedprocess of enamel of the next tooth behind until it touches this hindertooth ; in E. heinrichii the tongue-like process is cut off from thehinder tooth by a prolongation of the anterior one.4. The teeth of the two species are of different forms. In E. crenu-latus the angle between the two lines carried from the apex of the "Geol. Surv. 111., vol. 4, pi. 1, fig. 1; Aun.,N. Y, Acad. Sci., vol. 4, pi. 5, fig. 2. 46 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. 37.tooth to the ends of the anterior and posterior cutting edges respec-tively is close to a right angle; in E. heinrichii the angle measuresabout 75?. The height of the teeth of E. heiiirichii is six-tenths thelength of the base, while in E. cremdatus it is seven-tenths. Never-theless, the teeth of E. heinrichii appear more acute than those ofE. creniddtus because of the concavity of the cutting borders, thoseof the latter species being nearly straight. The last tooth ofE. heinrichii seems to be less elevated than the others. Judgingfrom the apical angle and the straight cutting edges of the tooth, itis believed that Newberry's figure 2b " represents a specimen ofE. crenulahis.It has been mentioned that there is in the specimen here described anarrow groove that runs forward from the first tooth, becomingbroader at the end of the shaft. Immediateh^ in front of this firsttooth there is a rough surface just like that found bordering theenamel of each of the teeth. These facts suggest that some of theanterior and earliest formed teeth are now missing, that long beforethe animal died some teeth had dropped off. ' Certainly it is not prob-able that the animal became adult before it developed any teeth.There must have been a series of teeth from very small ones up to thefirst tooth now present. Nevertheless the writer does not believe thatthe shaft ever extended any considerable distance, if at all, in front ofAvhat is now its anterior end. No doubt the trough-like processes orsheaths of these earliest teeth, like those of the ones present, shortenedrapidly toward the front, so that they probably never extendedbeyond the shaft as we now have it. Dr. Charles R. Eastman ^ hasfigured a series of three teeth which diminish rapidly in size, and inwhich the sheath of the smallest does not reach beyond the succeedingone.'' Doubtless still smaller teeth occupied the space in front of thesmallest one present. The significance then of the groove found infront of the first tooth of the type of E. crenulatiis is found in theexistence there, in the eacj,^' youth of the animal, of a series of smallteeth that were shed perhaps earl}' in life.Newberry '^ appears to have regarded the separate segment repre-sented by his fig. 2?, pi. 5, as that of a young animal. Its sheath is,however, too long for this. It nnist have been one of the later teeth.It seems possible that on sufficient maceration all the teeth, with their ? Ann. N. Y. Acad., vol. 4, pi. 5 ; Pal. Fishes N. A., pi. 39.? Bull. Mus. Couip. Zool., vol. 39, p. 76, fig. 7. '^ In case the reduction of Eastman's figure is really one-half, it seems possiblethat his specimen belongs to an uudescribed species. Otherwise great variationin size of teeth in E. heinrichii is indicated. The length of the anterior teethof the type is only about 30 mm., whereas the largest tooth figured by Eastmanhas a length of 37 mm. ?^Ann. N. Y. Acad., vol. 4, p. 122. NO. 1699. Oy EDESTUS AND RELATED GENERA?HAY. 47 sheaths, even of old individuals, might have separated from one an-other. The t3'pe of E. I'arpinslii lateh^ described by Missuna ? isevidently a segment produced by an adult indi-vidual and freed from all its predecessors.At a short distance behind its anterior endeach sheath of E. crenidatus is divided on themidline below into right and left portions, thelower edges of which run forward and upward.This is the origin of the groove which occupiesthe lower border of the front of the shaft. Itis j^robable that in this groove at the point ofsej)aration of the lateral portions of the sheaththe nerves and blood vessels entered and ex-tended backward into each sheath and tooth.It is, of course, possible that another toothwould have been produced behind the sixth ofthe type of E. crenidatus ; but, in as much as thatlast tooth is considerabl}" thinner than its prede-cessors, it is not unlikely that old age and de-cadence of powers had come on and that no moret^eth would have been developed.EDESTUS SERRATUS, new species.The history of the type of this species isexactly that of the specimen described as E.crenulatus. It is Cat. No. G049 in the U. S.The length of the fossil (pi. 12, fig. 4), as found, is 150 mm. Thesixth and last tooth of the series had been brokenoff before the time of entombment. If an allow-ance of 27 mm. is made for this missing part thewhole length will be 177 mm. The greatestheight, from the apex of the second tooth down-ward, is 51 mm. The greatest height of the shaft,between the second and third teeth, is 35 mm.Its greatest thickness is 21 mm. ; that at the rearof the fifth tooth is 9 mm. The form of the shaftand sections of it (text figs. 3, 4) resemble thoseof E. crenidatus^ but the hinder section, thoughnot so high as that of E. crenulatus^ is fully asbroad. The last sheath occupies two-thirds ofthe width of the shaft. The anterior half of theshaft is rough, with close-set patches of enamel,irregular in size and form. The hinder halfalso is rough for some distance below the teeth,but most of the surface is nearly smooth. It seems possible that thehinder part of the shaft had been embraced by a sheath that had not Fig. 3. ? Edestus serr>-TUS, Xl. TYPE. SEC-TION BETWEEN FIRSTAND SECOND TEETH.1, HINDER END OFFIRST TOOTH ; 2,ENAMELED PROCESS OFSECOND TOOTH ; 3,SHEATH OF THIRDTOOTH ; 4, SHEATH OFFOURTH TOOTH ; 5,SHEATH OF FIFTHTOOTH ; 6, UPPER BOR-DER OF SHEATH OFSIXTH TOOTH ; 7,GROOVE ALONG LOWERBORDER OF ANTERIORHALF OF SHAFT.National Museum. Fig. 4. ? Edestus serra-tds, x 1. type. sec-tion between thirdand fourth teeth.3, hinder end ofthird tooth ; 4,enameled process offourth tooth ; 5,sheath of fifthtooth ; 6, upper bor-der of sheath ofsixth tooth. * Bull. Soc. Imp. Nat. Moscow, vol. 21, 1908. p. 528. 