A \< ~ \ KK AY \ _ a . _ \ \\ AX \ \ ‘ ~ AX 2 IY NX \ oo AY WN \ A \ WS ~~ GZz3 ty Ye tj tj ee yy tye tty ie Li SY SS \ \ \\ XY _ Vi i ; oh . a it pa r a , i (on ee I : { i A ‘1 H , 4 7 i a . ; Lt i) i u H i ) i i - i i i Oi A i \ ; | oh : wi i : (5) mi f ‘ i t { i , if f on Wy ia " aie ee ; 4 ic i ; od i Ar a f Ae (sub if Y Aye aa n , 1 i 1 . i f ! i ; ie a i i J i 7 i | i 1 : if : . ' val ih J } ‘ t 4 be: We as ia i ; by | . h iw NV ; ty a a bf A ae ey ie rig ii (ge nw ui ; eae jie { i 1 N a ae Mg “tne aa ee Say ok if ‘ 7 ree eM yi ra i! o re ; a ; “ “ rt 7 SMITHSONIAN INSTITUTION UNITED STATES NATIONAL MUSEUM Bu.tuetin 82 A MONOGRAPH OF THE EXISTING CRINOIDS BY AUSTIN HOBART CLARK VOLUME 1 THE COMATULIDS Part 4a.—SUPERFAMILY MARIAMETRIDA (except the family Colobometridae) UNITED STATES GOVERNMENT PRINTING OFFICE WASHINGTON : 1941 For sale by the Superintendent of Documents, Washington, D.C. - - - - =---+---- Price $1.50 —_ : wAIa AT 7 i halenetit? blhamateee oil iuthond $4 sive amulet to dda d wet a ever faglbatediy 15 abdifanecy gorvil 16 aquery vordteial hus salary adt sabrevinninny Se iat - Ye daots alt loboload LECT si havfeiilerg ade iw jabbdenainal A ylinretisque saan add abbilour jog iaonmnd aide : eubtuienroedoleD ylivint 43 To seodiqeaxa + intel ado eaonexilde laioade valu opwd wvad Yad Med Wo ngiletagnd adit ol 30) eiet ines aaa Gil Whol 6 yore LintgofooN odd Yo to sob bedeal yon baal £ dt) sruswits aaibal oft Yo poitoollos orld. wi puinatiooge ih Lomn vrineata Mi pwadasivel add lo wiomib neared WA a De tolweril agnitotodl EE ntl '0d pecan eld cea eooitnollo oil ut ehaiwamoe toad doin oi) ai aban iotallon snabzoqeni add oor tos ode nent ML brn fared ADD sunlet) TW 1 ot nao’ ronda acwmrivece ot Jase odw ,(qiodelll luutavi) numer deine alt P natok od) yd-ebom aottorllos gaienedai yoy adi cele Ina aoitwiitadt " fe aw gop bates Vi A 6) baa tase aatbal ait af bo-Ctul dtaitib ar 3 - amisqaga? upsentd gillieT 9d1 to miinallos edi unt dae ante ution wld to otal. sreiergt Aedwld A of enobagilda hav nela ose I feout wed toy bil) ai aa wi, ree ogbidena’) da quota ovitetad MD AG of ins ad t4bow eetanivegs oil) sulmers ot cin hattionsg = pateroT ACI of ;dremwne(l asgadasqeD stunt lavigaload } ‘aeilichixealh al of :aitineaggiie haa atéinisics oldauley Jeon 70h ee j bolistel a0) eivach mond] do muse qd oda, Sobenih onidolt Wax whA ot has yapads snd af soide Yo noinlos odd it = oe alt ooteof ,2itd ool to mitonht mio! ot baa sisi Tanoitays evtare, hatin ads ui esrgsation yo oT mebubertsl gaifietens tind? 10) alustitary Yo dab « wero J ytardil e'mtdaw ee useunaae dunidancony cokqiniie yeas orl} (7 testo of eyaw euotuny | ion wht Yo srg ed) adil deena yLaatenos oved dail) scroitaonp qargurodg lo actoitouboryos -leitas Joomla ors snoiletwvlienT miilqjwryolnule eal of Baten yA aC to noivivioque ods tobou rswotdy vortuts pain dyad wenldailt 314 bamibetaall) oT do snvmeulé old To dite wane menild iw alluiors sldiseng vend aft etuasa of atzolte siodd > apaidoe : mii Lae Bebe tate act wh waa gateha ait 20 bowyi acamiseg# off” 7. Jb zheitlo ivewod. ynaMd meas Innate eebat2 Latta Se ecsorncets part oid eiritsearny oesal'T acini See? baa nebo ieee lwiaY) isee¢ eth 3.8.90) seul 15 ( MO) thaenes cosgedneqo) ume noseillg@O! eared ilies! shquatwdi/ wi\ smeeu A 36.4) taeda | eer thnalloH ,oteinsteniA ,mvow lM, iM. (MB sibel jettuoled muaiolt Aq eaiaD. envesw 4 orin”) ois Dera’ aw CONTENTS IPTCLA Cee eee enna eee ee ae Sea an See ee ne eee ERA eS HiittrOduCtiOn esa a sa ae ee oe ee ee Se ee eee a See OrdertComatuliday (Continued) pase = = a a ee Cera Seen nee eee Suborder/Oligophreatax(continued)=eese===-— == sense meen ee ee eee Sosa ees Supertamily: Marigmetrida===="=2"==2= "ss serene doe ween s ee oUt ee eee Ramily.Zycometrdaee serene == sae en eee eee ee eee ee eee (Genus-Zy Lometraa a= een ee ee ee ee mere Sees ee esa eee ke AY COMetr ay ClOCISCUS sae =m aaa apap mere erinnn Ny a aen Ne eee = NPE age ayn FY COMetra Clee ans ease ae = wae eR eee Le tae Zygometra andromeda_-- -- mt abate ean bem pat die Apnea et he eet tee kes AY SOME WA COMMAS = Oe ea ae eee ee ee a ee Tee ee ee Zygometra punctata_-_----- oe Ae re ee et Se ea Zycometra pristina === == ae ae eee ee ene Seen es Mek Genus Catoptometrassa ne == a= = mer eee marae pare N er ierens sh hi) 2 etree ee Catoptometranmagnifical==2s5s= ese sala ae eee eee ee Catoptometramrubrolav ass = 2s see ie eae a ee ee See ee See Catoptometrarophiural=2="2 esse eer ee. eae aaa ne ee eee Catoptometrathartiaubi======s ees sp anes ene ee ee nee Ranily sh udiocrinidaes ss ses a2 == ae ee eee een» BE eae eee Genus? Budiocrinusss222 3 22. seee see eee rns p= eee ease re Eudiocrinus junceus__------ Se ee eee eee: ee oe ee ee Eudiocrinus philenor-------__ Ee ae een en en eo = TO Sa Eudiocrinus pinnatus_-_---_- ee ee a ae ae ee eS Hudiocrinuslovenis== = 22s a eee eee es ee ee HUdiocrinusvariegatuss s\n sans Se bs semen ene) i Eudiocrinus venustulus__--_--_ ee eer Oe ee ene 2 oh SoeN rc Ee Eudiocrinus eoa- --=------=-- te See Ree ee RA ee Eudiocrinus indivisus- - - - -- ne ee eS a ree Ane Bed Eudiocrinus serripinna - - - - - a a ne Re ae I Se Pg ot ae HUIGIOCHINUS OLN SUS Seen ee ee ee ee Eudiocrinus gracilis_------- SI a ye oh pe aah ea ee Hudiocninusipulchellus] === =55= == ae seen amen Se ey eae Family Himerometridae_----------- eee ee ee ye ee Se Genus Himerometra__---------- Bo SSE eS eee eee HinierometrarsOle= sso 8225 = a Se eee ne ae ee ee ae oe Himerometral magnipinnas=-ce sean en emo ne eee Himerometrarrobustipinna == =n ae eee See ee oe Himerometra martensi- - --- ee ee ee ere eee eek SFeys Se Himerometray bartschii225====5es= cee See ee eee es eee ae Himerometra persi¢as= 2522s --— ee pat 8 Aas Sn BPE Genus Craspedometra_-_-------- Se en rete re wee ee os a Se et Craspedometraracuticitrac== =. ee eee eee eee eee eae GenusrHeterometra sean so tees melee oe ee ere eee ee eee Heterometra sa vipnlis=292s see on eee ae ee eo te Heterometramema todonse es eee eee ee a ee eee Vi BULLETIN 82, UNITED STATES NATIONAL MUSEUM Order Comatulida—Continued. Suborder Oligophreata—Continued. Superfamily Mariametrida—Continued. Family Himerometridae—Continued. Genus Heterometra—Continued. Page Heterometra quinduplicava-----_------------------------------- 245 Pletaromotra crenata. == 2- bee ann ea sa ee 253 Heterometra producta------------------------------------------ 275 Heterometra propinqua----------------------------------------- 277 Heterometra variipinna....--..---.-.._-.-------------------.- 278 GCENOMUOLIS SH pC RD ES e209 aon ne nee oe a eee ee 284 reterometie Afiies <== er oe ee 285 PETG ECOG BRE [DUEL CN Se a ee ee eee 287 TER USKOTNG GTO CRW AGU esta oe ee er ee 289 IER er peOTIG Gre) BINA LAT ot ce ee re 290 leterormenre (Com pth on =—2 = eee ane oe esa ew aaa nasoeeon ee 293 Pleterometra madagascarenkls- -- = one 2 eis 295 Heterometra,.amboinne- ~~ -- 2-50 oe ee een nee 297 Bleterometra xeynaUGlee-—-o-= 2a ee ee nas seen 302 Heterometra stent] son c «=> 3 ae ee ee eae ae 311 LePerOMIOLAy OUDUN = a =e momen see a eee eae a nee 312 eterOmeura phil Deltla= sa = ae nn eae eee oe 313 ML @Leronietra saTae. - = 2ccsn are eee UR soe et eee 318 ISLOROMG Gls DARIUS =e aE ie ee ee ee 319 Heterometra\bengalensis-=22- (oe ete oa aScece 321 Meteromotra africans =~ ——— — 24 see ee ee Seek 325 Hleterometra schiegellin=. == 2 ase so ene eee a Be eee noe 329 Heterometra flora_----------- See ee ae ee eee eeeoee 333 Heteromotra CelIagORG 222-2 am 2. ee ee ene 334 Bieterometra delicate 2.22 ——-o. ssee et ap ee og 336 Genus Hommlometras. oes eee ere ree Neg hla 337 HomAlometra dentiwlnteac. ose a ee go se es ee ee 338 CrenUR AIM PHIMOtd oe oo — ant Seen mee en ee 340 Amipnimetra spectADWS=-- soo ee ce 347 AMG MMeEtr a WnOUGN . a ss— 5 oa a ee et ee 349 DINphimMetrA) pPinOit ONMIGs as ee ee ee ene 358 Amphimetra tessellata tessellata__......_....-.__...._.--_-__--_. 360 Amphimetra tessellata discoiden_-—- en ec 376 Amphimetra tessellata papuensis.5=—_.. - 8 ose ne eee 385 ATP HIMoetra ONslOb. =. = Soa ae ne Een 387 pInphimein lAeviNiINss— seo = 8. == See eee ee 389 AMOI EATIAIIOUOROL Oe cee ee ene ea. 8 eae SP deer 391 LE EITEMO; ape) e: pane S Sele Eee Sr eeelsa Ste: ee ee 396 OX VINGULOLINAUCA. (ose ae ee eg ek Ce Bi Se 398 Oxyineira enw Oir a ae ae te ee ae 401 (eV eLrA TINBO Re | oon ee eee ee eee le 402 RSETIIN SLAP NMNOMIGLT A 92 see a eg 407 Btephanomoetra echinvssscss2.2 ees ee 409 Stepnanometrs tenuous = ss eee ee 413 Stephanometra spinipinas oo ea ee ee 415 Stephanometra oxyacantha.-.__.-.-...-------u-----u.---------e 418 HLePUANOMOUIA BPICA tA. ooo op ooee neue Ee ee wc eee 424 MUSDUMMOMP LA TONGA NCION 3 ee ere ee et a 436 CONTENTS VII Order Comatulida—Continued. Suborder Oligophreata—Continued. Superfamily Mariametrida—Continued. Family Mariametridae—Continued. Page Genus*Belometrar=2* =s. S22 Ses Sa ey Sine se es Le ae Se ee ac 459 ‘Pelometrarambonensisetsa. =- sa en ee ee UE ees AN Seek 459 Genusthiparometras sss se es wee eee eee PW, DE ete 460 biparometravarticulataesse=— see ee ee ee ee oo ee ee 461 iiparometrarprandise = sessess = a= ae ae aa ee ee eee 467 ibiparometrarepalish <.s ee oe ese nes ee en eee Dee one 470 Genus lamprome transac ke pe ee oe ee i Rh wh A Es he ree 472 samprometraspalmata palmate sme = sme eae ee ee eee 474 Mamprometraipalmatargy ces saseee ee ae ee ee 517 Mamprometravidunzin geri ses se ee ee ee ee ea ee 527 Genus Dichrome tray eoe ee ee tee ay Deere ps alg aes ee Se ag 536 Michrometraisty lifer seen e a eee are eee ee Se ee 538 Dichrowmie tray iis Cuil Gee eee ee ee re ee een ee ee eee 541 Dichrometra fapella tae we et tee ee ne ee oe ree ee ten een Sd 544 Dichrometrancenil CLerye sae ees esr ee pee eee ee ee eens 556 Dichrometragafra tse ey Be sake Sete elis Says eh 8 HU ee ee a 558 Dichrometra dod erlein ieee sees ees ee ee eee renee 562 Dichrometrayciliatas = se es se) ce Lea hcg Se Se KAS 565 Genus: Mariametracss2 ean tees Baye Sanne es Fe pam, Al es See A es nee 566 Mariametraysubcarins tase nee ee nee I ee 568 Mariarnctra; viCaTi a ees eter aac wees reer eee Seah EE ice tetas De Mariametradelicatissima === ssa says a a snl ye eee ees weeny. Se eee 579 Mariametraytuloenct sit sexs sere em ae meee eee ee enepee any eee Spee 580 Mariametra;tenuipes soe ee chee i ae eee ns ee 582 - | - 7 J art “ae ee ta aye vie é ice “igh ue. sia cv ' a an os Ltt uy i? “an can 7 : | - Vide on bel . : _* ; fy ~~ 7 7 Stuy? ; a o™ a “0 53] | _s / am f bet ingDhe atify down’) aybedD - : - e Supavibhirs” “A mrtiquy |! bruh © ihobathaiOeebi fanmail vial aqck ae ; : hoi iy nal briantahe th viet 4tgree 018424 ee oe fh den Maret eet oi we ee 2.2 Wirerdline weihoyisl e a ivi Cees oe Peer Ta em 8) : —. - g=4 eS Selisihn iiteyote; =e ome a hea 6 Aes A i oth rieren LT ee eee oe ee Ee ee ae tot uae - eek dell,» .b-«) ata. ei be aia) See ee pete Sao A> Blganfany aecter HitAgcel n> Sere \. eG Alaileg thao riigeet Aalfee 1S - ; Pradirl tae A otic ; bts deck) Ge keel ee Ree) W-- 4 a pases aiiagrdanttl wasko ice weet 4 2 ee ~ woe ls atbatonisht * a ee Ar ne tye a el on wr Pied ee i ie 5 ae siaiia nh aslanostoltT ~ os ; ; ; fi ave fice) BthveeretAsal 7 anes a ee = ere fale wd j litilinti®h siioautidl ms er by : ov fs _itallig etenamtallt ; - = oe bias oe ta siding BE iaved? Sy alte exe oe = alanivevdiw mhenanalh 4 ti ; 7 + .. Atal y a aehtely: ~~ a : o - cuban aveaceitaly ; wien. >~2 loan Jee : pg e~wwhalipred® eiigciriel? om nt ce eral sad eR es eo Ma Bier, «yd ~:~ SP OP etree Oye u Ont deecks «46s a4 eee. ae ins hel ene aaah = ate “* f S ‘ be - - me ‘ - 7 = ~*~ s : —_ * i ? = - , ‘ 7: A MONOGRAPH OF THE EXISTING CRINOIDS By Austin Hopart Criark INTRODUCTION Tur present part, Part 4a, of Volume 1, Bulletin 82, is a continuation of Part 3 preceding. Part 3 contained the account of the superfamily Comasterida, and the present part includes the account of the second of the three superfamilies of the Oligophreata, the Mariametrida, with the exception of the family Colobometridae. The arrangement of the families and higher groups adopted in the present work is given in Part 3, p. 65. Keys to the families into which the superfamily Mariametrida is divided are included in the keys to the families and higher groups of the comatulids given in Part 3, pp. 69-74. Since the portion of this monograph dealing with the structure and morphology of the comatulids was published a large amount of additional work has been done by others, particularly by Prof. Torsten Gislén, of the University of Lund, Sweden, and many interesting and important new facts have been brought to light. Some of the interpretations of the structural peculiarities of the comatulids advanced by these authors differ more or less widely from those given earlier im this work. Though thoroughly appreciative of the vast amount of original work upon which these interpretations are based, I fail to find, upon careful analysis, any reasons for altering my opinions already expressed. The main point of divergence between my opinions and those of my colleagues hinges upon the weight to be given the Paleozoic and earlier Mesozoic forms in an elucidation of the structure of recent types. I maintain that until the recent crinoids are far better known than they are at present, especially in regard to their younger stages, it is futile to attempt to interpret the details of their structure from compari- son with earlier forms. For most of the earlier forms, as we know them, represent the terminal twigs of developmental branches of which the generalized beginnings are unknown, just as the adults of the recent forms represent the ultimate product of an unknown ontogeny. And even if the earlier stages of all the recent forms were known, it is by no means certain that these would give any definite clue to their ances- try and morphological relationships when compared with the adults of earlier types. In my opinion the most valid and logical conclusions regarding the comparative structure and morphology of the recent crinoids are to be reached by detailed and intensive study of the recent types alone. Until we are in possession of sufficient facts to enable us to understand the recent types, and thus to be sure of our ground, 1 2 BULLETIN 82, UNITED STATES NATIONAL MUSEUM we are likely to be led astray rather than aided by any detailed comparison with fossil types. However, I believe it only fair to my colleagues to give a summary of their work and of their views, without further comment, so that future workers may be able to judge for themselves the relative merit of the various opinions expressed. PINNULES AND ARMS Dr. Gislén, in common with Dr. Th. Mortensen, disagrees with my interpretation of the pinnules of the recent crinoids. Mortensen said that, guided by his researches and observations on pathological and embryological material, he considered himself bound to uphold the theory of P. H. Carpenter—that pinnules correspond morpho- logically to dwarfed arms. Gislén noted that Bather had stated that the only difference between pinnules and arms is that pinnules contain fertile gonads. But he pointed out that this dis- tinction does not always exist, for many stalked crinoids and even comatulids, as for instance Notoerinus and Comatula, have fertile gonads in the arms. Furthermore, distal pinnules, and all the pinnules of the anterior radii of many comasterids, also lack fertile gonads. He said that the only difference that should be made use of in a definition of the different types of brachial ramification is that the pinnule is a small unramified arm, as a rule considerably shorter than the main arm. He noted that, so far as he had observed, the articulations between the pinnule segments in recent crinoids are always unlike those between the brachials, if the articulations between the first two segments are excepted. Syzygial articulations between the pinnulars are found in Hyocrinus, and also between the segments of P, in Stephanometra. In Hyocrinus, as Gislén remarked, the pinnules are very long, approaching the main arm in size, and therefore the pinnule-bearing brachials may have the appear- ance of axillaries. When pinnules are of nearly the same size as the arms that bear them Gislén said it is a matter of opinion whether they are to be regarded as pinnules or as arms. He concluded that the phylogenetic development has doubtless been that an isotomic type of ramification was succeeded by a heterotomic, which in its turn was replaced by a metatomic. The lateral ramifications began to appear at regular inter- vals, diminished in relative size, and became unramified; thus the holotomiec or pinnulate type was reached. The arm of the recent crinoids, according to Gislén, should thus be regarded as a sympodium that has arisen in the following way: Right and left ramifications alternately have remained at their full development, while corresponding left and right ramifications have been suppressed into pinnules. Gislén gave a detailed account of the phylogenetic development of pinnules from arm ramifications, based chiefly upon some of Angelin’s type specimens in the Riks Museum at Stockholm. Mortensen showed that Dr. W. B. Carpenter’s account of a sympodial growth of the arm during ontogenetic development is incorrect. The pinnule segments are weaker from the very first, though on the other hand their longitudinal growth is at first more rapid. From an examination of specimens of Comatula pectinata and of Comanthus parvicirra Gislén found that in the anterior radii the arms terminate in the manner A MONOGRAPH OF THE EXISTING CRINOIDS 3 described by Mortensen for Antedon petasus—the main arm always retains its char- acter of chief branch, and the pinnules originate as small buds alternately to right and left. On the posterior arms the only difference is that the growth of the main arm is retarded, so that the pinnules reach their full size while the main arm grows not at all, or only slowly. Therefore these arms do not terminate in an axillary supporting two pinnules, but on the last segment that gives rise to a ramification there is an almost full grown pinnule, and a main arm composed of a few segments. As soon as more than one or two segments have been formed on the main arm, a new pinnule appears. The main arm, therefore, keeps its ramified character and never develops as a pinnule. According to Gislén the obstructive or inhibitive factor that causes the formation of pinnules instead of arms may in certain abnormal cases develop two pinnules on a brachial instead of one pinnule and one main arm, or two main arms. He con- cluded from this that sometimes the suppressive factor may produce a reversed effect—that what ought to have been a main arm has become a pinnule, and the reverse. He regards hypertrophied pinnules as the result of removed obstruction. Gislén classified arm regeneration in the crinoids as follows: 1. Restorative: Replacing of a single arm broken off; this occurs in all crinoids. 2. Reproductive or pseudo-augmentative: From a fracture one or more axillaries with two or more arms are reformed; thus the number of arms is the same as before the breakage. Presumably this occurs in all more or less full grown crinoids. It is called pseudo-duplicative when one axillary and two arms are reformed, and pscudo- multiplicative when several axillaries follow each other, and therefore more than two arms are reformed from a single fracture. 3. Augmentative: From a fracture a greater number of arms are formed than the broken limb possessed. This occurs in comatulids, and presumably also in Diplo- crinus and Teliocrinus. It may be duplicative, when one axillary and two arms are formed from the fracture, or multiplicative, when several axillaries following one another, and therefore more than two arms, are formed. 