48 PROCEEDINGS OF THE XATIOXAL MUSEUM. vol. 37.become consolidated, and which was lost when the specimen becameinterred. The point in the lower border from which the two con-cavities def)art is directly opposite the space between the second andthird teeth. In E. crenulatus it is opposite the apex of the secondtooth.The following- are the dimensions of the teeth of this specimen:Tooth, 1 ; length of base, 22 mm. ; thickness, 8 mm. : tooth, 2 ; lengthof base, 24 mm. ; height, 10 mm. ; thickness, 8 mm. : tooth, 3 ; length ofbase, 25 mm. ; height, 19 mm. ; thickness, 8 mm. : tooth, 4 ; lengthof base, 26 mm. ; height, 19 mm. ; thickness, 8 mm. : tooth, 5 ; length ofbase, 27 mm. ; height, 19 mm. ; thickness, 8 mm.The apical angle of the teeth is 80?. That of E. minor is 35? or40? ; that of E. crenulatus., as already mentioned, is 90?. The an-terior edge of the teeth is concave; the posterior is nearl}^ straight.The teeth are furnished with denticles, but the edges of these areperfectly smooth. The surfaces of the teeth are enameled. Fromthe base of each tooth a number of sharply defined and frequentlyanastomosing ridges rise to the apex. The forwardly directed proc-ess of each tooth is brought up close to the base of the next tooth infront.In front of the first tooth there is, as in E. crenulatus^ a groovethat was originally occupied by a series of earlier-formed teeth. Thehigher and more pointed teeth, with concave anterior cutting edgeand smooth denticles, distinguish this species from E. crejiulatus,which it most resembles.EDESTUS MINUSCULUS, new species.Edestus ef. minor IvARriNSKv, Verliamll. rnss.-kais. min. Gesellsch. St.Petersb., 2d ser., vol. 26, 1898, p. 379, pi. 4, figs. 12, 13.As cited above, Karpinsky identified provisionally as E. minor anddescribed with illustrations a single tooth of an Edestus which hadbeen found in the lowermost Permian, the Artinskian stage, near Mos-cow, Russia. A comparison of Karpinsky's description and figureswith the numerous good figures that have been published of E. minorNewberry has convinced the writer that the tooth in question belongedto a species quite distinct from E. minor.The tooth in Karpinskj^'s possession was a small one, the heightfrom the base of the enamel to the apex, measured along the axis ofthe tooth, being onh' 14 mm., plus 1 mm. or 1.5 mm. that had beenbroken from the apex. It was therefore only about half as large asthe specimen figured by Newberry ** and by Eastman .^ We must ? Geol. Surv. 111., vol. 4, pi. 1, fig. 2. "E. vorax."* Mark Anniversary Volume, pi. 21, figs. 2, 3. L NO. 1699. OX EDESTUS AND RELATED GENERA?HAY. 49conclude, therefore, either that it belonged to a much smaller species,than E. minor or that it belonged among the teeth of a half-grownanimal.If now, from Eastman's beautiful figures, apparently the best yetjDublished, one compares the basal length of each tooth with its height,it is found tlir.t the ratio of the base to the height is 0.83, 0.82, and0.81 in the first, second, and fourth teeth, respectively, 0.91 in thethird and sixth, 0.93 in the fifth, and 0.9T in the seventh. "W^iilethere are some irregularities here, no encouragement is given to con-cluding that the ratio would rise in the earlier-formed teeth. Now,the ratio of the base to the height in the tooth described by KarpinskyIS 1.18, This means that in E. minor the base is considerably shorterthan the height, while in the Russian tooth the base is considerablygreater than the height.An examination of the figures of E. minor shows that the hinderborder of each tooth meets the anterior border of the next at an acuteangle. Karpinsky's figure shows that the hinder free border of thetooth was turned at a right angle with the hinder cutting edge, anarrangement that would have made the angle between successiveteeth quite ditferent from that in E. minor. A somewhat similarprocess is seen at the hinder end of the last tooth of E. crenulatusand even of E. minor., but to assume that the Russian tooth was thelast of the series is to abandon the supj)Osition that it was the toothof a young animal. Attention may also be called to the fact that inKarpinsky's specimen the apex of the concavity of the anterior borderis placed between the middle and lower thirds of the border, while inE. 7ninor it is placed considerably lower down; also that the hindercutting edge of E. minusculus is far more strongly convex than thatof E. minor.In the specimen studied by Karpinsky the height of the sheath,taken at the front end of the tooth, is 0.3 the basal length of thetooth. If the last tooth of E. minor had the same length as the oneimmediately in front of it, the height of the sheath, obtained at thehinder border of the last tooth present, would be 0.75 of the length ofits tooth.The section of the sheath of his specimen that Karpinsky has pub-lished requires notice." This section shows that the lower border,close to the tooth, was rounded, not sharp, as it is in E. crenulatus.,E. serratus., and E. heinrichii. No section of E. minor has hitherto,so far as the writer knows, been published. Prof. F. S. Loomis, ofAmherst, Massachusetts, has kindly sent me an accurate drawing ofthe broken hinder end of the type of the species, now deposited in the " Verhaudl. rnss.