4. Reducing: From a fracture fewer arms are regenerated than the lost limb possessed. This is theoretically conceivable, though it is not known with certainty. Gislén said that the young of the stalked crinoids are known only in a very small number of examples, and ordinarily they are in such an advanced stage that not much can be determined regarding their method of arm augmentation. Augmentative arm regeneration in pentacrinites probably occurs in Teliocrinus and also in Diplocrinus, according to Déderlein. The drawing of a young Neocrinus decorus left by P. H. Carpenter seems to show that the formation of arms from the IIBr axillary takes place in the same simple way as the formation of the two arms on the IBr axillary in the comatulid pentacrinoid—the arm tip forks into two processes that are equally favored in further growth and so develop into two new arms. Déderlein said of the young of species of Metacrinus from the Valdivia expedition that he examined, “Ich konnte bei den jungen Exemplaren nachweisen, dass sich an einen oder dem anderen Armstrahl noch unmittelbar vor seinem Ende eine Axillare ausgebildet hatte, d. h. eine Gabelung angelegt wurde.” In his somewhat earlier work on the genus Metacrinus in the Siboga report he published a figure of an arm ramification in M. acutus in which “die Arme auffallend ungleich sind.” 4 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Gislén said that the young of Metacrinus described by himself in 1922 are very instructive on this point. His observations showed distinctly that the formation of 11Br, I11Br, and IVBr series takes place in the following way: On the simple arms certain pinnules begin to gain strength and to grow. On the sides of the strengthened pinnules new small pinnules arise, and the new arm finally reaches the same length as the main arm. The place for this strengthening of the pinnules is to be found in the region of transition between the large and rudimentary pinnules of the main arm. As that part of the arm provided with the latter is shorter the younger the individual is, it is clear that the arm ramification for the formation of [Br axillaries will occur fairly near the tip of the arm. Therefore the main arm and the new arm are not very unequal in length there, whereas the new arms on the IIIBr axillaries, and to a still greater extent on the [VBr axillaries, are very different in length at earlier stages. Thus the lengths of a main and a side arm on a IIBr axillary (Metacrinus interruptus, specimen 16) are 2 and 3 mm. respectively; of two young arms from a IIIBr axillary (M. interruptus, specimen 17) 5 and 1 mm.; of arms from a IVBr axillary (VM. inter- ruptus, specimen 13) 12 and 2 mm., and (M. nobilis tenwis, specimen 9) 5 and 1.2 mm., or 3 and 0.8 mm. Here, said Gislén, the factor of obstruction only succeeds in acting temporarily on the pinnule that is destined to become an arm. Thus arm ramification in Meta- crinus is more direct and primary than in the comatulids—a pinnule is strengthened, it ramifies, and then it grows until it is equivalent to the main arm. In the pentacrinites there is regeneration of broken arms, just as in the coma- tulids, and as in the latter the fracture is most often at the syzygies. When a post- radial series has been broken before a last axillary, regenerates appear that are like the augmentative regenerates of the comatulids. It is only by following series in different stages of development that the difference in principle between comatulids and Metacrinus in the way their arms increase in number can be ascertained. Thus the pseudo-augmentative arm regeneration of which examples have been given is only of a reproductive nature. In this connection Gislén pointed out with regard to the comatulids that repro- ductive arm regeneration appears in them also. It is necessary to avoid interpreting an axillary regenerate as evidence of augmentative regeneration, but the comatulids undoubtedly have augmentative arm regeneration as a means of increasing the num- ber of their arms, whereas in Metacrinus the method is more direct and primitive. A forerunner of augmentative arm regeneration is reproductive arm regeneration, which occurs in Metacrinus together with the primitive method of augmentation. In the more specialized comatulids the strengthening of a pinnule into an equiva- lent of the main arm never occurs normally, except possibly in Comatula etheridgei (the young of C. rotalaria). In these arm regeneration becomes augmentative as the surfaces of syzygies in the proximal parts of the arms acquire the power of forming axillaries with a greater number of arm branches than the lost portion of the post- radial series had. The apparently simpler postradial series of the comatulids must be considered as potentially forked, but the ramification does not materialize because the factor of obstruction permanently restrains the efforts of the pinnules to develop directly into arm branches. This ramification does not appear until after the break- A MONOGRAPH OF THE EXISTING CRINOIDS 5 age of an arm, (in the comatulids autotomy) when the dormant tendency is released following the inception of regeneration. Gislén said that on an examination of the genus Metacrinus one is struck by the extent to which the odd numbers are more frequent than the even in the division series—that is, if the hypozygals are counted as independent segments. Among the specimens of Metacrinus nobilis tenuis examined by him there were in the IIBr to VBr series 233 odd to 26 even; in M. rotundus 217 odd to 70 even; in M. interruptus 372 odd to 59 even. Moreover, if the number of syzygies in the different division series are examined it is soon found that the series with an odd number almost always have an odd number of syzygies, while those with an even number have an even number of syzygies. So if syzygial pairs are considered as units, as was done by Carpenter, the division series have an even number of segments. Counted in this way, the specimens of M. nobilis tenuis examined had 258 series with an even number of segments and only one with an odd number; M. rotundus had 275 even to 12 odd; and M. interruptus had 422 even to 9 odd—a total of 955 even series to 22 odd. Sperry made a curve for the variability of the I[Br and IIIBr series in M. rotundus, and he also gave a strong preponderance to the even series. He looked for the reason for this in the fact that pinnules on the inside of the arm are closer together than on the outside. If the number of segments is even there will be a smaller number of pinnules on the inside of the division series than on the outside. According to Gislén this observation is quite correct, but it does not give an explanation of the phenomenon. The real explanation presumably is that if there is an even number of segments (syzygial pairs counted as units) the new arm will be formed as a branch on the inside of the arm where it is more protected during its early growth than on the outside. How far-reaching this rule is may be seen from the figures given. It can also be confirmed by other facts. Gislén cited some examples of reproductive arm regenera- tion. Measurements of the regenerated arms show that the inner arms, even in the case of reproduction, are at first weaker than the outer (main) arms. This is explained by the rule just given, for the regenerate suggests the form that the actual ontogenetic development took. When, as sometimes happens, a new arm develops on the outer side of the main arm, the ramification often takes an abnormal form. The base of the new arm bends outward in a wide curve, and the two arms diverge not at equal angles from the main axillary, as is usually the case. Sometimes, in spite of an uneven number of components in the series, the new arm still appears on the inner side, owing to abnormal pinnulation. According to Gislén the explanation of an abnormal number in the division series is generally that the main arm has aborted and become a pinnule, while a pinnule has hypertrophied and taken over the function of the main arm—in other words a shifting of the sup- pressive factor has taken place. Gislén said that so far as can be judged from Agassiz’s work on Calamocrinus, in this genus too it is most frequently an inner pinnule that is strengthened into an arm. Carpenter’s figure of the young of Neocrinus decorus shows that in this genus the new arm is formed by the splitting of the growing point. Thus it cannot be said in this case that a new arm developing on the inner side of the main arm would be more protected than one developing on the outer side. In this case, therefore, the number of even segments (syzygial pairs counted as units) in the IIBr and following series is 6 BULLETIN 82, UNITED STATES NATIONAL MUSEUM not strikingly higher than the number of odd segments, which, to judge from the figures, occur pretty often. Gislén pointed out that in the cyathocrinids odd and even series occur mixed. In the pentacrinites two separate series develop, one with an arm in the place of an inner pinnule, and another with an arm in the place of an outer pinnule. In Telio- crinus, with its division series composed of few elements, the new arm is usually, though by no means always, in the place of an inner pinnule. In Diplocrinus, on the other hand, the number of segments in the division series is odd (counting syzygial pairs as units), and the new arm is therefore in the place of an outer pinnule. The genus Cenocrinus to some extent forms a link between Neocrinus and Metacrinus, since the outer series of divisions seem usually to have an even number of segments, while the inner have an odd number. Among the comatulids, according to Gislén, it is only in the subfamily Capil- lasterinae of the family Comasteridae that the form of pinnulation appearing in the genera Metacrinus, Teliocrinus, and Neocrinus and represented by the formula 3(2+3) reappears. So far as pinnulation is concerned this type is to be compared with division series of 4(3+4) or 4(1+2, 3+-4), though there the first syzygial pair has coalesced into a single ossicle. Ail the other comatulids have either two or four components in their division series. In regard to pinnulation these are to be consid- ered as one and two ossicles, respectively, since segments 1 and 3 in the division series never have pinnules. Thus in the former case the new arm must be considered as having developed on the outside of the main arm, in the latter case on the inside. The indifference as to whether it is an outer or an inner pinnule that is developed into an arm that is found in certain pentacrinites, which is to be explained by the shortening of the division series, recurs in certain primitive families of comatulids. Thus in the family Charitometridae there are about as many forms with two as with four components in the division series. The mixed type is also found in the Capil- lasterinae, Comasterinae, and Zygometridae (in the last two with a tendency for four components to predominate), and in the Calometridae and Thalassometridae (though most of the genera in the last two families have division series of two elements). The family Himerometridae has almost exclusively forms with four components in the division series, while the Comactiniinae, Mariametridae (including the Stephanome- tridae), Colobometridae, and Antedonidae have taken the other path and have practically always two elements in the division series. As the division series here have few elements, the young arms will be situated near the base of the arms. As in the case of Neoerinus decorus an inner pinnule is not more protected during development than an outer pinnule. Moreover, in the comatulids direct increase in the number of arms has been replaced by augmentative regeneration, and the small arm regenerates, hidden between the bases of the arms, are not very much exposed to breakage. In Gislén’s opinion this must be taken to be the explanation of the indifference to the number of segments in the division series. In the tendency of the axillaries to shift toward the proximal part of the arms, a tendency that can be traced in the form series of all crinoids, there is this advantage—if the arms are exposed to a bending that will result in fracture, these fractures will occur distally to the axillaries, and will therefore affect only single arms. On the A MONOGRAPH OF THE EXISTING CRINOIDS 7 other hand, in Metacrinus and forms of a similar type whole clusters of arms are easily broken off, and so there is a greater percentage of loss of ambulacral furrow. Occasionally in the comatulids also there occurs a direct strengthening of pinnules to an equality with arms. In a case of this kind in Antedon petasus described by Mortensen (1920) in 5 arms out of 10 both P, and P, have developed into more or less complete pinnule-bearing arms, P, being often stouter than P;. In another individual examined by Gislén P, on the left anterior ray had developed into an arm 75 mm. long that was as stout 9s the other arms. From the sixth brachial, where the pinnules begin on the hypertrophied P,, there are stout segments and oblique articulations, as on the other arms. Syzygies of normal appearance and distribution also occur, the distal intersyzygial interval being 3 muscular articulations. Gislén noted that a similar case, possibly normal, is found in the so-called Comatula etheridgei. If C. etheridgei (see Part 3, p. 309) is the young of C. rotalaria there is here direct arm development like that in Metacrinus. In a specimen of Cyllometra pulchella (=manca) from Mortensen’s collection (station 10) the right posterior radius has a normal undivided arm on the right; on the left there is a I[Br 3 series, the external arm from this being normal with a synarthry between brachials 1 and 2, syzygies between brachials 3-+-4 and 15+ 16, and the first pinnule on the second brachial, and the internal arm having syzygies between brachials 1+2 and 16+17, and the first pinnule on the fourth brachial. Gislén recalled that Springer described a specimen of Uintacrinus socialis with 11 arms, as the result of the development of a P, into an arm. Gislén said that occasionally, for one reason or another, pinnules develop directly into arms in more distal portions of the arms also. He cited a specimen of Asterometra anthus in which on the right arm of one of the postradial series the seventh brachial is axillary, the pinnule having developed into an arm with syzygies between brachials 4+5, 12+13, 19+20, and 27+28, a pinnule on each side of the third brachial, but the pinnules following normally distributed. At the point where the arm divides an entoparasitic gastropod is encysted, and Gislén suggested that it was possibly the irritation caused by the parasite that stimulated the pinnule to develop into an arm, though he noted that many other specimens of the same species had encysted gastro- pods without arm division having taken place. For some reason or other there are no soft parts on the eighth-twelfth brachials, though the distal part of the original arm still has an ambulacral furrow. He said that this is the most probable cause of the hypertrophy of P; on the original arm. The fiow of body fluids to the distal parts of the main arm has been cut off, and P; has received the surplus and so has developed into a complete arm. It is to be noted, however, that no such abnormal flow of body fluids can be adduced in the three cases following. In a specimen of Comanthus pinguis from Mortensen’s station 10 one postradial series has two IIBr 4(3-++4) series; the inner arms from each of these are undivided. One of the outer arms has a II[Br 4(3+-4) series, the other a I1IBr 8(3-++4, 5+6, 7+-8) series. In a specimen of Cyllometra pulchella (=manea) from Mortensen’s station 10 one ray bears an undivided arm on the right and a I1Br 8(3 +4) series on the left. In a specimen of Neometra multicolor from Mortensen’s station 24 one postradial series has two IIBr 2 series, that to the left bearing two normal undivided arms and 8 BULLETIN 82, UNITED STATES NATIONAL MUSEUM that to the right bearing externally a normal undivided arm and internally a I1IBr 11 (3-4-4) series, on which the outer arm has the proximal syzygies between brachials 4+-5, 9410, 13-+14, and 17+-18, and the inner has them between brachials 2+3, 7+-8, and 11-+-12. Gislén said that in the first two cases it is an inner pinnule that has developed into an arm. The first example is perhaps the most remarkable, presenting a picture of the repetition, so far as possible normal, in more distant parts of the arm of the proximal arm ramification and its distribution of nonmuscular articulations. Gislén noted that Mortensen (1920) has described a case of P; developing into an arm in Antedon petasus, and that the author has described and figured an arm of Thaumatocrinus sp. forked twice in its distal portion. The proximal fork, he said, seems to be due to a mere splitting of the growing point, but the distal branch seems to have been caused by an hypertrophied pinnule. Gislén said we have already seen how the new arms in the genus Metacrinus develop in normal cases on the inner side of a main arm, and asks: What, then, is the condition at the first ramification, the [Br axillary? Here there can be no question of the new arm developing on the inner side of an arm branch. A comparison of the material that has been considered shows that in a great majority of cases it is the right hand pinnule that has been strengthened and has become equivalent to the main arm, corresponding to a IBr 7(1+2, 4+-5) series in M. interruptus. In 9 specimens of M. nobilis tenuis the right pinnule is strengthened into an arm in 24 cases, the left pinnulein 11. In7 specimens of M. rotundus 25 right and 11 left pinnules are strength- ened into arms. In 19 specimens of M. interruptus 73 right and 15 left pinnules are developed into arms. In all 122 right pinnules as against 37 left pinnules have been strengthened. Gislén said that the reason for this may be ascertained. In by far the greater number of cases the first pinnule is on the right side of the second brachial— in other words it is the right branch that is suppressed at the first ramification. When the first real arm ramification at last takes place, it is on the right side that the sup- pressed impulse to arm formation is stronger, and therefore it is oftener a right than a left pinnule that develops into an arm. In 9 specimens of M. nobilis tenuis P, is on the right side of the second brachial in 24 cases out of 35, in 7 specimens of M. rotundus in 26 cases out of 30, and in 21 specimens of M. interruptus in 85 out of 100— in all, in 135 cases out of 165. Thus it is the right arm that is suppressed in the genus Metacrinus. In the genera of the family Hyocrinidae, according to Gislén, Calamocrinus has its first arm ramification after the first pinnule, while the other genera have simple arms. Examples of the arm base in Calamocrinus up to and including the first axillary are: Br 10(1+-2, 546, 7+8), with pinnules on brachials 4 (left), 6, 8, 9; and IBr 10 (1+2, 5+6, 8+9), with pinnules on brachials 4 (left), 6, 7, 9. The first pair of pin- nules is thus completely suppressed, and the first pinnule develops in the great majority of cases to the left of the fourth brachial—the suppressed first pinnule would also have appeared on the left (of the second brachial) if it had developed. The first arm ramification is also usually caused by the strengthening of a left pinnule. Thus here also there is a connection between the development of the first pinnule and of the first arm ramification similar to that which may be made out in the genus Metacrinus. A MONOGRAPH OF THE EXISTING CRINOIDS 9 The difference is that P, and the first arm ramification developed to the right in Metacrinus, while they are found on the left in Calamocrinus. The other genera of the family Hyocrinidae are, like Calamocrinus, defective in their pinnulation. Information regarding the position of the pinnules is sporadic and incomplete. A reconstruction of the defective rows of pinnules shows, however, that in the great majority-of cases the suppressed P; must have been on the left of the second brachial. A specimen of Ptilocrinus pinnatus examined had, up to the eleventh brachial, on all five arms syzygies between brachials 1-+2, 5+6, and 8+9, with pin- nules on brachials 4 (left), 6, 7, 9, 10, and 11. Of Gephyrocrinus there are only three specimens known. The pinnulation and distribution of the syzygies in the arm bases up to and including the sixth brachial are for the most part as in Ptilocrinus. The first pinnule is on the fourth brachial. According to the description and figures the first (suppressed) pinnule was on the left of the second brachial in 10 cases out of 11. In Thalassocrinus the first developed pinnule is on the fifth brachial, and on the left. In Hyocrinus bethellianus the first pinnule is on the left of the sixth brachial. Accord- ing to Gislén Carpenter’s figure of this species is reversed. In the Phrynocrinidae the genus Nawmachocrinus is known only from a single very defective specimen. In Phrynocrinus the first arm ramification is at about the twentieth-twenty-fifth brachial. The lowest pinnule is on the seventh or eigth brachial, to the right in two cases and to the left in three. A reconstruction of the pinnulation on the five rays of the only known specimen of Phrynocrinus nudus shows that the (suppressed) P,; was on the right of the first epizygal in three cases, and on the left in two. The family Bathycrinidae is represented in the recent seas by six genera, of which three have 10 arms with [Br series of two elements, and three have five undivided arms—Rhizocrinus, Bythocrinus, and Democrinus. The genera of Bathycrinidae fall into two series, one with short and coalesced basals, the other with very long basals joined by suture. Gislén remarked that it is interesting to note that the position of the first pinnule is different in these two series. In Rhizocrinus lofotensis the first developed pinnule appears almost always on the eighth brachial, and usually to the right—in 22 cases out of 30 in specimens from the Trondhjem Fjord. P, would therefore, if developed, usually appear on the left of the second brachial. In the genus Bythocrinus, on the other hand, the arrangement is different. In Bythocrinus cf. braueri in the four specimens studied the first pinnule in 9 cases out of 11 appears to the left on the fourth brachial, or to the right on the sixth brachial. Therefore P, would, if developed, usually appear on the right of the second brachial. In the comatulids four genera have undivided arms. In the only one of these four genera belonging to the oligophreate type, Hudiocrinus, the first pinnule is invariably on the left, whereas of the three remaining genera, belonging to the Macro- phreata, at least two have the first pinnule to the right in the great majority of cases. The three macrophreate genera with undivided arms are Thawmatocrinus, with 10 arms, and Pentametrocrinus and Atopocrinus, each with 5 arms. In Thaumatocrinus Gislén found that the first pinnule is as often to the right as to the left in the species he was able to study. In the specimen of 7. jungersenit in the Upsala Museum P, is on the right of the second brachial in five cases out of nine. 208244492 10 BULLETIN 82, UNITED STATES NATIONAL MUSEUM In 10 additional specimens in the Copenhagen Museum the first pinnule is on the right of the second brachial in 50 cases, and on the left in 48. Gislén said that it might be supposed this peculiarity might be accounted for by the genus being 10- rayed; but the same distribution may also be seen in the 5-rayed young. In Thau- matocrinus renovatus P, appears to the right of the second brachial in three cases and to the left in two. The position of P, also varies in a young individual of 7. gungersent. In Pentametrocrinus and Atopocrinus the first syzygy is usually between brachials 4-5, and the first pinnule is on either the second or the fifth brachials. Carpenter said that in Pentametrocrinus japonicus the lowest pinnule (on the fifth brachial) appeared to the right in 11 cases out of 12. P, also appeared to the right on the second brachial in P. varians. It is also stated to be commoner on the right than on the left in P. semperi. Gislén examined two specimens of P. diomedeae; in these the lowest pinnule was on the right in 9 cases out of 10. Judged from Koehl- er’s figures the same seems to be the case in P. atlanticus. In Atopocrinus sibogae P, in 4 cases out of 5 is to the right on the second brachial. Gislén gave some examples of the suppression of a right or a left arm in other comatulids. The examples given showed a tendency for the suppression to be located differently in certain oligophreate families on the one hand and certain macrophreate types on the other. In a 9-armed specimen of Comatula pectinata from Java the left arm, as in Hudio- crinus, is rudimentary. It is represented by a small calcareous lump of two ossicles united by syzygy. The right arm has the first 6 brachials united in 3 syzygial pairs. Springer described a 9-armed specimen of Uintacrinus socialis in which a left arm is suppressed into a pinnule. Gislén found a similar case in a 9-armed specimen of Antedon petasus. On the undivided arm the second and third ossicles are united by synarthry, and the fourth and fifth and tenth and eleventh are united by syzygy. The undivided arm is as stout as the others. On both sides of the arm the first gonad is on the third pmnule. The unusual distribution of nonmuscular articulations is the same as in the specimen of Uintacrinus socialis just mentioned. According to Gislén these cases may be ex- plained in two ways—either the reduction has gone so far that both the suppressed arm and the ossicle to which it was attached have disappeared, or the arm alone has been suppressed, the [Br, and IBr, having been coalesced into a single ossicle. Since there is a muscular articulation between the first and second ossicles following the radial, the second alternative is believed by Gislén to be the more probable. In one of these cases it is clearly a right arm and in the other a left arm that has been sup- pressed. Gislén said it is to be desired that more data regarding the pinnulation of similar forms with undivided postradial series should be made available in order to enable us to decide whether the tendency to suppression is more frequent in one or the other, or, as in Thawmatocrinus, equally frequent in both. Gislén noted that during ontogenetic development the first arm ramification is formed by two equally stout arms. Sometimes, however, it seems that a difference in length between the two arms which is very insignificant, and in the larger young soon disappears, arises, at least in certain cases. Thus we have here a difference in length between the two arms of a pair, not a difference between pairs of arms such as is illustrated and described by Perrier. This last, indeed, seems not to occur. The A MONOGRAPH OF THE EXISTING CRINOIDS 11 slicht differences between the two arms of a pair has heretofore scarcely been discussed, only occasionally illustrated, and possibly sometimes overlooked, because of its transitory nature. Mortensen, indeed, denied it. When it is marked it seems to be strongest in arms with from 10 to 15 brachials. Observations made by Gislén on very young individuals of Asterometra anthus and Stenometra dentata point to the conclusion that one arm, usually the right arm, is less favored in the earlier stages of development. It seemed possible to Gislén that in reproductive arm regeneration the growth of one or the other arm might be more or less favored, and that this might open up a possibility of drawing certain conclusions regarding the tendency to suppression. In order to determine whether this could be established for regenerates from the IBr, Gislén carried out experiments in regeneration with Antedon petasus at the Kristineberg station. Owing to a number of unfortunate accidents he obtained only 10 individuals with regenerated arms out of all the hundreds operated upon. In 1918 he amputated the postradial series between the IBr, and IBr, in 15 individuals. The animals so treated were placed in a fairly large box with a lid perforated with holes 10 mm. in diameter. The box was sunk off Blabergsholmen, Kristineberg, at a depth of about 15 meters and was anchored by a line to a crevice in a rock just above high-water mark. When it was taken up a month later it was found that all the individuals but one had escaped through the perforations. Fine-meshed netting was then nailed over the perforations, and in 1919 the box was again placed in position, this time with 50 amputated individuals. The strong September storms, however, tore away the cable, and the box was lost. In 1922 the same procedure was repeated with a new box, but the circulation within the box was too weak and the animals died. In 1923 he secured nine individuals with regenerated arms. The regenerates obtained, however, did not give a final solution to the problem, even though they furnished certain indications. The speci- men obtained in 1918 had the anterior ray amputated. Of the two regenerated arms the left was 1.5 mm. long, with 8 (9) brachials, and the right was 1.2 mm. long, with 7 brachials. But an opposite result was shown by one of the nine regenerates obtained in 1923. In this also the anterior postradial series is regenerated. The left arm is distinctly the shorter, 1.4 mm. long, with 10 brachials, the right arm being 1.6 mm. long, with 11 brachials. No pinnules have developed. Four other individuals showed either no difference or an almost imperceptible difference in the length of the arms. The first of these, with the right posterior postradial series amputated, had 3 brachials on each side of the regenerated IBr axillary. The other three, two with the anterior and one with the left anterior ray regenerated, had almost exactly equal arms with from 15 to 20 brachials, supplied with from two to four pairs of minute pinnules distally. About 12 proximal brachials lacked pinnules, as in the young, though sometimes there were rudiments of P, and P,. Of the four remaining regenerates three had only three or four regenerated segments developed. In two of these the regen- erated ray was the left anterior, and in one it was the anterior A minute ramification consisting of a single segment appeared in one case to the right, and in one case to the left. The last individual had regenerated in the left anterior radius an undivided arm 2.5 mm. long with the tip bent to the left consisting of about 11 segments. 12 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Gislén said that judged from the slight available information regarding undivided arms in Antedon the indications are that there is here the same variability as in Thaumatocrinus. In summing up Gislén said that if the first ramification is suppressed this process occurs differently in different families, but with a certain regularity in closely related forms. Usually it can then be demonstrated how the tendency to suppression affects predominantly either the right or the left ray, which becomes a pinnule or disappears altogether. If the arm is defective in its pinnulation the first pinnule and, beyond, the first arm branch, appear as a rule on the same side of the arm as that on which the suppressed first pinnule would have been according to the reconstruction. In the family Hyocrinidae usually both the suppressed and the lowest pinnule, and in Calamocrinus the first arm ramification, are to the left. Among the pentacrinites Metacrinus has both the lowest pinnule and the first arm branch to the right. In the Phrynocrinidae too little material is available to permit judging the position of the first pinnule. The family Bathycrinidae may be divided into two natural groups. To one group belongs Rhizocrinus, which has the first developed pinnule to the right on the eighth brachial, though the suppressed first pinnule was to the left on the second brachial. According to Gislén this type is the only exception to the rule that the lowest pinnule appears on the same side of the arm as the suppressed first pinnule. In the other group, exemplified by Bythocrinus, the suppressed first pinnule is usually to the right. Among the comatulids, in Hudiocrinus Pe, the lowest pinnule, is on the left. Where arms have been suppressed a similar tendency seemed also to show itself in a few examples in the families Comasteridae and Uintacrinidae. An opposite tendency is seen in Pentametrocrinus and Atopocrinus in which P, usually appears to the right. In some young of certain species of Thalassometridae the right arms were often for a time more weakly developed than the left. In the genus Thawmatocrinus the position of P, is variable, and this is also the case in a few individuals of Antedon petasus with undivided postradial series. Regenerates of this species pointed in the same direction. Gislén said that a connection, in the sense that the families having their lowest pinnule developed in the same way are more closely related, can not be assumed if there are no further facts forthcoming to support such a supposition. He remarked that the Thalassometridae and certain Macrophreata agree in a number of features, and it is probable that they are rather closely related. Possibly also, according to Gislén, Thaumatocrinus is more closely related to the Antedoninae than it is to Penta- metrocrinus. At least the majority of the comatulids presumably have not developed from forms with undivided postradial series. It may be supposed that forms with undivided postradial series arose in different comatulid families at different times, and are thus only parallel types. The fact that in certain Zygometridae (Hudiocrinus) and in the Comasteridae and Uintacrinidae the left arm branch is less strongly developed or suppressed, and in certain Macrophreata and in some of the Thalassometridae it is the right arm branch that is less strongly developed or suppressed seemed to Gislén to support Kirk’s assumption that the comatulids are of polyphyletic origin. A MONOGRAPH OF THE EXISTING CRINOIDS 13 REVERSIBILITY OF DEVELOPMENT Gislén maintained that pinnulate forms have developed phylogenetically from nonpinnulate types with extensive arm ramification. The holotomic recent types are descended from the metatomic. Pinnules are thus simplified armlets or ramiculi. That the tendency toward arm ramification is in all cases present, though it may be latent, is shown by the fact that pinnules in certain cases abandon their character of small unramified arms and develop into complete arms with pinnules. Gislén said the appearance of pinnules can most easily be explained by assuming the existence of obstructive factors on the removal of which the pinnule again develops into a ramified arm, the phylogenetic development having presumably involved a great number of obstructive mutations. From this point of view the appearance of axillaries with two pinnules and cases of abnormal pinnulation otherwise difficult to explain are easily accounted for as examples of two sided and reversed obstruction respectively. In Metacrinus also obstruction of a temporary nature may be said to occur, retarding the development of the inner arms so that they remain in the form of pinnules for a fairly long time. Dollo in 1893 laid down the law of irreversibility of development, which was later formulated by Abel as follows: ‘‘An aborted organ never recovers its former strength, and an organ that has disappeared never reappears—at least not developed in the same way.” Gislén said that, broadly speaking, this rule is certainly correct. But it is also clear that the facts recorded and the conclusions to which they lead suggest that exceptions may be found to Dollo’s law. From the point of view of heredity it may be maintained that Dollo’s law is valid if the tendency to an organ has really disappeared—in that case a return to the original type would be out of the question. When a return really appears this may be interpreted as the removal of a factor of obstruction. Gislén said that this idea might perhaps be considered from the paleon- tological standpoint as rather perilous, though in reality it is not. As a matter of fact such reversibility presumably occurs very seldom in nature. The reason for this would seem to be that the type of organization reached is the one that is most suitable from the point of view of selection and structure, and most economical; and it has often been reached in a roundabout way by an infinity of adaptations. That in certain cases the conditions necessary for ‘“Riickschlage”’ really occur is shown, however, by the specimen with hypertrophied pinnules. Hypothetically the conditions might be pictured thus. In any collection of some thousands of comatulids there are always a number with one or another of their pinnules replaced by a more or less ramified arm. Thus in 270 specimens of Antedon petasus that Gislén secured at.a single haul of the dredge off the Kristineberg zoological station at Flatholmen, three showed abnormal arm ramification. Assuming that these variations were of any advantage from the point of view of selection, and provided that they were inheritable, the old character would appear normally in all individuals after some few generations. Gislén said that there seems actually to be a good example of a reversion of this kind in the phylogeny of the comatulids. At one time, toward the end of the Paleo- zoic, the proximal pinnules were suppressed by the disk, which rose high up between the arms. They reappeared when the disk again retreated. But, according to Gislén, there is still a relic of this development in the gap that appears in the pinnula- 14 BULLETIN 82, UNITED STATES NATIONAL MUSEUM tion of young comatulids, though in most fully grown individuals it is filled in by the subsequent development of the proximal pinnules. MUSCULAR ARTICULATIONS Gislén studied in great detail the brachial articulations of the crinoids and their significance, his conclusions differing somewhat from mine. He pointed out that when examining the obliqueness of an articulation two different kinds of obliquity must be distinguished. If an arm be viewed from the dorsal side it is noticed at once that the lines that mark the articulations do not run at right angles to the longi- tudinal axis of the arm, but cross the arm more or less obliquely. This Gislén called exterior obliqueness. On the articular faces of the brachials the dorsoventral crest and the transverse (fulcral) ridge do not usually meet at a right angle, with the result that the muscles and interarticular ligaments of the two sides are not of the same size. This he called interior obliqueness. A straight muscular articulation is characterized by two points of contact at the same distance from the mediodorsal line; an oblique muscular articulation has a distinct dorsolateral point of contact, and a less distinct ventrolateral one. This difference is obvious only in those types in which the course of the articular lines across the dorsal surface of the arm is fairly oblique; in those comatulids in which the articular lines run nearly at right angles to the longitudinal axis of the arm it is in- distinct, or not visible at all. As this feature stands in a certain relation to the exterior obliqueness of the brachials, Gislén believed that there would be a certain amount of interest in ascertaining the degree of obliqueness in the course of the articular lines across the dorsal side of the arms in the comatulids. Comissia parvula, arms 45 mm. long: (15)! 