-kais. iiiin. Gesellsch. St. Petersb., 2(1 ser., vol. 26, 1898, p. 380,fig. 16.Proc.N.M.vol.37?09 1 . f)0 PROCEEDINGS OF THE NATIONAL MVSEU2I. vol. 37. collection of Amherst College, This drawing, here reproduced (textfig. 6) , represents a section across the sheath immediately in front ofthe eighth tooth, now missing. For comparison with it is shown Kar-pinsky's section of his specimen (text fig. 5). The difference willbe readily observed. It may be noted here that Mr. E. T. Newtonpublished ? a description of a species of Edestiis, E. triserratus^found in the coal measures of Britain. The shaft appears to havebeen much curved. Beneath the tooth the lower border is thin andangular. In front of the tooth the border isbroadly rounded. 2. ON THE HISTOLOGY OF EDESTUS.The organ called Edestus^ whatever its positionFIG..J.?EDESTLS MINUS- aud Its functiou, is composed of dentine whichcuLus, x3. TYPE. -^^ penetrated bv numerous terminal arteries andSECTION OF SHEATH }-JUST IX FRONT OF veius aud capillary canals. On the surface of '^?"'^"' what are called teeth is a thin layer of what isprobably true enamel. The dentine must be classed with that calledby Tomes vasodentine, although, like osteodentine, there existed nodistinct pulp. Some sections have been made, in order to show theminute structure of the organ in question. These have been preparedfrom two specimens of what are regarded as Edestiis heinrirhii,which have been most kindly sent me by Dr. Bashford Dean, of theAmerican Museum of Natural History, New York. The specimenscame originally from western Indiana. There is rejjresented in fig.1, pi. 13, a section across one of the segments, so taken as to includethe front of the tooth. That part which belongsto the tooth broke away from the part below itduring mounting. An examination of this figureshows that in this genus the central core ofdentine, which contains the larger vessels, wasnot sharply marked off from the- more suioer-n ? ^ i.-^ -i. ? ^ 1 IT ? rr T - Fiu.6. ? Edestus minor,ncial portions, as it is marked oii m Helicopnon^ ^j ^^^^ sec-as shown by Karpinsky. Below the center of tion of sheath justthe section there is a large vessel that probably '^ front mcorresponds to Karpinsky 's " Liingscanal." The section appearsto have fallen where the canal was sendinp- off a large branch. Inthe specimen figured all the larger vessels and many of the smallerones apparently have the lumen ojoen. They are really filled with atransparent mineral, probably calcite. Each, however, has a nar-row black border which represents a deposit of pyrite or marcasite.Many of the capillary channels aj^pear to be filled with pyrite,but this appearance seems often to be due to the position ofthe vessel in the section, for when the near and the distal walls "Quart. Jour. (4e(>l., vol. GO, 19()4, pp. 1^. pi. 1. NO. 1699. ON EDESTUS AND RELATED GENERA?HAY. 51have been cut away the hinien appears. As the surfaces of the organare approached, the filling of the channels with pja-ite becomes morecomplete. In fig. 1 of pi. 13 and also in fig. 2, in order to bring outthe structure, the lumina of the vessels are represented as black.The light spaces between the network of black lines represent thedentine substance.Examination of the section shows that the longitudinal canals,large and small, are abundantly connected by anastomosis, so thatthe vascular apparatus formed a dense network. In the lower por-tion of the section, that corresponding to the root of the tooth, manyof the larger capillary canals approach the surface, and probablysome of them passed out into the surrounding tissues. Fig. 2 onpi. 13 represents a median sagittal section of another small segmentof Edestus. This is taken in front of the tooth and includes no partof it. Most of the vascular canals run longitudinally. The mainlongitudinal canal is seen near the bottom of the section. As seenin favorable situations, fine branching lines run away from theborders of the capillary canals. These lines are regarded as markingthe dentinal tubes. Often, especially near the capillaries (pi. 13,fig. 3, taken from near the anterior border of a tooth) , they havethe lumen filled with pyrite, and then they resemble the canaliculseof bone. Where not indicated by pyrite filling, the tubes may never-theless often be traced out under the microscope, and they constitutea network of fine lines in the dentine. Nowhere does there appearto be any layer of dentine made up of tubes running parallel withone another.The layer of enamel is so deeply stained with pyrite that fewobservations can be made on it. In one spot it is sufficiently thinand translucent to allow it to be seen that the enamel is penetratedb}^ nearly parallel black lines, which stand at right angles with theouter surface of the tooth, but do not quite reach this surface. Thisis to be taken as that variety of enamel described by Tomes as beingpenetrated by dentinal tubes."KariDinsky * has noted the resemblances between the teeth of Hell-coprion and those of various sharks, living and extinct. The presentwriter has wished to compare Edestus with the spines of fossil sharks,and has accordingly made sections of a fragment of the spine ofCtenaccmthus varians (pi. 13, figs. 4, 5). Although differences be-tween this genus and Edestus may be observed, the writer regardsthe structure of the two as being essentially the same. In the speci-men of Ctenaennfhits nearly all the capillary vessels are probablyfilled with limonite, while few of the dentinal tubes are thus filled. "Manual Dental Anat, 6th ed., p. 30.