17 brachials 1.1 mm. broad in each 10 mm. in the proximal part of the arm, and (12) 15 brachials 0.6 mm. broad in the distal; relation of longer side of the brachials to width 1:1 to 3:2 angle of the distal articulations with the longitudinal axis of the arm 66-72°. Comissia peregrina magnifica, arms 160 mm. long: (9) 12 brachials 2.5 mm. broad in each 10 mm. in the proximal part of the arm, and (12) 16 brachials 1.2 mm. broad in the distal; relation of longer side of the brachials to width 1:2 to 2:3; angle of the distal articulations with the longitudinal axis of the arm 69-77°. ; Comatula solaris, arms 100-+-mm. long: (9) 10 brachials 3.5 mm. broad in each 10 mm. in the proximal part of the arm, and (13) 15 brachials 2.0 mm. broad in the distal; relation of longer side of the brachials to width 1:2; angle of the distal artic- ulations with the longitudinal axis of the arm 72-76°. Comatula pectinata, arms 70 mm. Jong: Relation of longer side of the brachials to width 2:3; angle of the distal articulations with the longitudinal axis of the arm 67-79°. Capillaster sentosa, arms 80 mm. long: (11) 11 brachials 2.4 mm. broad in each 10 mm. in the proximal part of the arm, and (13) 15 brachials 1.7 mm. broad in the distal; relation of longer side of the brachials to width 1:2; angle of the distal articu- lations with the longitudinal axis of the arm 80-89°. Comantheria delicata grandis, arms 105 mm. long: (7) 8 brachials 3.0 mm. broad in each 10 mm. in the proximal part of the arm, and (11) 14 brachials 1.3 mm. ' The figures in parentheses show the number of segments when the syzygial pairs are counted as units. A MONOGRAPH OF THE EXISTING CRINOIDS 15 broad in the distal; relation of longer side of the brachials to width 1:2 to 2:3; angle of the distal articulations with the longitudinal axis of the arm 73-82°. Comanthus japonica, arms 120 mm. long: (9) 10 brachials 2.4 broad in each 10 mm. in the proximal part of the arm, and (18) 22 brachials 1.6 mm. broad in the distal; relation of longer side of the brachials to width 1:5 to 1:4; angle of the distal articu- lations with the longitudinal axis of the arm 87—90°. Zygometra elegans, arms 105 mm. long: (16) 16 brachials 1.7 mm. broad in each 10 mm. in the proximal part of the arm, and (22) 24 brachials 1.3 mm. broad in the distal; relation of longer side of the brachials to width 1:4; angle of the distal articu- lations with the longitudinal axis of the arm 82-90°. Eudiocrinus loveni, arms 45 mm. long: Proximal brachials 1.4 mm. broad; (11) 14-15 brachials 0.8 mm. broad in each 10 mm. in the distal part of the arm; relation of the longer side of the brachials to width 1:1 to 6:5; angle of the distal articulations with the longitudinal axis of the arm 75-80°. Amphimetra tessellata, arms 120 mm. long: (11) 138 brachials 2.4 mm. broad in each 10 mm. in the proximal part of the arm, and (25) 28 brachials 1.5 mm. broad in the distal; relation of longer side of the brachials to width 1:5; angle of the distal articulations with the longitudinal axis of the arm 86-90°. Himerometra magnipinna, arms 105 mm. long: (13) 13 brachials 1.7 mm. broad in each 10 mm. in the proximal part of the arm, and (22) 23 brachials 0.8 mm. broad in the distal; relation of longer side of the brachials to width 1:4 to 1:3; angle of the distal articulations with the longitudinal axis of the arm 81-89°. Stephanometra spicata, arms 130 mm. long: (11) 12 brachials 1.5 broad in each 10 mm. in the proximal part of the arm, and (15) 16-17 brachials 1.1 mm. broad in the distal; relation of longer side of the brachials to width 1:2; angle of the distal articula- tions with the longitudinal axis of the arm 78-87°. Liparometra grandis, arms 120 mm. long: (13) 14 brachials 2.0 mm. broad in each 10 mm. in the proximal part of the arm, and (16) 18 brachials 1.2 mm. broad in the distal; relation of longer side of the brachials to width 1:4 to 1:3; angle of the distal articulations with the longitudinal axis of the arm 82-89°. Cenometra bella, arms 90 mm. long: (17) 20 brachials 0.8 mm. broad in each 10 mm. in the distal part of the arm; relation of the longer side of the brachials to the width 2:3 to 3:4; angle of the distal articulations with the longitudinal axis of the arm 78-82°. Cyllometra disciformis (=manca), arms 90 mm. long: (14) 17 brachials 1.2 mm. broad in each 10 mm. in the proximal part of the arm, and (15) 20 brachials 0.7 mm. broad in the distal; relation of the longer side of the brachials to width 1:1 to 4:3; angle of the distal articulations with the longitudinal axis of the arm 72-82°. Tropiometra encrinus, arms 105 mm. long: (9) 11 brachials 2.2 mm. broad in each 10 mm. in the proximal part of the arm, and (14) 16 brachials 1.0 mm. broad in the distal; relation of the longer side of the brachials to width 3:4; angle of the distal articulations with the longitudinal axis of the arm 80-85°. Pectinometra flavopurpurea, arms 65 mm. long: (12) 13 brachials 1.0 mm. broad in each 10 mm. in the proximal part of the arm, and (15) 17 brachials 0.7 mm. broad in the distal; relation of the longer side of the brachials to width 5:6 to 1:1; angle of the distal articulations with the longitudinal axis of the arm 64—72°. 16 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Asterometra anthus, arms 85 mm. long: (12) 14 brachials 1.6 mm. broad in each 10 mm. in the proximal part of the arm, and (15) 18 brachials 1.1 mm. broad in the distal; relation of the longer side of the brachials to width 1:2 to 2:3; angle of the distal articulations with the longitudinal axis of the arm 69-77°. Diodontometra bocki, arms 60 mm. long: (11) 12 brachials 1.3 mm. broad in each 10 mm. in the proximal part of the arm, and (14) 15 brachials 0.7 mm. broad in the distal; relation of the longer side of the brachials to width (3:4 to) 1:1; angle of the distal articulations with the longitudinal axis of the arm 69-75°. Antedon petasus, arms 75 mm. long: (12) 15 brachials 1.2 mm. broad in each 10 mm. in the proximal part of the arm, and (14) 17 brachials 1.0 mm. broad in the distal; relation of the longer side of the brachials to width 1:1; angle of the distal articulations with the longitudinal axis of the arm 64-75°. Heliometra glacialis, arms 200 mm. long: (7) 9 brachials 3.2 mm. broad in each 10 mm. in the proximal part of the arm, and (12) 15 brachials 1.5 mm. broad in the distal; relation of the longer side of the brachials to width 1:2 to 2:3; angle of the distal articulations with the longitudinal axis of the arm 68-78°. Poliometra proliza, arms about 100 mm. long: (9) 12 brachials 1.6 mm. broad in each 10 mm. in the proximal part of the arm, and (11) 14 brachials 1.0 mm. broad in the distal; relation of the longer side of the brachials to width 1:1; angle of the distal articulations with the longitudinal axis of the arm 66-—74°. Pentametrocrinus diomedeae, arms 90 mm. long: (10) 13 brachials 0.6 mm. broad in each 10 mm. in the proximal part of the arm, and (11) 14 brachials 0.5 mm. broad in the distal; relation of the longer side of the brachials to width 3:2 to 2:1; angle of the distal articulations with the longitudinal axis of the arm 59-66°. Gislén said this study showed that distinctly oblique articulations, that is, artic- ulations in which the articular line makes an angle with the longitudinal axis of the arm of from 59° to 79° (average 71°), are found in the Comasteridae (exclusive of the Comasterinae and Capillaster), Thalassometridae, Charitometridae, and Calome- tridae, and in the Macrophreata; while on the other hand indistinctly acute to right angles of from 72° to 90° (average about 83°) are found in the Comasterinae, Capil- laster, the Zygometridae, Himerometridae, Mariametridae (including the Stephano- metridae), Tropiometridae, and Colobometridae. He said that from the figures given it may also be seen that the number of brachials for each 10 mm. and their relative length has a certain relation to the thickness of the arms—that is, generally with the size of the animal. In the majority of cases, how- ever, the greatest number of segments in each 10 mm. and the shortest segments with- out comparison are found in the group in which the articular line makes nearly a right angle with the longitudinal axis. The figures given are from the middle or distal parts of the arms. The figures following give the size of the angles in the proximal parts of the arms. As in the previous cases the figures are taken from the distal borders of the brachials and refer to the angles of the pinnulate side. Figures omitted denote hypozygals. Comanthus japonica: I1Br,-80°; IIIBr.-78°; brachials 1-110°; 2-87°; 4-84°; 5-90°; 6-84°; 7-80°; 8-78°; 9-76°; 10-78°; 11-78°; after the twenty-third brachials the angles are larger, up to 85° or 90°. A MONOGRAPH OF THE EXISTING CRINOIDS 17 Eudiocrinus loveni: TBr,-89°; brachials 2-88°; 4-85°; 5-83°; 6-74°; 7-69°; 8-60°; 10-69°; 11-64°; 13-66°. Amphimetra tessellata: Brachials 2-86°; 4-92°; 5-96°; 6-100°; 7-101°; 8-95°; 10-90°; 11-87°; 12-84°; 13-83°; 14-88°; then (81°) 86-90°. Pontiometra andersoni: Brachials 2-88°; 4-91°; 5-92°; 6-94°; 7—-93°; 8-93°; 9-90°; 10-89°; 11-85°; 12-78°; 13-74°; 14-80°; distal brachials 80-90°. Pectinometra flavopurpurea: Brachials 2-80°; 4—80°; 5-80°; 6-78°; 7-78°; 8-76°; 9-76°; 10-71°; 11-72°; 12-73°; 14-72°. Asterometra anthus: Brachials (1—about 100°); 2-84°; 4-87°; 5-92°; 6-95°; 787° 8-840 9=77°? 10-6995 12-719 "13--70°: Diodontometra bocki: Brachials 2-83°; 486°; 5-87°; 6-93°; 7-96°; 8-97°; 9-93°; 10-91°; 11-90°; 12-81°; 13-78°; 14-73°; 15-75°; 16-73°. Clarkometra elegans: Brachials (1-103°); 2-68°; 4-87°; 5-105°; 6-106°; 7-83°; 860°; 10-69°; 11-65°; 12-64°; 13-62°; 15-65°; then about 64°. Heliometra glacialis: Brachials 2-67°; 4-87°; 5-95°; 6-105°; 7-106°; 8-108°; 10-108°; 11-96°; 12-87°; 13-75°; 15-82°; 16-70°; 17-74°; then about 70°. Hypalometra defecta: Brachials 2-78°; 4-96°; 5-100°; 6-97°; 7-86°; 8-80°; 10-62°. Thaumatocrinus jungerseni: Brachials 2-98°; 3-101°; 5-93°; 6-90°; 7-86°; 8-82°: 9-77°: 11-62°; 12-63°; 13-58°; 14-63°; 15-61°; 17-60°. Pentametrocrinus diomedeae: Brachials 2-98°; 3-98°; 5-88°; 6-84°; 7-80°; 8-78°; 10-60°; 11-66°; 12-53°. Ina larger specimen: Brachials 2—98°; 3-96; 5-92°; 6-88°. Atelecrinus helgae: Brachials 2-64°; 4-67°; 572°; 6-67°. Metacrinus nobilis tenuis: IBr,-82°; IBr,-90°; IIBr,-90°; I1Br.-82°; IIBr,-80°; I1Br,-88°; IIBr,-84°; Il1Br,-98°; TIIBr.-91°; IL[TBr.-85°; TIIBr;_y.-about 90°; IVBr,-102°; IVBr,-99°; IVBry_;)-about 90°; brachials 1—93°; 2-98°; 4 and following about 90°. Gislén said that from the preceding figures, which might have been supplemented with others from about 40 additional specimens examined but not tabulated, all of which point in the same direction, we learn the following: The first two brachials following an axillary always have a relatively long outer and short inner side. The distal articular line of these brachials therefore makes a relatively large angle with the proximal. The arms, which following the arm division would diverge widely, are thus brought closer together, which because of the limited space is of considerable importance. In reality the length of the outer sides of these brachials is so much greater than that of the inner sides that the distal articular line of the second brachial slopes inward. In Metacrinus the proximal and distal margins of the brachials after the first or second postaxillary are fairly parallel, but in the other types examined a more or less pronounced obliqueness usually sets in sooner or later. In the first syzygial pair (composed usually of brachials 3-+-4) this obliqueness is generally only slight—the greater length of the inner side in this pair has counterbalanced almost completely the inward slope that was apparent on the distal margin of the second brachials. The distal edge of the fourth brachial is therefore very often nearly at a right angle to the longitudinal axis of the arm. Subsequently there usually follow a greaterfor lesser number of brachials in which the proximal and distal edges show a tendency 18 BULLETIN 82, UNITED STATES NATIONAL MUSEUM to be more or less parallel, then the obliqueness increases again after the seventh— thirteenth brachial, reaching its maximum at from the fifteenth to the twentieth brachials. In the middle portion of the arms the obliqueness decreases, but it increases slightly again in the distal portion. THE REVERSION PHENOMENON This exterior obliqueness manifests itself by the acute angle formed by the articular line with the longitudinal axis of the arm, facing now outward and now inward. The brachials are therefore longer alternately on the outer and inner sides. Usually the pinnule is attached to the longer side of the brachial; this is always the case in the middle and distal portions of the arms. In the proximal portion of the arms Gislén pointed out that the opposite is often found—the pinnules arise from the shorter side of the brachials. He called this reversion of the articulations. This feature is lacking, or is very little developed, in certain Comasteridae, Eudiocrinus, Calometridae, and Atelecrinus, in which the pinnular side of the brachials is the longer along the entire arm. Reversion is most conspicuously developed in the Antedonidae and Charitometridae. The second, and usually also the fourth, brachials, bearing P,; and P, are more strongly developed on the pinnular side, but from the fifth brachial onward reversion appears in a variable number of segments, to be gradually effaced and followed in the middle of the arm by the normal condition. Reversion is only slight in Capillaster sentosa, Comanthus japonica, Tropiometra afra macrodiscus, and Pterometra trichopoda, in which the fifth brachial is only very slightly, sometimes not perceptibly, narrower on the pinnular side, and the sixth brachial has parallel ends, or an obviously longer pinnular side. The same is the case in Notocrinus virilis, in which the sixth or seventh brachial has a longer pinnular side, and in Clarkometra elegans and Stephanometra spicata in which the seventh or eighth brachial is longer on the pinnular side. The reversion appears more distinctly in Comantheria delicata grandis, in which the seventh brachial has a longer pinnular side, in Stenometra diadema, Cyllometra manca, and Himerometra magnipinna, in which the seventh or eighth brachials have longer pinnular sides, in Zygometra elegans, Heterometra crenulata, and Asterometra anthus, in which the seventh—ninth brachials have longer pinnular sides, and in Liparometra grandis, in which the eleventh brachial first shows a longer pinnular side. This phenomenon appears most conspicuously, however, in Pentametrocrinus dio- medeae, in which brachials 4+-5 have a longer pinnular side, in Jsometra vivipara and Hypalometra defecta in which the eighth brachial, in Crossometra septentrionalis and Antedon petasus, in which the seventh-ninth brachials, in Diodontometra bocki, in which the tenth—-twelfth brachials, in Heliometra glacialis, in which the twelfth brachial, and in Promachocrinus kerguelensis and Monachometra cf. fragilis, in which the thir- teenth brachial first has the pinnular side longer. Gislén said one might suppose that the slenderness of the pinnule bases had some- thing to do with this phenomenon, and this idea seems to be favored by the Calome- tridae, Pectinometra flavopurpurea and Neometra multicolor having the pinnular side of the brachials longer from the arm base outward. But the lack of reversion in Coma- tula solaris, Comanthus parvicirra, and Atelecrinus cannot, however, be explained in A MONOGRAPH OF THE EXISTING CRINOIDS 19 this way. The two first have slender pinnules, and in the last pinnules are entirely lacking in the proximal portion of the arm. OBLIQUITY OF THE JOINT FACES, OR INTERIOR OBLIQUENESS In order to determine the degree of obliqueness in the angle between the transverse ridge and the radial intermuscular fossa (the dorsoventral axis of the ossicle), Gislén measured the angle between them in series of brachials from the arm bases tomore distal portions of the arms. The angle measured was always that of the antipinnular side, and unless otherwise stated was on the distal articular facet. Comanthus japonica: IIBr2,, 75-76°; IIIBr., 68—72°; brachials 2, 60°; 4, 52°; 5, 47°: 6, 57°; 7, 58°; 8, 83°; 9, 87°; 11, 85°; 12, 85°; 13, 86°; 14, 89°; 16, 87°; 17, 87°; 18, 82°; 19, 86°; 20, 86°; 22, 81°; 23, 82°; 24, 81°; 25, 83°; 27, 78°; 28, 82°: 29, 76°: 30, 80°; 31, 81°; distal brachials, 60-76°. Eudiocrinus indivisus: [Br2, 75°; brachials 2, 64°; 4, 54°; 5, 56°; 6, 47°; 7, 60°; 9, 69°; 10, 78°; 11, 80°; 12, 82°; 18, 86°; 14, 87°. The arm was regenerated from the tenth brachial. Pontiometra andersoni: An arm arising from a VBr axillary: Brachial 2, 66° or 67°: An arm arising from a IVBr axillary: Brachials 4, 62-67°; 5, 61°; 6, 66°; 7, 62°; 8, 68°; 9, 67°; 10, 68°; 11, 72°; 12, 77°; 18, 78°; 14, 80°; 15, 81°; 16, 84°. From another series, probably beginning with the ninth brachial: Brachials 9, 69°; 10, 72°; 11, 74°; 12, 84°; 13, 86°; 14, 83°; 15, 84°; 16, 87°; 17, 87°; 18, 90°; 19, 85°; 20, 87°. Pectinometra flavopurpurea: Brachials 2, 86°; 4, 72°; 5, 80°; 6, 63°; 7, 72°; 8, 78°; 9, 75°; 10, 88°; 11, 85°; 12, 90°; 14, 90°; 15, 97°; 16, 91°; 17, 90°; 18, 93°; 19, 90°: 21, 92°; 22, 91°; 23, 86°; 24, 86°; 25, 87°. Asterometra anthus: Brachials 2, 81-85°; 4, 61°; 5, 64°; 6, 65°; 7, 63°; 8, 74°; 10, 86°; 11, 88°; 12, 80°; 13, 82°. Clarkometra elegans: Brachials 2, 79°, 80°, 80°, 83°, 84°; 4, 67°, 69°, 70°; 5, 65°; 6, 57-61°; 8, 80°; 10, 82°. Antedon petasus: Brachials 2, 71°; 4, 68°; 5, 62°; 6, 65°; 8, 71°; 10, 79°; 11, 81°; 12, 82°; 13, 85°; 15, 86°; 16, 87°; 17, 86°; distal brachials about 80°. Heliometra glacialis: Brachials 2, 71°; 4, 67°; 5, 64°; 6, 64°; 7, 60°; 8, 67°; 10, 65°; 11, 70°; 12, 75°: 13, 83°; 15, 78°; 16, 87°; 17, 79°; 19, 92°; 20, 84°; 21, 86°: 23, 85°: 24, 88°; 25, 85°; 27, 90°; 28, 84°; 29, 89°; 31, 86°; 32, 83°; 33, 84°; 34, 86°; 36, 87°: distal brachials 70-80°. Thaumatocrinus jungerseni: Brachials 2, 76°; 3, 72°; 5, 66°; 6, 64°; 8, 69°; 10, 75°; 12, 74°: 18, 85°; 14, 80°; 15, 83°; 17, 86°. Pentametrocrinus diomedeae: Brachials 2, 62°; 3, 60°; 5, 65°; 6, 62°; 7, 65°; 8, 70°; 10, 72°; 11, 75°; 12, 75°; 18, 82°. Atelecrinus helgae: Brachials 2, 80°, 83°; 4, 87°; 5, 85°; 6, 89°. Llycrinus carpenterii: Brachials 2, 82°; 3, 86°; 5, 89°; 6, 90°; 8, 90°; 9, 89°; 11, 92°; 13, 99°; 15, 90°; 17, 92°; pinnules are developed from the eleventh brachial onward. Rhizocrinus lofotensis: Brachials 2, 90°; 4, 90°; 6, 90°; 8, 84°; pinnules are developed from the eighth brachial onward. Metacrinus nobilis tenuis: IBr2, 84°; IBrz:, 75°; I1Br,, 75°; IIBr., 72°; I1Br,, (393) LIBrs742 > LiBr.; 74°; TILBry, 69°; TLLBr.;' 77°; TLBr; 742;| TITBr;, 74°; 11 Bre, 702; ALEBr;,.71°> TT Br,;'72°; INL Br, 72°? MEBr 79°s LVBr, 71°; LV Brs, 75°; DVBrs 20 BULLETIN 82, UNITED STATES NATIONAL MUSEUM 72°: IVBrs, 69°; IVBry, 70°; IVBr;, 64°; IVBrs, 69°; IVBro, 69°; [VBryo, 76°; brachials 1, 75°; 2, 73°; 4, 80°; 5, 72°; 6, 73°; 7, 69°; distal brachials, 72—75°. From these figures it is evident that as a rule obliquity of the muscular articula- tions is most strongly pronounced immediately following the axillaries, and that the more distal articulations become less oblique, and later in the distal portions of the arms, again somewhat more oblique. An exception to this is found in IJlycrinus carpenterii in which the distal angle of the axillary (as viewed dorsally) is very large. Here the proximal brachials, which also lack pinnules, are only slightly oblique, whether viewed from the dorsal side or from the articular face. As a rule—in the comatulids practically without exception judging from 20 speci- mens examined in addition to those listed above and agreeing with them—the oblique muscular articulations show the following sequence: After the first brachial articula- tions at the base of the arm they tend toward the straight type. Gislén’s figures of the twenty-seventh brachial of Heliometra glacialis and of the thirteenth brachial of Heterometra crenulata show how close the oblique muscular articulation on certain proximal brachials can come to the articulation on the radial articular face. This straight type, according to Gislén, occurs more or less completely developed after a varying number of brachials. It should be noted, however, that discoidal brachials, that is, brachials with externally straight articular lines, appear before the tenth brachial whereas the interior straightness of the articulations does not generally appear until after the tenth or fifteenth brachial. These two types are therefore usually separate, and seldom appear on the same brachial. Gislén remarked that Clarkometra elegans shows, for an antedonid [I consider it a colobometrid], a rather high figure for the angle on the second brachial. He sup- posed that one might imagine this to be caused by the lack of pinnules; but the fourth brachial, which is also nonpinnulate, shows a normal figure. That inner obliqueness can be extraordinarily strong is shown by Pentametro- crinus, in which the first pinnule appears on the fifth brachial. The species of Atele- crinus lack pinnules to about the fifteenth brachial. In spite of this, external oblique- ness is well developed. Inner obliqueness, on the other hand, is very little marked, and straight articulations are met with already at about the sixth brachial. Gislén inferred one can understand from this that in the comatulids the presence or absence of pinnules plays a rather small part in regard to external obliqueness in the proximal portion of the arms. It is of more importance in regard to inner oblique- ness, although it is obvious that in the more proximal parts of the arms other factors appear that influence the obliqueness of the muscular articulations. The low figures of the angles in Kudiocrinus may be caused by the fairly thick pinnule bases. If the morphological reasons for the oblique angle between the intermuscular fossa and the transverse ridge are studied the following facts appear: The difference in size between the two muscular fossae is usually of less importance; the difference in size between the interarticular ligament of the pinnular side and that of the anti- pinnular side is, however, often much greater. This difference reaches its maximum in the antedonids. Through the interaction of muscle, pinnule socket, and interartic- ular ligament the pinnular side of the distal facet of the segment is enlarged ventral to the transverse ridge, and as a result the dorsoventral intermuscular furrow looks as if it were forced over toward the antipinnular side. In reality, however, this is not A MONOGRAPH OF THE EXISTING CRINOIDS 21 entirely so. If one studies the ventral side of an arm from which all the soft parts excepting the ligaments have been removed, one will find that the intermuscular fur- row or crest is always oriented strictly dorsoventrally. It is therefore more correct to express the inner obliqueness of the articulation by saying that muscle, pinnule socket, and interarticular ligament displace the transverse ridge toward the dorsal side of the segment, or the pinnular side of the distal end of the ossicle—or the anti- pinnular side if the ossicle is observed from the proximal end. Gislén examined the distal articular faces of the axillaries to determine whether they should be interpreted some as straight and others as oblique, or as intermediate between the two types. The following figures were obtained by measuring the distal axillary angles (the lesser angle between the transverse ridge and the dorsoventral crest). Comanthus japonica: (IBr,; proximal 87°; IIBr, proximal 78°); I[Br axillary 67-70°; I[1Br axillary 60°. Himerometra magnipinna: I1Br axillary 50°; II1Br axillary 63°. Pontiometra andersoni: [Br axillary 87°; I1IBr axillary 82-85°; IVBr axillary 78-82°; VBr axillary 80°. Tropiometra afra macrodiscus: [Br axillary 80°. Cenometra bella: I11Br axillary 76°. Oligometrides adeonae: Br axillary 77°. Pectinometra flavopurpurea: [Br and IIBr axillaries 90°. Asterometra anthus: [Br axillary 87°. Notocrinus virilis: [Br axillary 87°. Clarkometra elegans: [Br axillary 84°, 84°, 85°, 89°. Heliometra glacialis: [Br axillary 86°. Tlycrinus carpenterii: [Br axillary 74°. Metacrinus nobilis tenuis: Radials 89°; [Br axillaries (4) 70°; IIBr axillaries (7) 63°; I1]Br axillaries (11) 67°; IVBr axillaries (11) 68°. From this it appears that Comanthus, Himerometra, (Pontiometra), Tropiometra, Cenometra, Oligometrides, Ilycrinus, and Metacrinus have oblique articulations, while the other types examined have nearly straight articulations. It is evident from these figures, according to Gislén, that no sharply defined boundary can be found between straight and oblique muscular articulations. Gislén said that if one accepts the later of the standpoints taken by the author one should be able to maintain that the straight muscular articulation on the distal face of the radial is always single, while all the other ossicles with muscular articula- tions have double articular faces, that is, form axillaries or bear an arm and a pinnule on their distal ends—though in this connection one must remember that there are types with defective pinnulation. In such a case Z; and Z, would always correspond to the first two brachials which might also, with a certain degree of justice, be called homologous in the recent crinoids. Gislén concluded it can be stated that the homologi- zation of certain proximal brachials based upon a supposed distinct difference between straight and oblique muscular articulations is erroneous. Gislén noted that the arm divisions are brought closer together by the first two brachials being longer on the outer side of the arm than on the inner. This is of some importance, for after the first arm division the arms in 5-rayed forms cannot, if they 22 BULLETIN 82, UNITED STATES NATIONAL MUSEUM are to expand in the same plane, diverge at an angle greater than about 72°. As a consequence of this the distal angle of the axillary ought not to be less than 108°. It turned out, however, that among the forms examined Ilycrinus carpenterii and Monachometra cf. fragilis only have an angle rising to any great extent above this figure. In these two cases it reaches 162° and 145°, respectively. All the other types except two have a [Br axillary angle of less than 108°—Comanthus japonica 88°; Capillaster sentosa, 105°; Amphimetra tessellata, 106°; Pontiometra andersoni, 105°; Tropiometra afra macrodiscus, 105°; Pectinometra flavopurpurea, 104°; Diodontometra bocki and Asterometra anthus, 110°; Notocrinus virilis, 83°; Clarkometra elegans, 103°; Hypalometra defecta, 96°; Antedon petasus, 78°; Heliometra glacialis, 68-72°; Atelecrinus helgae, 81°; and Metacrinus nobilis tenuis, 98°. In the 10-rayed Promachocrinus kerguelensis the [Br axillary ought not to fall below 144°; but in reality it is 107° here. Axillaries of higher rank, assuming the highest possible number of arms allowed by the full number of the division series, would have angles as follows: IIBr axillary, 144°-++ ; IIBr axillary, 162°+ ;IVBr axillary, 171°+-. The types examined, however, gave the following figures: Comanthus japonica, I1Br axillary, 104°, IIIBr axillary, 104°; Capillaster sentosa, I1Br axillary, 126°; I1[Br axillary, 120°; [VBr axillary, 131°, VBr axillary, 120°; Pontiometra andersoni, I1Br axillary, 130°, I1/Br axillary, 130°, IVBr axillary, 115°; VBr axillary, 105°; Pectinometra flavopurpurea, I1Br axillary, 126°; Metacrinus nobilis tenuis, I1Br axillary, 104°, I1IBr axillary, 102°, IVBr axillary, 108°. All the examples given above, other than the exceptions mentioned, have their arms diverging far too rapidly. But owing to the first postaxillary ossicles being longer on the outer side of the arms than on the inner this is counterbalanced. If the arms were narrow proximally and not in lateral contact, the basal portions of the free arms could certainly diverge at angles greater than those theoretically possible. In such a case, however, the divergence must soon be lessened to prevent the arms interfering in the more distal portions. If the distal angle of an axillary is smaller than the theoretically possible angle, the sides of the brachials immediately following, in order to restore the arms to the requisite angle, must be longer on the outside of the arms than on the inner side, and the angle that the proximal and distal ends form with each other must reach at least the difference between the actual distal angle of the axillary and the minimum possible. It seems, however, to be a general rule that the divergence of the arms is still further diminished, so that they become nearly parallel with each other. The earliest brachials are usually so much longer on the outside than on the inner side that the distal end of the first brachial forms almost a right angle with the longitudinal axis of the pre- ceding axillary, and the distal end of the second brachial slopes somewhat downward and inward. The too strong convergence of the arms resulting from this is compen- sated by the first syzygial pair (composed of brachials 3+-4) being longer on the inner side of the arm than on the outer, so that the distal margin of the epizygal (fourth brachial) is almost at right angles with the longitudinal axis of the arm. The effect of the arm branching on the obliqueness of the brachials is thus counteracted. The conditions found in Clarkometra and in Hypalometra show that the presence or absence of P, and P, has no influence worth mentioning upon the obliqueness of the first four brachials. A MONOGRAPH OF THE EXISTING CRINOIDS 23 In the comatulid brachials special cases sometimes occur in which the greater length of the pinnular side is pronounced from the beginning; more often, however, a series of discoidal segments follows the fourth brachial, or the reversion phenomenon pein INFLUENCE OF THE PINNULATION ON THE OBLIQUENESS OF THE BRACHIALS In the distal portions of the arms the position of the pinnules evidently has a controlling influence upon both the outer and the inner obliqueness of the brachials. The pinnular side of the brachial becomes more strongly developed; this causes its outer obliqueness. The pinnule’s point of insertion enlarges the pinnular side of the articular face; this causes its inner obliqueness. As examples of the decisive réle that the position of the pinnules plays in regard to obliqueness Gislén mentioned a specimen of Pectinometra flavopurpurea in which pinnules appear on the right side of both the ninth and tenth brachials, the proximal and distal borders of the tenth brachial being, as a result, parallel; and a specimen of Heliometra glacialis in which on one arm the ninety-second brachial bears two pinnules of normal size (19-20 mm. long) and has the distal articular face perfectly symmetrical and lying in a plane perpendicular to the longitudinal axis of the arm. Gislén remarked that in certain cases it seems as if the idea that phylogenetic causes may have played a certain part in bringing about obliqueness cannot altogether be dismissed. Thus in Atelecrinus, which lacks pinnules as far out as the twelfth or fifteenth brachials, if the arm division and pinnulation were the only causal factors, it might be expected that straight articulations would be found between at least the fourth and twelfth brachials. This seems to be the case regarding interior, though not exterior, obliqueness. The side of the brachial that would have borne the pinnule, if a pinnule were present, is in fact always longer than the opposite side. It is pre- sumable that phylogeny must be invoked here in order to explain the exterior oblique- ness of the brachials. The lateral flexibility, except at the synarthrial articulation, here seems to be extremely slight, so that there is no trace of reversion in this case. On the proximal brachials the pinnule socket always more or less invades the articular face, either intruding between the muscular and interarticular ligament fossae, or lying ventral to the muscular fossae. On more distal brachials the pinnule socket may wander out on to the lateral portion of the brachial so that is it completely isolated from the articular face of the brachial, or it may lie on the ventral side of the brachial and therefore, as in the proximal portion of the arm, ventral to the muscular fossa. In the former case the brachials are more or less elongated with markedly oblique articulations, while in the latter the brachials are usually short and discoidal. Here, however, the size of the pinnule base plays a certain réle. On account of the insertion of the pinnule bases the Calometridae ought to be placed in the latter group. However, the extraordinarily stout proximal pinnule segments do not permit such a pronounced shortening of the brachials as is seen in the Mariametridae and Himerometridae. In the Calometridae the pinnule attachment is extended at the expense of the muscular fossa of the pinnular side, which is reduced and acquires a more horizontal position on the distal side of the articular face. In other cases where the pinnule has a ventral origin the shortening of the brachials (ontogenetically as well as phylogenetically) may lead to discoidal brachials on which the proximal and distal articular faces are parallel. If, on the other hand,’ the origin 24 BULLETIN 82, UNITED STATES NATIONAL MUSEUM of the pinnule is more lateral, then with the shortening of the brachial, the pinnule being of normal size, a tendency toward greater shortening of the antipinnular side may make itself felt, and a certain inclination to dichostichality or biseriality appears in the arms. Among recent forms this is strongly pronounced in certain Comasteridae, in which there may be observed a thickening of the arm bases and at the same time a relative shortening of the brachials. This tendency toward biseriality is very evident, for example, in Comatula pectinata, but is best developed in Comatulella brachiolata. Gislén said it might be imagined that the thickening of the arms has been stimu- lated by two causes. First, in these types the gonads are developed in the arms; but this circumstance can scarcely play any very large part since in various other recent erinoids (Metacrinus, Notocrinus, and Isometra) gonads in the same position do not induce any thickening of the arms. Second, one may refer to the role that the creep- ing motions typical of the Comasteridae undoubtedly play. THE ACTIVE FLEXING POWER OF THE LIGAMENTS In the interbrachial articulations the transverse ridge serves as a hinge or fulcrum. On each side of this ridge are ligaments that are about equally powerful—the inter- articular ligaments and the dorsal ligament. Gislén remarked that if the dorsal liga- ment were antagonistic to the muscles alone it would be difficult to understand of what service the interarticular ligaments could be. Possibly it might be imagined that with the relaxation of the muscles they serve to counteract a too hasty flexion of the arms dorsally, which might involve the risk of breakage. It might also be sup- posed that with the ventral flexion of the arms the interarticular ligaments become com- pressed so that when the muscles are relaxed they would supplement the action of the dorsal ligament. Both these suppositions are confuted, however, by the fact that when the muscles are cut through no very strong dorsal bending of the arms occurs. A little dorsal flexion often appears which, however, is soon effaced. Another circum- stance speaking in favor of the ligament’s capacity for motion is the fact that the two ligament bundles of the synarthries may sometimes be rather unequally contracted, thus giving the arm or arm pair an oblique position. From this Gislén believes that the ligaments play a certain active part in the flexions of the arm. Gislén said that the arms of stalked crinoids seem to move quite slowly in com- parison with those of recent comatulids according to Bock’s observations on living Metacrinus and his own observations on Rhizocrinus. It should be noticed, too, that the ligaments between the proximal cirrus segments in the pentacrinites are very strongly developed so that even a slight contraction gives a powerful effect in the distal part of the very long cirrus. DEVELOPMENT OF THE LIGAMENTARY ELEMENTS AMONG CREEPING TYPES The Oligophreata in general and the comasterids in particular have only incon- siderable muscular attachments in comparison with their ligamentary interarticular connections. In certain comasterids the proximal portion cf the arms is greatly thickened, and this thickening, so far as regards the calcareous ossicle itself, is caused by the specially strong development of the ligamentary connections. Gislén pointed out that for a swimming type, like Antedon, a very important con- dition is that the structure of the arms should be light and slender. For creeping A MONOGRAPH OF THE EXISTING CRINOIDS Dn types, on the other hand, a stout and solid structure is of less consequence, if only the levers are powerfully built. It is therefore difficult to avoid the supposition that the thickness of the arm bases bears a certain relation to the creeping habits of the comas- terids. As the ligament fibers are much more intimately connected with the calcareous mass of the ossicle than are the muscle bundles, as the muscles in these types are rather poorly developed, and as it must be presumed that the ligamentary elements possess a capacity for active motion, Gislén sees no objection to accepting the hypothesis that an arm of the type mentioned must answer better the demand for an efficient pull and push lever than is the case with the more graceful and delicate arms found in the other comatulids that are better adapted for swimming. Gislén noted that a large number of the oldest (Jurassic) comatulids also had very stout arm bases that were often thickened proximally, where the same tendency to biseriality seen in Comatulella and Comatula appeared, although often it went much farther. Here, too, the muscular attachments, as in all the older Articulata, are weakly developed in comparison with the ligamentous connections. Gislén said that one can not help thinking that these phylogenetically old comatulids, which had fairly recently given up their sedentary habits, had not yet gained any great capacity for swimming, but were chiefly creeping types. On the other hand the antedonids, which are good swimmers, would then be a later type, more suited for a free and active mode of existence. SOLUTION OF THE REVERSION PROBLEM The causes of the shortening of the pinnule-bearing side of the proximal brachials in certain comatulids according to Gislén are the following: The degree of the effect caused by an articulation’s capacity for flexing the arm as a whole both ventrally and laterally increases as the position of the articulation becomes more proximal. The greatest capacity for flexion must therefore be con- sidered to exist in the most proximal brachial articulations. In the proximal articula- tions the interarticular ligament on the distal articular face of the brachials is developed much more strongly on the pinnular side than on the antipinnular side. As a result of the encroachment of the pinnule socket, which is fairly large in