* Verhandl, russ.-kais. min. Gesellsch. St. Petersb., 2d ser., vol. 26, 1898, p. 420. 52 PROCEEDINGS OF THE JATIOXAL MUSEUM, vol.37.The larger blood vessels do not aj^pear to be so richly connected bycapillaries as in Edestus. Under high power a dense network ofbright lines, which are regarded as representing the dentinal tubes,is to be seen, running irregular courses and branching dichotomously.The shadowed areas seen around most of the capillaries are producedby the network of dentinal tubes, slightly stained Avitli iron. 3. DESCRIPTION OF NEW SPECIMENS OF LISSOPRION FERRIEUI.In 1907 " the writer described a fossil Avhich he regarded as re-lated to Edestus^ but still more closely to Helicoprion. The typespecimen, now the property of the U. S. National Museum, Cat.No. 6091, had been found in Upper Pennsylvanian dejiosits, nearMontpelier, Bear Lake County, Idaho, by Mr. AV. F. Ferrier. Atthe time of publishing the description it was impossible to deter-mine whether the complete structure would prove to be straight orslightly bent, as the species of Edestus, or strongly bent, as the fossildescribed b}?^ Dean as Edestus lecoxtci^ or spirally coiled, as Heli-coprion hessonoici Karpinsky. Immediately after the appearanceof that description the writer received from Mr. Ferrier two ship-ments of specimens from the same horizon at Thomas Fork, "Wyo-ming, not far from the type locality. These showed that the series ofteeth and their shaft formed a spiral resembling closelv that ofHelieoprion. From the best of these specimens have been preparedfigs. 1 and 2, on pi. 14. At a later time, about October 1, 1907, Mr.Ferrier made a fourth shipment, consisting of a block of lime-stone, in which there was a complete example of this curious fossil(pi. 15). Unfortunately the limestone is excessively hard andtough, while the fossil teeth and their shaft are friable. As a resultthe plane of cleavage has passed through the shaft and most of theteeth instead of over their surfaces. Nevertheless the specimen dis-plays well the coils of the spiral and the outlines of most of theteeth. Taking all the specimens together, the most important factsregarding the structure are made known. Credit is due Mr. Ferrierfor his interest in collecting so nnich material belonging to thisspecies. He has, moreover, presented to the IT. S. National Museumthe type of the species and important parts of the other specimens.Mr. Ferrier is a geologist and paleontologist of much experience,having been for some years assistant to Sir William Dawson, ofthe Geological Survey of Canada, and being now engaged as miningengineer in charge of phosphate mining for a commercial company.Besides the specimens of Lissoprion Mr. Ferrier has collected manyinvertebrate fossils from the deposits that furnished Lissoprion^ and ? Science, vol. 26, p. 22. NO. 1690. ON EDB8TUS AND RELATED GENERA?HAY. 53these fossils are being described by Dr. George H. Girty, of the U. S.(jeological Survey. Doctor Girty has verj^ kindly furnished mesome information regarding these fossils and their relationships. Hewrites me that the specimens of Lissoprion were obtained from phos-phate beds of from GO to 100 feet in thickness and placed near themiddle of the Preiiss formation. The fauna has a facies stronglyunlike anything known from the Pennsylvanian of eastern NorthAmerica, and man}^ features tend to ally it with the upper Carbo-niferous faunas of eastern Europe and Asia. In fact Doctor Girtyfeels little doubt that it is equivalent to a part of the Gschel stageof the Russian section. Some of the characteristic fossils of the phos-phate fauna are Chonetes ostiolatus, Pugnax weeks/, Amhocoeliaarciiata, Nucula inontpelierensis, Yoldia mcchesneyana, Leda ohesa,Plagioglypta canna, Omphalotrochus ferrieri, O. conotdeus, andGastrioceras simtdator. The genus Productus is poorly representedin the phosphate fauna, but contains four species closely relatedand perhaps identical with forms occurring in the Gschelian ofEurope. These species, with the species of OmphalofrocJins andothers occurring in the overlying strata of the same formation, arethe forms in which the affinities with the Russian fauna are especiallymanifested.In the case of the specimen which furnished fig. 1, pi. 14, the rocksplit in such a way as to expose the right-hand side of the firstfive large teeth, those at the termination of the shaft, and the leftside of the fifth of these (counting from the end), two others suc-ceeding this, and several small teeth of an interior whorl, as well asa part of the shaft. The figure has been prepared by combining twophotographs, that of the left side having been reversed. The desig-nation of the sides as right and left is made on the assumption thatthe base of the spiral, the larger end, was directed backward in theanimal, a view that may require