the proximal portion of the arms, the interarticular ligament of the pinnular side, and also the transverse ridge, is displaced dorsally on the pinnular side of the brachial’s distal articular face. On the proximal articular face of the succeeding brachials it is the end of the trans- verse ridge on the antipinnular side of the brachial that is displaced dorsally. In other words, we find in the dorsal part of the articulation a certain likeness to a synarthry, although a very oblique one. The dorsal end of the transverse ridge, which runs between the great interarticular ligament and the dorsal ligament, is now produced so that a process from the proximal margin of the more distal brachial fits into a notch in the distal margin on the more proximal brachial. We thus get something very similar to an oblique synarthrial projection where, as in the latter, the synarthrial process is directed backward. At the same time as this prolongation of the proximal margin of the distal brachial takes place posteriorly, the point of attachment for the great interarticular ligament spreads out on the lateral inner side of the increased attachment surface. As the projection from the more distal ossicle fits into the more proximal ossicle on its pinnular side, it encroaches upon the breadth of the pinnular side of this brachial, which thus becomes narrower than the antipinnular side. On the 208244403 26 BULLETIN 82, UNITED STATES NATIONAL MUSEUM other hand, the more distal brachial, which bears the process, is strengthened on the antipinnular proximal side by the posteriorly directed process, and thus becomes broader there. It is evident that such a structure gives the arm increased capacity for flexion laterally. An articulation in which both the end points of the transverse ridge are strictly lateral evidently would not permit of any flexion at all laterally, whether the articulation in other respects is oblique or not. The greater this reversion of the conditions of breadth in the proximal brachials, the greater the capacity for lateral flexion in the arms. A certain amount of lateral flexion is possible in all arms in which synarthries are developed. In fossil types where synarthries are found and in recent stalked crinoids this restricted possibility for lateral motion seems to be adequate. The same seems to be the case in the primitively organized Atelecrinus in which no decrease at all of the breadth of the pinnular side of the brachials appears in any part of the arms. Here one is able, from the primitive- ness of the type in other respects, to assume an inconsiderable lateral mobility. An increased lateral mobility here would, according to Gislén, most likely give rise to discoidal brachials. Under certain conditions a case apparently similar to that of Atelecrinus may occur because of enormously enlarged pinnule bases, as for instance in the Calometridae. In other cases the habit of creeping, as in a number of genera of Comasteridae, possibly combined with original primitiveness, may bring about a similar phenomenon. In all the other comatulids reversion may be observed. In the cases in which it is little developed the influence of the pinnule is about equal in importance to the counter- acting, reversional, lateral flexibility; here we get in the proximal portion of the arms a number of discoidal brachials. In the cases in which it is more strongly developed, as for instance in the Antedonidae, the capacity for lateral flexion is the strongest, and a number of brachials appear which are considerably narrower on the antipinnular than on the pinnular side. THE MUSCLES Gislén’s conclusions in regard to the origin and development of the musculature in the crinoids are as follows: A pair of longitudinal muscle bands, corresponding to those in the holothurians, presumably existed on the ventral side of the arms of the oldest crinoids. With the demand for increased and more rapid motion of the arms these muscle bands began to play a more active part in flexing the arms, and therefore acquired stronger attachments to the brachials. The brachials developed crests and caleareous processes, and these extended outward into the longitudinal muscle band and divided it into segments, which now correspond to the interbrachial muscles. For the only slightly movable arms of most of the Paleozoic crinoids the liga- mentary articulations in the proximal part of the arms, which still lacked true muscles, were sufficient for the flexion required. In these crinoids we therefore find only syzygial, or more commonly more or less synostostical, articulations in the middle and distal portions of the arms. Later, with the necessity for increased speed and effectiveness in movement, the true muscles, which heretofore had only been found in the ventrally situated soft parts, became associated with the calcareous skeleton. This association took place earliest in the articulation between the radials and the arms, standing alone in its type, in the Flexibilia, according to Springer becoming A MONOGRAPH OF THE EXISTING CRINOIDS 27 extinct in the upper Carboniferous without leaving any descendants. The Articulata, which according to Gislén may be traced back to Poteriocrinites, which in its turn is descended from some group of Dendrocrinites, have independently followed a similar course and developed the type usual in recent crinoids. Among the Jurassic comatu- lids the muscle attachments are still relatively insignificant. In the radial muscular articulations they may sometimes even be lacking, as in Pontiometra and Stephano- metra; but this seems to be a condition of secondary origin. Gislén said that the ligamentary articulations would thus appear to be the oldest arm flexors, which, through their action, have developed the transverse ridge. There- fore the interarticular ligaments are the original antagonists of the dorsal ligaments, The ligamentary articulations can be derived from the close sutural connections be- tween the ossicles of the Cambrian, Ordovician, and Silurian crinoids. The role of true muscles as arm flexors would be of later date, arising from the necessity for in- creased and more rapid motion. The muscular articulations first reach their full development in certain recent eleutherozoic comatulids, while other comatulids less suited for swimming, many recent stalked crinoids, and the Permian to Jurassic forms, have muscular attachments that are very modestly developed. LIGAMENTOUS ARTICULATIONS IN THE ARMS Gislén divided the articulations lacking striated muscle fibers into immovable and movable ligamentary articulations. The immovable ligamentary articulations he subdivided into (1) close synostosis (with a flat smooth joint face, as the junctures between the radials); (2) syzygy; (3) pseudosyzygy; and (4) anchylosis. The movable ligamentary articulations he divided into (1) those movable in all directions, and (2) those movable in one plane only. Those movable in all directions he subdivided into (a) articulations between the columnals of pentacrinites and comparable types, with petaloid sculpture; (b) loose synostosis—the apposed joint faces somewhat concave, without scultpure; and (c) pseudosyzygies as developed between the calyx plates in the Flexibilia. Those movable in one plane only he divided into (a) synarthry (the eryptosynarthry being a transition between the synarthry and synostosis); and (b) ligamentary articulation, with at least three ligamentary pits (occurring in place of syzygies in the Bathycrinidae, and according to Gislén between the radials and the IBr; in Pontiometra and Stephanometra and perhaps also in burdigalocrinid stems; this is the trifascial articulation of P. H. Carpenter). The articulations including striated muscle fibers Gislén divided into two cate- gories: (a) those movable dorsoventrally, with one dorsal and two interarticular liga- ments and two fairly strong muscle bundles—muscular articulations, sensu stricto; these occur between the brachials, between the brachials and first pinnule segments, and between the first two pinnule segments; and (b) those movable in a lateral direction only, with no ligament fossae and usually small muscle fossae—pinnular articulations; these occur between the pinnule segments from the second onward in the proximal part of the arms. Gislén said that of the ligamentary articulations the oldest both ontogenetically and phylogenetically is the synostosis. The synarthry can thus scarcely be derived from the syzygy, or the syzygy from the synarthry; both are differentiated varieties of 28 BULLETIN 82, UNITED STATES NATIONAL MUSEUM synostosis, specialized for particular purposes. One or the other form has developed in response to the different demand made by the articulation in question. Some- times, however, the syzygy seems to develop into the synarthry. In Crossometra septentrionalis, in which usually the first two brachials are united by syzygy, Gislén observed in one case two lateral concavities near the nerve lumen. The center of the joint face therefore represents a synarthry, while the peripheral portion is of the normal syzygyial type. Catoptometra (magnifica minor) shows possibly a somewhat different direction in its development. In fully grown individuals the articulation between the IBr, and IBr; is syzygial; in the young this articulation is furnished about its periph- ery with weak syzygial septa with a slightly stronger dorsoventral thickening which gives the joint face a somewhat synarthrial aspect. Typical syzygies scarcely occur in recent stalked crinoids. Calamocrinus and Ptilocrinus have typical synostoses, as have a large number of pentacrinites also. In the pentacrinites, however, there is an evident tendency toward the development of syzygies. Here, as in the fossil forms, there is found the first tendency toward the formation of syzygies, foreshadowed by a weak fluting of the dorsolateral margin of the joint face. The marginal ridges gradually extend toward the center, and when they reach it the typical syzygy appears. The species of Metacrinus have almost typical synostoses, which, however, sometimes show a very indistinct marginal stria- tion. Syzygies are indicated about as feebly in Annacrinus. On the other hand, a syzygial type is rather distinct in Hypalocrinus naresianus and in Cenocrinus asteria. In describing the syzygies of the comatulids, Gislén called the thickened margin of the nerve lumen the areola. The radiating ridges he called the septa. They are complete when they reach the areola, or incomplete when they run only part way in from the periphery of the joint face. They are called finished when they arise within the margin of the joint face and continue to the center. They are forked when two fuse a short distance within the margin. When the septa are both incomplete and finished, there result small, low, more or less rounded protuberances—the tubercles. In many of the larger forms the ventral septa especially show a tendency toward divi- sion into rows of tubercles. Sometimes concentric ridges—synapticulae—run at cer- tain distances from the nerve lumen. The areola is often connected with the ventral, usually grooveless, part of the joint face by a more or less broad smooth line—a ven- tral beam. Usually the median portion of this ventral beam is occupied by a pit or furrow, so that there are two parallel ventral beams. These ventral beams may be developed as coarse bow-shaped septa that run out to the margin of the joint face. These ridges and other elevations as found in different forms are marked by dif- ferent degrees of sculptural relief. Gislén denoted the strength of relief by figures from 1to4. The figure 1 was assigned to forms having the septa only feebly indicated, as for instance certain Charitometridae; syzygies in which the septa stand out in high relief, such as those in Heliometra glacialis, are given the figure 4. The elevations on the distal face of the hypozygal are matched by corresponding elevations on the proximal face of the epizygal. When the syzygial articulation is observed from the dorsal side of the arms it is seen as a dotted line, the dots marking the ends of the canals between the septa which are traversed by the ligaments. Gislén noted that the development from synostosis to syzygy evident in the phylo- genetic history of the crinoids appears again in the ontogeny of the comatulids. In A MONOGRAPH OF THE EXISTING CRINOIDS 29 the comatulids very young individuals have the ligamentary connections in the form of synostoses. The distal face of the hypozygal is somewhat concave, and the proxi- mal end of the epizygal somewhat convex. The change from synostosis to another form of ligamentary union seems to take the following course. The synostosis becomes less close—in other words, the ligamentous fibers between the ossicles in certain places become longer. Sometimes the contact is retained about the lumen, resulting in an areola, and the contact seems to be closer at the margin of the ossicles. If the articu- lation is to develop into a synarthry, a strong areola always appears from which a ventral and dorsal beam grow out, resulting in the dorsoventral articular ridge. But if the articulation is to develop into a syzygy an areola may appear and persist, or it may be altogether lacking. The septa, very feeble at first, appear in the dorsal and lateral parts of the joint faces and are extended outward, because with the apposi- tional increase of growth in the articulation they continue in contact with the margin. During the development the septa increase in number, and the new ones that are formed later arise in the same way from the border of the joint face, intercalated between the septa first formed, which extend into the neighborhood of the central lumen. That the septa must also to some extent increase in growth toward the center is seen by the increasing number of complete septa in older invididuals. It is cer- tainly true that the lumen is somewhat enlarged during growth, but this is not suffi- cient to explain the increase of complete septa. In a young Antedon petasus the diameter of the ossicle is 0.31 mm., and of the central lumen 0.06 mm. There are two complete septa. In a full grown individual the lumen is 0.19 mm. in diameter, and there are 12 complete septa. The variation in the number of septa in any given species at any given age is rather small and unimportant. In a Mariametra subcarinata, for example, Gislén found the number of septa on the distal end of four third brachials to be 15-16. In a specimen of Heliometra glacialis the number of septa on the distal end of the third brachial was 36-36, and in another of Promachocrinus kerguelensis the number was 27-30. The type of syzygy was here, as everywhere else, uniform within the species. In strongly multibrachiate forms the size of the segment and the number of septa may vary a little more, according to whether one examines an inner or an outer arm, and the number of axillaries preceding the arm, as for instance in Capillaster, Coman- theria, and Stylometra. Beyond the first syzygy the number of septa in a few syzygies may increase slightly, but fairly soon the septa decrease in number, as a natural result of the distal portion of the arms being younger and slenderer than the proximal. The ninety-first brachial (about the thirtieth hypozygal) in a fully grown Heliometra glacialis with arms about 200 mm. long had only 22 septa, consisting of 7 ven- tral rows of tubercles and 15 more or less incomplete septa. The hypozygal follow- ing had the same number of septa. In the small Jsometra vivipara with arms about 60 mm. long the distal face of the third brachial had 14 septa. In this specimen the fourth or fifth syzygyial articulation was developed as a synostosis, although not a close one, and lacked septa. Large and stout types have usually an unusually large number of septa for the systematic units to which they belong, while small and more delicate forms have an unusually low number. Also young and immature individuals have fewer septa than fully grown individuals of the same species. A young Heliometra glacialis from 30 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Taimyr with arms about 18 mm. long had in two cases 18 septa on the end of the third brachial; on the ninth brachial there were 18 and on the fourteenth 15 septa. Another young individual of the same species from East Greenland with arms 65 mm. and cirri 18 mm. long had on the distal end of the third brachial 20 septa. A young individual of Catoptometra magnifica minor with arms 35 mm. long had 11 septa on the distal end of the IBr,, two of them complete and forming an indistinct dorso- ventral ridge. On the distal face of the third brachial there were 16 septa, 6 com- plete septa and 2 ventral beams. The degree of strength was 1. No areola was present. On the distal face of the ninth brachial there were 14 septa, and on the distal face of the fourteenth brachial 12. A full-grown Catoptometra rubroflava with arms about 100 mm. long had 28 septa on the IBr,, 25 on the third brachial, and 24 on the ninth brachial. A young Antedon petasus with the arms about 16 mm. long had two (complete) very weak septa, visible only with the strongest magnification. CoMASTERIDAE Comatella brachycirra.—In a specimen with the arms about 50 mm. long the distal face of the third brachial has 15 septa, 5 complete broad septa and 2 ventral beams; the degree of strength of the septa is 3. In another case there were 17 septa on the distal face of the third brachial, and 13 on the distal face of the sixteenth brachial. Comissia peregrina magnifica.—In a specimen with arms about 150 mm. long the distal face of the third brachial has 30 septa, 11 complete septa, one of them forked, 16 incomplete septa, and 2 ventral beams. The degree of strength is 4. The areola is of medium breadth, stout and well marked. The distal end of the thirteenth brachial has 27 septa, 13 incomplete septa and 3 ventral rows of tubercles. The septa run out into the dorsal margin of the ossicle. Capillaster sentosa.—In a specimen with the arms 90 mm. long the IIBr. has 45 septa, 6 complete septa and 2 ventral beams. The IIIBr, has 44 septa, 8 com- plete septa, 2 ventral beams, some septa forked, a distinct areola, strength 4. In three other cases the septa are 44, 43, and 42. The IVBr, has 37, 38, 39, and 43 septa. The VBr, has 33 and 38 septa. The second brachial, following a VBr axillary, has 28 and 25 septa, 2 complete. In a fifth case the number of septa on a IVBr, is 44, 2 of them ventral beams. On the following second brachial the septa were 40, two of them complete; there was a narrow but distinct areola; the degree of strength was 3. In another case the second and third brachials would not separate in spite of violent boiling in KOH. On the twenty-ninth brachial (second hypozygal) there were 34 septa, 3 complete septa, 6-7 very small tubercle-shaped marginal septa, and 2 ventral beams; the degree of strength was 4. In another case the second hypozygal (about the thirty-first brachial) had 29 septa. Comatula pectinata.