modification. The 5 or 6 teethseen in the lower part of fig. 1, pi. 13, formed probably the begin-ning of the second whorl, no remains existing in the specimen ofthe first or innermost whorl. Some traces are found in the matrix ofthe remainder of the second whorl. The large teeth would thenbelong to the third whorl.It will be observed that the shaft of the specimen extends back-ward (toward the left) some distance bej'ond the last tooth produced,and the same will be found to be true of the species called by DeanEdestus lecontei. The last tooth present can hardly have been thelast one that would have been developed had the animal lived longer,for this tooth lacks much of having the size of the teeth of the typespecimen. In this the largest tooth has a height of 36 mm. and awidth of 17 mm., while the last tooth of fig. 1, pi. 14, has an axialheight of 30 nun. and a width of 11 mm. 54 PROCEEDINGS OF THE XATtOXAL MU8EUM. vol.37.The small teeth of fig. 1, pi. l-t, present only a part of their upperportion, or blade.The specimen represented by fig. 2, pi. 14, presents wholes orparts of 13 teeth and the corresponding part of the shaft. Plate 15is taken from the specimen that displays the whole of the spiral.Unfortunately the matrix is of such a dark color that the fossil doesnot show as distinctly as is desirable. However, from this it islearned that the structure, dentition or spine, whichever it may be,consisted of a shaft of a little more than two and a half coils and aseries of enameled teeth occupying the outer border of the shaft. Theinner coil with its minute teeth was, of course, first produced. Itis not probable that the smallest teeth seen are the first that the animalpossessed. Some smaller teeth and their shaft may be hidden in theobdurate matrix, but it is more probable that they had been lost bythe animal long before its death.The greatest diameter of the specimen, measuring from the apexof the last tooth to the apex of the one on the opposite side of thecoil is IGO mm. The inner coils were not in contact with the outercoil nor with each other. The apices of the teeth at the beginning ofthe second coil are removed by about 10 mm. from the inner borderof the shaft ; the innermost teeth approach within 3 mm. of the shaft.It is impossible to determine exactly the whole number of teeth. Anestimate made as accurately as possible indicates that there were 32teeth in the outer coil, the same number in the preceding coil, and 22teeth in the portion remaining of the innermost coil, in all 86 teeth,as against 130 in Ilelicoprion hessoyiowi; but the latter species pos-sessed about one more coil than did the species here described. InKarpinsky's species there were 36 teeth in the innermost coil, 43 inthe next, and 51 in the outer.Karpinsky showed that the teeth of his species might be regardedas consisting of three portions. The first includes the cutting blade,extending from the apex to the points where the edges of the bladecome into contact with the blades of the succeeding and the precedingteeth, respectively; the third portion includes that part that is nar-rowed and turned toward the older teeth of the series; the secondportion is found between these two. In the larger teeth of Helicop-rion the intermediate portion occupies half or more of the height ofthe tooth; in the case of the smaller teeth it becomes reduced inimportance and may become merged into the third portion. InLissoprioii this middle part may be said to be present in all theteeth, but to be relatively unimportant. In Helieoprion the blade isrelatively longest in the oldest, or smallest, teeth, forming some-times more than half the height of the tooth, while in the largestteeth it forms onlv about a fourth of the height. In all cases NO. 1690. ON EDESTU8 AND RELATED GENERA?HAY. 55the blade forms, in Lissoprion, about one-half the total height of thetooth.The apical angle of the teeth before us is obtained by drawing linesfrom the apex to the opposite ends of the cutting edges. This anglevaries with the size of the teeth. In the teeth originally described,the largest yet found, the apical angle is 48?. In the largest teeth ofpi. 14, fig. 1, the angle is 35?, while in the teeth of the specimenrepresented by pi. 14, fig. 2, it is 32?. The smallest teeth appear tohave the same angle as just given. It is seen, therefore, that the angleincreases rapidly in the largest teeth. Karpinsky has stated ** that inIIelicop7'ion hessonowi the apical angle is 30?. The present writermakes it 45?.The cutting edges of the teeth of LissopHon were originallydescribed as being smooth ; but some of the newer specimens showthat these edges were sometimes feebly crenulated.The middle portion of each tooth is short, convex posteriorly, con-cave anteriorly. It passes insensibly into the third portion. Thelatter is naiTowed to a point below and turned toward the older partsof the coil. In the smaller and medium-sized teeth its extremityreaches forward to a point opposite the hinder border of the secondtooth in advance. In the larger teeth it extends forward only to themiddle of the tooth immediately in front. Each tooth touches itsl^redecessor and its successor only at the base of the blade. Themedian and third portions of the successive teeth are separated by aspace very narrow and varying little in relative width throughout theseries. In IleVicopr'ion hessonoiri the interdental spaces vary con-siderably, being much wider relatively between the smaller teeth.''All the teeth of Lissoprion were covered with enamel, but this has,in the specimens at hand, been altered or removed. It seems to havebeen traversed by narrow ridges, which radiated from the apex of thetooth.Fig. 2, of pi. 14, furnishes a good illustration of the shaft and itsrelation to the teeth. It will be observed that a wide band of theshaft is exposed below the enameled processes of the teeth, the widthin the case of the specimen figured being G mm., one-fifth the heightof the teeth and the shaft taken together. In Helicoprion there is farless of the shaft visible below the teeth; according to Karpinsky'sfigures, about one-fifteenth of the height of the teeth and the shaft.Text fig. 7 shows a section through the axis of the second tooth fromthe right. It is seen that the sides of the shaft are convex and that inthe lower border there is a rounded notch. This represents a gutterthat runs along the inner border of the shaft. A similar gutter occu- " Verhandl. russ.-kais. miu. Gesellsch. St. Petersb., 2d ser., vol. 26, 1898,pp. 383. 402, fig. 23.^Idem, p. 394. figs. 24-29. 56 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. 37.pies this border in llelicoprion; and Karpinsky thinks that it mighthave conducted some kind of vessel. According to his view, thegutter was comj^leted below by a layer of shagreen, but I find noevidences of any such a covering. The sides ofthe shaft are unbroken and the edges boundingthe gutter are smooth.The lateral surfaces of the shaft are coveredi)y a layer which looks as if it might be enamel.For some distance below the teeth this is pittedso as to resemble in miniature the pittings of thecarapace of a trionj^chid turtle; but low downthe enamel is raised into delicate ridges thatRiERi, XI. SECTION I'u^ i^arallcl with the shaft. Doctor EastmanOF SHAFT AND TOOTH, has, lu defining Helicoprion, stated that the1, ENAMEL OF THIRD ., j.,i ic. i ii nilTOOTH; 2, GROOVE sidcs oi thc sliait are traversed b}^ a doubleBETWEEN NARROWED lateral groove. This is, however, an error, whichPROCESSES OF THIRD , ii.i ? /? !? ^ i ? j-AND SECOND TEETH; has (lOUDtless ariscu from a slight misconception .-?, ENAMELED PROCESS of tlic scctlous pubHshcd bv Karpinsky." InOP SECOND TOOTH ; 4, , , ? , i , , i ' i ? iEXPOSED PORTION OF ihosc figurcs tlic two iiotclies on each side repre-shaft; 5, GROOVE sent, not sections of as manv longitudinalALONG LOWER BORDER ' i;/> ' t / ^ l^ ^ iOF SHAFT. grooves, but or grooves between the downward])rolongations of the crowns of the teeth. Thereare no longitudinal grooves in Lissoprion and no room for them onthe sides of the shaft of HeUcojmon. 4. DESCRIPTION OF A NEW UENUS.TOXOPRION, new genus.The type of this genus is Dean's Edestv.9 lecontei. Doctor Eastmanhas recognized that this species did not belong to Fdestvs, inasmuch ashe included it in his genus Campyloprion; and afterwards, on remov-ing the type of the genus, O. amiectens to Heliocoprion, he essayed tomake lecontei the type. The writer called attention to this matter in] 907.^ Even were this procedure admissible it w^ould not be advisable,for the species annectens may yet prove to belong to a genus distinctfrom Helicoprion and would then require the name Campyloprion.The teeth of Toxoprion resembled most those of Lissoprion, but theshaft, though strongly bent, formed only a part of one coil. In thisgenus the present writer includes H. Woodward's Eclestus damsii,found in Australia. In this species it will be observed that the wndthis considerably reduced in passing from the newer to the older endsof the specimen, so that it is not likely that a complete coil was " Verhandl. niss.-kais. niin. Oesellsch. St. Petersb., 2cl ser., vol. 26, 1898, p. 397,figs. 30, 31. " Science, vol. 26, p. 22. NO. 1699. ON EDESTU8 AND RELATED GENERA?HAY. 57formed. It will be seen, too, that the teeth change considerably asthey are followed from one end of the shaft to the other. In thenewer ones the downw^ard prolongations are pointed and carried for-ward even to the extended axis of the fourth tooth in advance, whilein the smaller and older teeth the prolongations are truncated andreach only the extended axis of the second tooth in advance. Thepart of the shaft exposed is very narrow. 5. DEFINITIONS OF THE GENERA.It is evident that Lissopnon is closely related to Helicoprion^ butit is believed to be sufficiently distinct. It is possible that futurediscoveries may abolish the differences noted.Edestvs.?Shaft straight or slightly bent, roots of the teeth be-traying distinct traces of their original distinctness, and forming thegreater portion of the fossil. Blades of the teeth strongly denticu-lated. Type, E. vorax Leidy.T0X0prion.?Shaft bent, but forming less than a complete coil,mostly concealed under the bases of the teeth. Roots of teeth show-ing no traces of their original distinctness in the shaft. Blades ofteeth high, pointed, feebly denticulated. Type, T. lecontei (Dean).Lissoprion.?Teeth and their shaft forming a spiral, the coils notin contact. Roots of teeth indistinguishably consolidated. Shaftwidely exposed below the teeth. Inner border of shaft with alongitudinal groove. Teeth high, the middle portion short, thecutting edges smooth or feebly denticulated. Type, L. ferrieri Hay.HeUcoprion.?Teeth and shaft forming a spiral, the coils not incontact. No traces of the separate roots of the teeth. Blades of teethdistinctly denticulated. Little of the shaft exposed below the basesof the teeth. A longitudinal groove along inner border, as inLissoprion. Middle portion of teeth variable; in the larger teethgreatly developed. Type, H. hessonowi Karpinsky. 