—In a specimen with the arms about 70 mm. long the proximal face of the axillary had 58 septa with the degree of strength 3-4; of the septa 45 are dorsally or laterally directed; the 13 septa directed ventrally are incomplete; of the 45 dorsal septa 19 are complete and some are forked; the areola is indistinct; the surface of the ossicle is not, as usual, even or evenly curved, but is divided into three or four faces that stand somewhat obliquely to one another and on different levels in the horizontal plane when the ossicle is viewed from the proximal end; Gislén says A MONOGRAPH OF THE EXISTING CRINOIDS 31 it recalls the metacarpal III of a horse; the 58 septa run over these faces. On the distal end of the first brachial there are 42 septa, 34 dorsally and 8 ventrally directed; 15 are complete. In another case there are 43 septa, of which 4 are rows of tubercles. On the distal end of the third brachial in two cases there are 39 and 40 septa. On the eighth brachial there are 42 septa, on the fifteenth 43, on the twentieth 44, of which 10 are complete. The areola is narrower than in the preceding species. The septa are distinct, but broad and rather low. Comantheria delicata grandis—In a specimen with the arms about 110 mm. long the third brachial has 34 septa, with 6 ventral rows of tubercles, the degree of strength being 3. In another case there are 37 septa. The areola is broad, but low and rather indistinct. Comaster serrata.—In a specimen with the arms about 60 mm. long the articula- tion in the IBr series is a pseudosyzygy; the areola is rather broad with 5 finished tubercular septa and in addition two concentric rings of tubercles; the degree of strength is 3. The distal face of the third brachial has 19 septa, broad and low, of which 10 are complete. The areola is narrow. Comantheria grandicalyr.—In a specimen with the arms about 120 mm. long the IVBr, has about 31 septa with the strength 4. All the septa are incomplete, not reaching the stout areola. Between the areola and the inner ends of the septa there is a ring of 8 or 9 tubercles. There is a stout ventral beam and also a similar short finished one. A third brachial on the outer side of a division series has 31 incomplete septa, and an inner ring of 11 tubercles, with a distinct synapticle ring dorsally. The ventral beam is rather indistinct; the areola is stout. In another third brachial on the outer side of a division series there are 29 septa, three of them complete. The ring of tubercles is developed on one side only, and includes 6 tubercles. The ventral beam is stout and lobed. Hypozygals from the distal parts of the arm have 15-17 septa, all incomplete, with the degree of strength 3. The areola is narrow and insignificant. Comanthus japonica.—In a specimen with the arms about 110 mm. long the IIBr; has about 44 septa; of these 21 are complete, most of them finished, 18 are incomplete, and there are 5 ventral rows of tubercles. The degree of strength is 4. The areola is stout, well marked, and rather broad. The twenty-sixth brachial (fourth brachial hypozygal) has 35 septa of which 11 are complete. The areola is rather narrow. ZYGOMETRIDAE Catoptometra rubroflava—tIn a specimen with arms about 100 mm. long the distal end of the IBr, has 28 septa, of which 8 are complete; these are broad and rather low, with the strength 3 (—4); the areola is of medium width. The distal end of the third brachial has in one case 25 septa, and in another 23. The ninth brachial has 24 septa. Zygometra microdiscus elegans.—In a specimen with the arms 105 mm. long the distal end of the IVBr, has 30 septa, about 13 complete, with the degree of strength 3; the areola is absent. 32 BULLETIN 82, UNITED STATES NATIONAL MUSEUM EUDIOCRINIDAE Eudiocrinus indivisus—In a specimen with arms about 70 mm. long the IBr, has on the proximal end 16-17 septa. The third brachial has on the distal end 21 septa that are very low, especially toward the center, with the degree of strength 2; the areola is indistinct. HIMEROMETRIDAE Himerometra magnipinna.—In a specimen with arms 105 mm. long the IIIBr, has 35 septa, becoming rather indistinctly marked toward the areola, which is also indistinct and situated in a central concavity; the degree of strength is 2. Heterometra crenulata.—In a specimen with arms about 95 mm. long the distal end of the third brachial has 41 septa; there are two stout ventral beams and about 7 complete septa; the intervals between the septa are rather narrow; the areola is narrow and rather indistinct; the degree of strength is (2—) 3. The sixteenth brachial (second hypozygal) has 48 septa, of which about 13 are complete; there are two ventral beams. The thirty-first brachial (third hypozygal) has 50 septa. MARIAMETRIDAEB Stephanometra spicata.—In a specimen with the arms 135 mm. long the distal end of the third brachial has 31 septa, one complete and two ventral beams, with the degree of strength 3. On another third brachial there are 29 septa, of which 2 are complete; the areola is narrow, but rather distinct. Liparometra grandis.—In a specimen with arms about 120 mm. long the third brachial has 37 septa, 10 complete and 2 represented by indistinct ventral beams, the degree of strength being 3; the areola is distinct. The ninth brachial has 31 septa. Lamprometra palmata.—In a specimen with an arm length of 90 mm. the third brachial (following a IVBr axillary) has 24 septa, 6 complete and 2 ventral beams; the areola is distinct and of medium breadth; the degree of strength is 4. In another case there are 23 septa. The twenty-second brachial (second hypozygal) has 22 septa. The thirty-third brachial (third hypozygal) has 24 septa, 5 complete and 2 incomplete ventral beams. The distal syzygies, from about the one hundredth to about the one hundred and tenth brachials, or about the tenth hypozygal, have 16 incomplete septa. The hypozygal following has 14 incomplete septa, with the degree of strength (1—) 2. Mariametra subcarinata.—In a specimen with arms 60 mm. long the third brachial has 15 septa, including 2 complete septa and one ventral beam, with the degree of strength 4-3. In other cases the septa are 15, 16, 16, with 3, 5, and 6 complete septa. The areola is rather narrow, but usually stout. CoLOBOMETRIDAB Pontiometra andersoni.—In a specimen with the arms about 160 mm. long the third brachial, following a VBr axillary, has 36 septa, of which 5 are complete and 2 are ventral beams, with the degree of strength 4. In other cases the third brachial, following IVBr and VBr axillaries, has 37, 39, and 39 septa. The areola is stout, well marked, and of medium width. A MONOGRAPH OF THE EXISTING CRINOIDS 33 Cenometra bella.—In a specimen with arms 90 mm. long the third brachial has 21 septa, of which 10 are complete and 2 are indistinct ventral beams, with the degree of strength 8. The thirtieth brachial (second hypozygal) has 19 septa, of which about 6 are complete. Gislén says that this reminds one of the Liparometra type. Cyllometra manca (‘‘disciformis’’).—In a specimen with an arm length of 90 mm. the third brachial has 16 septa, of which 9 are complete and 2 are ventral beams, with the degree of strength 3; the areola is narrow and rather indistinct. The thirteenth brachial (second hypozygal) has 17 septa, the eighteenth has 14, the twenty-third has 15, the twenty-ninth has 13, and the thirty-fifth (sixth hypozygal) has 13. The areola is lacking. The degree of strength is 1-2. Oligometrides adeonae.—In a specimen with arms 60 mm. long the third brachial has 30 septa, of which 13 are complete and one is a ventral beam; the areola is indis- tinct; the degree of strength is 3. The nineteenth brachial (second hypozygal) has 24 septa, of which about 11 are complete; the ventral beam is indistinctly bifurcated. There are traces of synapticulae. Austrometra thetidis—In a specimen with the arms about 40 mm. long the third brachial has 15 septa of which 5-6 are complete; the areola is rather narrow, but distinct, with a broad smooth area united with the ventral part of the joint face; the degree of strength is 2. Clarkometra elegans.—In a specimen with arms 25 mm. long the third brachial had 11 septa, scarcely any complete, and a narrow indistinct areola; the degree of strength is 2-3. TROPIOMETRIDAE Tropiometra afra macrodiscus.—In a specimen with the arm length 190 mm. the distal end of the third brachial has about 52 septa, of which about 25 are complete, some of these being finished. The ventral portion of the joint face has rather irregular pits and a ring of tubercles which is synapticularly arranged. The areola is large and stout, protruding as a small cone that rises from a surrounding concavity. The degree of strength is 4. In another case there were 51 septa, about 27 complete. CALOMETRIDAE Pectinometra flavopurpurea.—In a specimen with arms 70 mm. long the third brachial has 14 septa, 2 complete; the ventral beam is very broad, and the areola is broad and stout; the degree of strength is 2. The thirteenth brachial has 15 septa. The twentieth brachial (third hypozygal) has about 15 septa, which are similar to those of Mariametra. Neometra multicolor—In a specimen with the arm length of about 60 mm. the third brachial has 19 septa of which 5 are complete and 2 are ventral beams; the areola is broad and stout, though rather low; the degree of strength is 3. The thir- teenth brachial (second hypozygal) has 15 septa, 5-6 of them complete. PTILOMETRIDAB Ptilometra miilleri—A specimen with an arm length of 50 mm. has the fifth brachial (first hypozygal) with 21 septa, two of them indistinct ventral beams; the areola is indistinct; the degree of strength is 3. A concentric furrow is present, as 34 BULLETIN 82, UNITED STATES NATIONAL MUSEUM in Asterometra, but closer to the periphery of the ossicle. The fifteenth brachial has 19 septa, about 9 complete. The twenty-third brachial (third hypozygal) has 18 septa, of which 7 are complete. ASTEROMETRIDAE Asterometra anthus.—In a specimen with arms 80 mm. long the third brachial has 29 septa, of which 14 are complete; in some cases the septa are forked; the areola is narrow; the degree of strength is 3. There is a concentric furrow about midway between the periphery of the joint face and the central lumen. In another case there are 27 septa. The ninth brachial (second hypozygal) has 28 septa. THALASSOMETRIDAE Stylometra spinifera.—In a specimen with arms about 90 mm. long the third brachial, following a IIBr axillary, has 23 septa, of which 9 are complete and 2 are indistinctly separated ventral bea ns; a couple of septa are forked, and some of the ventral ones are partly subdivided into rows of tubercles; the areola is broad and stout; the degree of strength is 3-4. Two third brachials following a IJBr axillary have 19 and 20 septa, with the ventral beam not subdivided; the areola is as before, but the septa are somewhat lower. Oceanometra annandalei.—In a specimen with arms 80 mm. long the ITBr; has about 18 septa. The third brachial has 19 septa, including 6 complete septa and one ventral beam; the areola is broad, though rather low; the degree of strength is 2-3. The septa are stout, running out as ribs to the margin of the ossicle, and there is a marginal growth ledge. Stenometra dentata.—In a specimen with an arm length of about 50 mm. the third brachial has 18 septa, 8 of which are complete and 2 are low ventral beams; the areola is rather narrow and indistinct; the degree of strength is 2-3. CHARITOMETRIDAB Perissometra cf. aranea.—In a specimen with arms about 80 mm. long the third brachial has 22 septa, 4 complete and 2 ventral beams; the areola is indistinct; the degree of strength is 2. The thirteenth brachial has 20 septa, 10 complete and 2 ventral beams. The eighteenth brachial has 19 septa and a narrow areola. The twenty-third brachial has 17 septa. The twenty-ninth brachial has 17 septa, of which 2 are complete. The thirty-fourth brachial (sixth hypozygal) has 14 septa, all incomplete, and no areola. Crossometra septentrionalis.—A specimen with an arm length of 70 mm. has the I1Br; with 17 very low septa, about 7 incomplete, with the degree of strength 1. Three first brachials have 14, 14, and 16 septa. A third brachial has 16 septa, 4 complete and 2 ventral beams; the areola is of medium width, but low. The eleventh brachial has 15 septa, 3 complete and 2 ventral beams, with the degree of strength 2. In one case the articulation between the first two brachials is a compromise between a synarthry and a syzygy. Diodontometra bocki.—In a specimen with the arms 60 mm. long the third brachial has 16 septa, 6-7 complete and the ventral beam weakly developed and partly divided; the degree of strength is 1; the areola is distinct. A MONOGRAPH OF THE EXISTING CRINOIDS 35 NOTOCRINIDAE Notocrinus virilis—In a specimen with the arms 75+ (probably about 125) mm. long the third brachial has 18 septa, 3 or 4 finished and only 2 incomplete; the degree of strength is 2. The areola is narrow, but there appears outside of this a very broad almost smooth area occupying about half the space between the lumen and the margin of the ossicle. Beyond this area there is a stout concentric synapticula, and beyond this a similar weaker one. In another case there were 19 septa, about 4 incomplete. Two eighth brachials had 19 and 20 septa and 2-3 weak synapticulae; the degree of strength is 1. ANTEDONIDAE Antedon petasus.—In a specimen with arms about 80 mm. long the third brachial has 21 septa, 12 complete, and an indistinct areola; the degree of strength is 2-3. In another case there are 19 septa. The ninth brachial has 19 septa with an abnor- mally large and stout areola. The fifteenth brachial has 19 septa. The nineteenth brachial has 19 septa, of which 12 are complete, aud a narrow distinct areola. In the last two hypozygals there is a distinct ventral beam. Compsometra serrata.—In a specimen with an arm length of 30 mm. the third brachial has 14 septa, 9 complete, and a narrow areola; the degree of strength is 2. In another case there are 12 septa, about 10 complete. The ninth brachial has 13 septa, including 2 indistinct ventral beams. The fourteenth brachial (third hypozygal) has 14 septa. Isometra vivipara.—In a specimen with arms 60 mm. long the third brachial has 14 septa, 10 complete; most of the septa are finished about halfway to the border of the ossicle; the areola is narrow; the degree of strength is 2. From the fourth or fifth syzygy onward septa are lacking, the union becoming a loose synostosis, with about the lumen a rather marked areola situated in the center of a somewhat con- cave face. Heliometra glacialis—In a specimen with an arm length of about 200 mm. the distal face of the third brachial has about 36 septa; there are 22 complete septa, about 6 ventral rows of tubercles, and 2-3 concentric synapticulae, of which the innermost is the stoutest; the areola is dorsally rather moderate, ventrally very broad; the degree of strength is 4. In two other cases there are 39 septa, 22 complete, 6 rows of tubercles, and 2 ventral beams. The twenty-sixth brachial (sixth hypo- zygal) has 29 septa, including two ventral rows of tubercles; the septa are somewhat lower than those on the third brachial, and the areola is narrow. Promachocrinus kerguelensis.—In a specimen with arms about 150 mm. long the third brachial has 27 septa, 11 of these incomplete, and on each side of the median line 3-5 ventral tubercle rows; the degree of strength is 4. In another case there are 30 septa including 8 rows of tubercles; the areola is narrow or of medium width, indistinct, sometimes lacking. The ninth brachial has 32 septa. The fourteenth brachial (third hypozygal) has 34 septa, including 7 ventral rows of tubercles; more- over, many of the septa are partly subdivided into rows of tubercles. Leptometra celtica.—In a specimen with an arm length of about 70 mm. the third brachial has 24 septa, about 12 complete septa, 3 of these finished, and 2 ventral beams; the areola is narrow and indistinct; the degree of strength is 2. In another case 36 BULLETIN 82, UNITED STATES NATIONAL MUSEUM there are 19 septa. The ninth brachial (second hypozygal) has 19 septa, and the fifteenth (third hypozygal) has 17 septa. Psathyrometra wireni.—In a specimen with the arms about 60 mm. long the third brachial has 16 septa, of which 11-12 are complete; the areola is narrow and indis- tinct; the degree of strength is 1; there is one indistinct concentric synapticula. The ninth brachial (second hypozygal) has about 16 very low septa. Poliometra proliza.—tIn a specimen with arms about 100 mm. long the third brachial has 21-22 septa, of which 13 are complete; the areola is ventrally in broad contact with the smooth ventral part of the ossicle, dorsally narrow and indistinct; the joint face is somewhat undulating; the degree of strength is 2. In another case there are about 22 septa. The ninth brachial (second hypozygal) has 21 septa, and the ninteenth (third hypozygal) has 18 septa. Hathrometra tenella var. sarsii.—In a specimen with an arm length of about 50 mm. the third brachial has 22 septa, including 9 complete septa and 2 ventral beams; there is no areola; the degree of strength is 3. In another case there were also 22 septa. The ninth brachial (second hypozygal) in two cases had about 18 septa. The fourteenth brachial (third hypozygal) in two cases had 18 and 19 septa; the degree of strength was 2-3. PENTAMETROCRINIDAB Pentametrocrinus diomedeae.—In a specimen with the arms about 100 mm. long the distal face of the fourth brachial (first hypozygal) had 18 septa, including 11 complete septa and 2 ventral beams; the areola was scarcely indicated; the degree of strength was 2-1. The ninth brachial had 16 septa. Thaumatocrinus jungerseni.—In a specimen with the arms about 70 mm. long the distal face of the fourth brachial (first hypozygal) had 11 septa, including 5-6 complete septa and 1 ventral beam; there was no areola; the degree of strength was 2-3. The ninth brachial (second hypozygal) had 11 septa, including 2 ventral beams. The sixteenth brachial (third hypozygal) had 11 septa, including 5 complete septa and 2 ventral beams. ATELECRINIDAE Atelecrinus helgae.—In a specimen with an arm length of perhaps 90 mm. the third brachial had 10-11 septa, including one broad ventral beam with a small ventral pit, about 10 complete and finished or interrupted very narrow septa, four of them tubercles only, and 3 ventral tubercles continuous with the ventral smooth part of the ossicle; there was no areola; the degree of strength was 2-3. The seventh brachial (second hypozygal) had 12 septa, including 11 very narrow and rather complete septa and one ventral beam. The degree of strength is 2. Atopocrinus sibogae.—Gislén said that if the author’s statement that two small ventral muscle bundles occur in the syzygies of this species can be verified this species presents syzygial unions of a very interesting and ancient type, reminding one of the oldest articulate type. He has, however, never been able to find any trace of rudi- mentary muscles in ligamentary articulations in sections through arms of young comat- ulids, in regenerating arms, or in ligamentary unions in Ptilocrinus pinnatus, Rhizo- crinus lofotensis, or Ilycrinus carpenterii. Until it has been verified, he wrote, he prefers to regard the statement with some reservation. A MONOGRAPH OF THE EXISTING CRINOIDS 37 SUMMARY Gislén said the author’s statement that the number of syzygial septa is “to a certain extent proportionate to the size of the articular face, and is greater in the Oligophreata than in the Macrophreata” in general holds good. But there are some important exceptions, especially to the latter assertion. For instance, the oligophreate families Calometridae and Charitometridae have unusually few septa, and the sume is true in regard to most of the Thalassometridae and certain of the Colobometridac. Comasteridae —The septa in the proximal syzygies are usually numerous (30-58), below 20 only in small forms. The degree of strength is 3-4. The areola is strongly developed, but rather narrow. Plenty of tubercles occur, often also ventral tubercle rows. Pseudosyzygies sometimes appear among species of the genus Comaster between the first two post radial ossicles (the elements of the [Br series). These, like typical syzygies, often replace synarthries. The distal intersyzygial interval is normally 3-4 muscular articulations (an exception in Capillaster). Zygometridae——There are 16-30 septa in the proximal syzygies. The degree of strength is 2-8. The areola is indistinct. A syzygy replaces the most proximal synarthry. The young stages of these syzygies often remind one of the pseudosyzy- gies in Comaster. The number of incomplete septa is large. The distal intersyzygial interval is fairly short, 3 muscular articulations in Hudiocrinus (now assigned to a special family, Eudiocrinidae), and 4-8 in Catoptometra. In Zygometra, however, the interval is long, 10-30 muscular articulations. Himerometridae.