6. THE NATURE OF THE OBJECTS CALLED EDESTUS, TOXOPRION,LISSOPRION, AND HELICOPRION.In discussing this subject it is not necessary to enter into the his-tory of opinions regarding the position occupied and the functionperformed by the structures that have been described above. Theliterature of the subject may be found cited in Doctor Eastman'spapers." In the first of these papers this author, who has devotedso much attention to the fossil fishes and with such profit to science,discusses the homology of the objects before us. He there frames astrong argument in favor of regarding them as the consolidatedsymphysial teeth of the lower jaws of sharks. Karpinsky had pre- ?Bull. Mus. Comp. Zool., vol. 39, pp. 55-99, and in the Mark AnniversaryVolume, pp. 281-289. 58 PROCEEDIXGS OF THE XATIOXAL MUSEUM. vol.37.viously expressed the opinion that the spiral of Helicoprion wascomposed of the symphysial teeth of the upper jaw, the spiralhaving been pushed outside of the mouth and carried above thesnout.Eastman based his conclusions on the fossil called Campodus^which he shows was composed of the symphysial teeth of probablythe lower jaw. This roAv of teeth would corresj^ond to the medianrow of lower teeth mlleterodontus phih'ppi. the Port Jackson shark. Ifin this shark the outer and older teeth should, instead of droppingoff, cohere with the younger teeth, there would be formed at leasttwo-thirds of a coil, a structure that would resemble that ofCampodxs. If. then, the teeth .should become strongly compressedthe mass would resemble considerably that object that we call hereToxoprion. Further coiling would result in a series like Lissoprion.However, when Ave come to homologizing Edestus, Lissoprion, andHeJicoprion with the teeth of Campodus and Heterodontus diffi-culties are encountered. In the case of the two latter genera, thedifficult}' is to determine what disposition to make of such largespirals. If in Heterodontus the symphysial teeth should cohere withone another, a spiral of several coils might eventually be formed;but unless there were some especial arrangement developed, the spiralcould be completed only by a pushing of the older end of it throughthe skin and into the flesh and cartilage of the jaw. This would notcontribute to the comfort of the animal or the strength of the jaw,however much it might aid our efforts at homologizing. It wouldbe necessary, too, to conjecture a shark with a lower jaw of tremen-dous proportions to accommodate a spiral like that of Helicoprion.,the diameter of which is sometimes as much as 260 mm. If it besaid that the spiral projected far enough beyond the jaw to escapeburial in the tissues, it may be objected that it would have been ina position to be troublesome to the animal and exposed to injuries.The slender and bent dental mass of Toxopnon., too, would have hungdown in a position dangerous to its existence.A strong objection to placing any of these fossils in the mouth of ashark is to be found in the fact that none of them show an}' indica-tions of wear. The species of Edestus, described above, present noattrition of the enamel or of the most delicate denticles or crenations.Dr. A. S. Woodward, in speaking of Helicoprion ? has sought toescape this objection by supposing that the rows of teeth were sofar apart that they did not rub against one another. Nevertheless,constant contact with the food taken into the mouth must have pro-duced some wear.It seems certain that the general conclusions of Karpinsky regard-ing Edestus and Helicoprion must be accepted, namely: (1) These ? Geol. Magazine, dec. 4, 1900, vol. 7, p. 33. NO. 1690. ON EDE8TUS AND RELATED GENERA?HAY. 59animals belonged among the Elasmobranchii ; (2) the organs thatrepresent to us these sharks were more or less imbedded in the softparts; (3) they must have been organs in the median plane of thebody; (4) a considerable part of each of these organs must havebeen exposed externally?that is, they were not wholly buried inthe flesh. If these supposed dental masses w^ere in the mouth theywere consolidated teeth. The blades and the processes of enameldescending from them correspond to the crown of the teeth, whilethe shaft was formed through more or less complete fusion of theroots of the teeth. Now, while the crowns of the teeth in Edestusresemble closely those of some kinds of sharks, it must not be sup-posed that the cutting edges and the denticles correspond to thoseof sharks. Through strong compression of the teeth the originalcutting edges would have been brought to occupy what is now themiddle of the lateral faces of the teeth, while the anterior and pos-terior midlines would have become the cutting edges. The denticlesof these edges were develo^Ded later and could not have been derivedfrom the original denticles. It will be seen, therefore, that the wholetooth, if a tooth, suffered great transformation.If the organs under consideration w^ere not teeth they must havebeen placed either in front of some of the median fins, like manyof the other ichthyodorulites, or possibly behind a dorsal fin, likethe stings of the Masticura, or on the back of the head, as the spineof Xenacanthus. As regards Edestus^ it does not seem to be im-portant whether the new segment of the compound spine, if spine itwas, came up before or behind the older ones, since probably thewhole shaft was buried in the flesh. If it came up behind the olderones the spine might have been directed horizontally from the fin;if the new^ tooth arose in front of the older ones the spine may havebeen directed upward and backward in the fleshy front of the fin.If in the case of HeUcoprion and Lissoprion the new tooth had arisenbehind the older ones the spiral would have been directed forward,and on being subjected to oblique blows would have been liable to betwisted from its socket. It seems almost certain, therefore, that thenew teeth came up in front of the older ones, in case, of course, theorgan belonged outside of the mouth. If this is true, the end thathas in this paper been called the front end is the hinder end and theend called the hinder is the front end.The stings of the Masticura appear to be shed and replaced by newones. In Aetobatis there may be as many as five or six of thesespines present at once. The statements regarding the origin of thenew spines do not agree. Giinther " says that in the Trigonidae thestings are shed from time to time and replaced by others growing "Study of Fishes, p. 342. 60 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. 37.behind the one in function. Newberry ^ states that the worn spineis succeeded by another from behind. Jsekel ^ writes that usually onefinds in front of the base of an old spine the germ of a new one.In a specimen of Rhinoptera honasus in the U. S. National Museumthere is found a very small spine, loosely attached, in front of theone in function. In the German edition of Giinther's Study ofFishes, translated by Hayek (p. 23G), is a figure of the tail of aMyl'tohatis bearing two spines, the smaller of which is in front ofthe larger one. Storer,<" speaking of Mijliohatis aeuta^ states that thesmaller spine is in front of the larger. It appears, therefore, thatin the Myliobatida^ the new spine comes up in front of the olderones. On the other hand, Mr. B. A. Bean, of the U. S. NationalMuseum, has shown me a specimen of Urolophus jamaicensis and oneof an undetermined species of Twninra in which a considerablysmaller spine is behind the functional one; from which fact it mayprobably be inferred that in the Dasyatida* the new spine arisesbehind the one in function.It appears to the writer, therefore, that the objects called Edestus.^Lissoprion, Helicoprion, etc., may for the present be most easily dis-l^osed of by supposing that some ancient elasmobranchs developedin front of a median dorsal fin, or in place of it, not a single spine,but a succession of them. The new compressed spine, serrated infront and behind, arose in front of the older ones. Nevertheless, theroot of the new spine became directed backward beneath and on eachside of the preceding one, so as partly to embrace it. At first prob-ably the older spines were shed, but in time they began to cohere andthus form a compound spine. In Edestus this was straight or slightlybent. All of it, or nearly all, except the serrated teeth, was buried inthe flesh. As more and more elements were added, the organ becamemore curved and finally in some species formed a spiral, which wasdirected backward and the last turn of the shaft of which was ele-vated enough to keep the teeth from cutting into the skin. Such aweapon could be brought into action if only its possessor had divedunder its victim and brought the spine across its abdomen, thus dis-emboweling it, a suggestion already made by Trautschold. It is inthis way, as Doctor Gill informs me, that Gasterosteus attacks itsenemies.If possibly these organs belonged in front of a dorsal fin, that ofEdestus might have had its shaft buried in the fleshy part of thefront of the fin and directed upward and backward. The spiral ofHelic'oprio7i may be supposed to have been coiled on one side of thefin to which it belonged. The fin would have formed a partial sheathfor the spiral. ? Paleoz. Fishes N. A., p. 224.*Sitz.-Ber. Naturfor. Freunde, Berlin, 1890, p. 124. " Fishes of Massachusetts, p. 270. NO. 1699. ON EDE8TV8 AND RELATED GENERA?HAY. 61EXPLANATION OF PLATES.Plate 12.Figs. 1-3. Eclestus crenulatus, Xf.Fig. 1. View of right side.2. View of section at tlie fracture tliroiigli tlie second tootli from tlieriglit.3. View of section at tlie fracture tlirougli tlie fourtli tooth from right.4. Edcstus scrratus, Xt- View of right side.In Figs. 2, 3, the narrow white lines are at the boundaries between the con-tiguous sheaths. Plate 13.Figs. 1, 2. Edestus hcinrichii, X G.Fig. 1. Cross-section through shaft and front of tooth.2. Vertical sagittal section of shaft and part of tooth.In both of the figures the vascular channels are represented in black.3. Edestus hcinrichii, X45.Part of sagittal section through tooth to show vascular canals andthe dentinal canals diverging from them.4,5. Ctcnacanthiis varians, X6. Cat. No. 604S, U.S.N.M.4. Transverse section. The vascular canals are black.5. Longitudinal section. Plate 14.Lissoprion fcrricri, Xts.Fig. 1. Part of the outer whorl, with 7 large teeth, and some small teeth ofan inner whorl.2. The axis and about 12 teeth of another specimen.Plate 15.EissoiJiion ferrieri, Xiu. View of spiral showing the whorls and some of theteeth. U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. 37 PL. 12 U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. 37 PL. 13 \sf''/'' /A'- '?' '..>' -' ? -? '^^v^^^'siir ' ?'//7i,"T? i,y>^ V *?' 2