—In the proximal syzygies the septa number 35-50, with the degree of strength 2-3. The areola is indistinct. Sometimes there are 2 clearly marked ventral beams. The number of incomplete septa is rather large. The distal intersyzygial interval is long, (6) 10-26 muscular articulations. Mariametridae.—The proximal syzygies have 25-40 septa, except in the delicately built genus Mariametra in which there are only 15. Gislén remarked that the heavily granulated disk also shows Mariametra to be a primitive type. The degree of strength is 2-4. The areola is well developed—in Stephanometra narrow but distinct. There are 2 ventral beams. Complete septa are few. There are occasionally incidental irregularities in the distribution of the proximal ligamentary unions, so that syzygies may replace synarthries. The distal intersyzygial interval is fairly long, 5-18 muscu- lar articulations. Tropiometridae.—The septa in the first syzygy number about 50, with the degree ofstrength4. The areolaisstout, raised onaconical prominence. There are a good many completesepta. The distal intersyzygial interval is 5-10 muscular articulations. Colobometridae.—The proximal syzygies have 15-30 septa, with the degree of strength 2-3. The areola is narrow, distinct or indistinct. There is rather a large number of complete septa. There is great diversity within this family, Cenometra reminding one of Liparometra, Oligometrides of Tropiometra, Cyllometra of the ante- donid type, and Austrometra most nearly of certain Thalassometridae. The distal intersyzygial interval is 3-10 muscular articulations. Calometridae——The proximal syzygies have 13-19 septa, with the degree of strength 2-3. The areola is broad and stout. There are few complete septa. The distal intersyzygial interval is 3-15 muscular articulations. 38 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Thalassometridae—The proximal syzygies have 19-29 septa, with the degree of strength 2-3. In the Ptilometridae (here included as a subfamily of the Thalassome- tridae) the areola is indistinct, whereas in the Thalassometridae it is generally broad and stout. Very often a concentric furrow appears parallel to the margin of the ossicle. The septa are sometimes forked. There are 1 or 2 indistinct ventral beams. The proximal syzygies in Ptilometra (Ptilometridae) and even in Asterometra (Astero- metridae) are often irregular, sometimes also replacing synarthries. The distal inter- syzygial interval is 5-10 (in exceptional cases 4-18) muscular articulations. Charitometridae—The proximal syzygies have 15-22 septa, with the degree of strength 1-2. The areola is usually indistinct. There are 2 ventral beams, more or less indistinct. The complete septa are rather few. Irregularities in the distribution of the proximal syzygies are usual, these often replacing synarthries. The distal intersyzygial interval is 5-10 muscular articulations. Notocrinidae—The proximal syzygies have about 20 septa, with the degree of strength 2. Beyond the narrow areola is a broad and almost smooth area. Syn- apticulae occur. Most of the septa reach to the large nearly smooth area, and to the areola. The distal intersyzygial interval is 4-10 muscular articulations. Antedonidae.—The septa of the proximal syzygies have 11-24, and in the very large Heliometrinae 27-39, septa; the degree of strength is 1-3, reaching 4 in the Heliometrinae. The areola is lacking or indistinct except in Heliometra, in which in full-grown individuals it reminds one somewhat of the conditions described in Noto- crinus. Rows of tubercles occur only in the Heliometrinae. There are many com- plete septa, these being usually more than half the total number. The distal intersyzygial interval is 24 muscular articulations. Pentametrocrinidae.—The proximal syzygies have 11-18 septa, with the degree of strength 1-8. There is no areola. Half, or even the majority, of the septa are complete. The distal intersyzygial interval is 2-6 muscular articulations, and is rather irregular. Atelecrinidae.—The proximal syzygies have few septa, not more than 12, with the degree of strength 2-3. There is no areola. The septa are generally complete, sometimes finished, and more or less like tubercles. The distal intersyzygial interval is 2-7 muscular articulations. MOVABLE LIGAMENTARY UNIONS IN THE STALKED CRINOIDS In the pentacrinites the ligamentary articulations are generally immovable, and more or less pronouncedly synostoses. In the division series and in the first brachial ligamentary union, however, evident indications of synarthries are sometimes present. This is the case in Hypalocrinus naresianus and in the species of Neocrinus. In Neocrinus decorus the distal synostoses seem to indicate a development from this toward the trifascial type of articulation. The latter is typically developed in certain bathycrinids, such as Jlycrinus car- penterii, Bathycrinus aldrichianus, and B. australis. In Rhizocrinus lofotensis, as Sars showed and as Gislén confirmed, there is a form of loose synostosis in the liga- mentary articulations. In Democrinus rawsonii there seems to be a type intermediate between a synarthry and synostosis. The Phrynocrinidae and Hyocrinidae lack movable ligamentary articulations. A MONOGRAPH OF THE EXISTING CRINOIDS 39 MOVABLE LIGAMENTARY UNIONS IN THE COMATULIDS Gislén noted that although the degree of flexibility is very differently developed in different forms, nevertheless the synarthrial type taken as a whole is generally the same. Cryptosynarthries may appear in the Comasteridae. In these, as in pseudosyzygies and in the usual syzygies, the flexibility is practically nil. As viewed from the dorsal side synarthrial articulations show a fair amount of variation. The articular line may be straight and not accompanied by any process; there may be a more or less strongly developed synarthrial tubercle; or there may be a more or less strongly developed synarthrial backward projection. Gislén said that it is important to differentiate between synarthrial tubercles and synarthrial backward projections. In the former only a dorsal prolongation of the synarthry occurs; this causes an enlargement of the synarthrial faces, but gen- erally no increased flexibility worth mentioning, as the synarthrial fibers are not elongated. To the latter type, on the other hand, great possibilities as regards increased flexibility are always attached, with long ligamentary fibers in large lateral notches between the ossicles. There is, as a rule, very little flexibility in the Oligophreata from the Comasteridae to the Calometridae. Among these the backward synarthrial projection is very slight or lacking. On the other hand in the Macrophreata, together with most of the Thalassometridae and Charitometridae, the synarthries permit especially strong lateral flexion, and the backward synarthrial projections are often enormously de- voloped. These reach their maximum in the Zenometrinae and Bathymetrinae in the family Antedonidae. Extraordinarily strongly developed synarthrial tubercles appear sporadically in different families, for instance in species of Amphimetra, in Neometra diana, and in Perometra diomedeae. Gislén said that in the youngest stages the synarthries tend toward synostoses, but a stout areola appears very early. Another type of mobile articulation, trifascial ligamentary union or ligamentary articulation, with dorsoventral flexion only, is found in some cases in the recent co- matulids. This appears only in the articulation between the radials and the IBr. It has been observed in Pontiometra andersoni and in Stephanometra spicata. Stand- ing very near this type is the radial articulation in Himerometra martensi in which the muscles are indicated only as rudiments in the shape of low bands. In the Comasteridae and Calometridae also there is a form of articulation between the radials and the IBr, that approaches this type. This type of trifascial articulation, according to Gislén, possesses its greatest interest in that several of the Jurassic comatulids seem to have a similar articulation between the radials and the IBry. Comasteridae——From a pronounced synarthrial type the articulation seems sometimes to approach the cryptosynarthry, observed among the slender armed varieties of the species of Comatula. The species studied by Gislén have fairly typical synarthries. Divergencies from this have been observed in the synarthry between the elements of the IBr series. In large species, such as Comanthus japonica, the areola around the lumen is very extensive, and the two ligament pits are not equally curved, but are excavated irregularly, here and there. Here the possibilities of flexion are slight. The same is the case in those species of Comaster in which the synarthries are replaced by pseudosyzygies. The small species of Comissia, as well 40 BULLETIN 82, UNITED STATES NATIONAL MUSEUM as Capillaster sentosa, have an articulation between the elements of the 1Br series that approaches a cryptosynarthry. Capillaster may sometimes have only a single synarthry in each arm—between the first two brachials. Comatula and Comatulella (?) alone lack synarthries. Synarthrial tubercles and synarthrial backward projections are lacking. Zygometridae.—The articulation between the elements of the IBr series is de- veloped as a syzygy. Normal synarthries always appear between the two ossicles following each axillary. There are scarcely any synarthrial tubercles. Himerometridae.—Gislén did not examine any of the synarthries in the IBr series, but he found that the following synarthries are typical. The synarthrial tubercles are often knoblike and projecting. There are no synarthrial backward projections. Mariametridae.—In Stephanometra the synarthry in the [Br series shows a pair of deep and stout pits on either side of a fairly narrow dorsoventral beam. About the periphery is a broad contact margin. The capacity for movement is extremely slight. The more distal synarthries are typical. There are no synarthrial tubercles. The synarthry in the 1Br series in Lamprometra palmata corresponds in a high degree to that in Stephanometra. The more distal synarthries are high and laterally com- pressed, as in all forms with numerous arms. Colobometridae.—In Pontiometra andersoni the synarthry in the [Br series cor- responds to that in Stephanometra. The more distal synarthries are high and laterally compressed, more so than in other multibrachiate types. There are indistinct syn- arthrial tubercles. In having rhomboidal joint faces Oligometrides represents a unique type. The synarthries in the other forms offer very little of interest. Synarthrial backward projections are absent or, if present, weak. Gislén studied the synarthries in the 1Br series in Austrometra, Cenometra, and Cyllometra. Tropiometridae.—The synarthrial faces of the articulation between the elements of the IBr series in Tropiometra afra macrodiscus remind one very much of the same faces in the large species of Comasteridae. The joint pits are scarcely marked plainly, are bounded indistinctly, and are rather shallow. The capacity for movement is slight. In the synarthries between the first two brachials the articular pits are not developed in a semilunar shape, but only the dorsal half of each half moon is found marked. Both the fossae, however, are well circumscribed here. Synarthrial tuber- cles are absent or indistinct. Calometridae—The radials are often produced anteriorly in the interradial angles between the IBr,, and possibly as a result of this the first synarthrial face shows large wing-shaped appendages on both sides of the joint face. The union as seen from the dorsal side is often very close, and the mobility therefore presumably very slight. The fossae are, however, well marked. The more distal synarthries are of the usual type. Usually the synarthrial tubercles are indistinct, but they are extraordinarily large in Neometra diana. In Calometra discoidea there are synarthrial backward pro- jections. Ptilometridae.—There is never any synarthrial backward projection. The ca- pacity for lateral motion is slighter than in the Thalassometridae. Thalassometridae.—The contour of the synarthrial face is often somewhat angular, owing to the lateral flattening or “‘wall-sidedness” of the post-radial series. The synarthrial face is somewhat produced into an acute angle dorsally, where synarthrial A MONOGRAPH OF THE EXISTING CRINOIDS 4] tubercles are found. ‘This is often developed in connection with longitudinal crests on the proximal brachials. There is often a synarthrial backward projection, and at the same time large notches between the ossicles, giving the capacity for increased lateral flexion. The appearance of the synarthrial face is much like that in the follow- ing family. Charitometridae.—The synarthrial faces in this family correspond in many ways to those in the family preceding. There is often a longitudinal crest on the proximal ossicles, and sometimes synarthrial backward projections. There are often large lateral notches. The synarthrial face often shows a likeness to that of certain ante- donids in being restricted to a rather small proportion of the joint face. Notocrinidae.—The synarthrial face reminds one of that in the Thalassometridae. The median dorsoventral ridge is, however, extraordinarily stout, the areola surround- ing the lumen very broad, and the ligament fossae deep. There are no synarthrial tubercles. There is a weak synarthrial backward projection. Antedonidae.—Great variation is found in this family. Gislén said it looks as though one might trace the appearance of the synarthrial faces back to two types. The first of these is like that in Asterometra and Notocrinus in that the articular facet occupies practically the whole of the joint face, and the synarthrial backward projec- tion is moderately developed; there are rarely any synarthrial tubercles (but these are strongly developed in the Perometrinae). The Antedoninae belong here, and also the Perometrinae and Isometrinae and probably the Thysanometrinae. The second type has the synarthrial facet restricted to the center of the joint face and beyond this a broad area without ligament fibers; the synarthrial facet is often produced dorsoventrally, and the backward synarthrial projection is generally enormous. Here belong the Heliometrinae, Zenometrinae, and Bathymetrinae. Pentametrocrinidae.—The synarthrial facet occupies practically the whole of the joint face, but relatively somewhat less in Pentametrocrinus. It is narrower in Thau- matocrinus, evidently because of the 10 arms. The fossae are moderately deep. The synarthrial backward projection is distinctly pronounced in Pentametrocrinus, but imperceptible in Thawmatocrinus. Atelecrinidae.—The synarthrial facet is very much like that in the Zenometrinae. The synarthrial backward projection is distinct. ARTICULATIONS OF THE PINNULARS IN RECENT/CRINOIDS Gislén said that on the whole the pinnular articulations seem to be very uni- formly constructed in all the recent crinoids, though there are wide variations in several details. The typical form may be described as follows. The articulation between a brachial and the first pinnule is an articulation of rather ordinary muscular type. While the angle between the transverse ridges of the two joint faces on the distal end of an axillary is about 90° (for instance in Heliometra and Himerometra—the axillary being viewed from the distal end), the transverse ridges of the pinnule facet and of the brachial joint face form a considerably greater angle usually (Heliometra 115°, Metacrinus 150°, Himerometra about 165°, and Calo- metridae and other types 180°). The pinnule facet on the brachial is thus twisted nearly halfway around, so that the side that should be the outer lateral becomes the 20824440——4 42 BULLETIN 82, UNITED STATES NATIONAL MUSEUM inner median, and vice versa. It reaches such a position that its muscular fossae still continue adjacent to the muscle pit of the pinnular side of the articular face of the brachial. The dorsal ligament of the pinnule facet, on the other hand, seems to be removed almost to the maximum distance from the dorsal ligament of the brachial facet, if the brachial be viewed in plane projection. The pinnule facet forms an angle of about 110° with the articular face of the brachial, in, for instance, Helio- metra and Himerometra. At the distal end of the IBr, in Metacrinus nobilis tenuis the corresponding angle was 140°. The dorsal ligament of the pinnule facet is some- times better developed on the side adjacent to the median part of the brachial, and narrower on the side facing the lateral portion of the brachial. On the ventral side of the transverse ridge the lateral portion, on the contrary, is always more strongly developed. The intermuscular crest or furrow therefore bends considerably toward the median part of the brachial. The proximal part of the first pinnular answers to the pinnular joint face described on the brachial. Here, however, the articular fossae are often indistinctly marked, and the joint face only bulges irregularly, especially in small forms. When the muscles in this articulation are contracted the pinnule is flexed aborally and toward the side of the arm. The proximal and distal joint faces on the first pinnular are very nearly parallel. Also the distal joint face of the first pinnular shows here, usually, an almost straight muscular articulation. This, however, is oriented at about a right angle to the proximal facet of the ossicle, so that the transverse ridges of the two joint faces form an angle of about 90°. The proximal facet of the second pinnular corresponds to the distal facet of the first pinnular. In some comatulids this articulation is most strongly developed on the adoral side. On contraction of the muscles in this articu- lation the pinnule is flexed ventrally. The distal end of the second pinnular also has a transverse ridge, and this again makes an angle of 90° with the proximal face of the ossicle; here, therefore, it is a dorsoventral crest. But the articulation between the second and third pinnulars is differentiated from the two preceding by the appearance of small ventral muscle bundles on either side of the ridge, and usually by the absence of the ligament pit which, if it is present, marks the boundary for the dorsal development of the dorso- ventral crest. In these two features this pinnular articulation differs from a synarthry that otherwise in its general character it rather strongly resembles. As is shown by examples from the distal segments of the distal pinnules it may be considered as derived from an ordinary muscular articulation through the disappearance of the dorsal ligament. When asymmetry occurs in this and the following articulations it is the aboral side of the joint face that becomes the more strongly developed. This is contrary to the conditions found in the articulation between the first and second pinnulars (with some exceptions, as for instance Metacrinus). This stronger aboral development of the articulations of the distal pinnules expresses itself very strikingly in the devel- opment of the combs in the Comasteridae, and of the calcareous flanges protecting the gonads in Austrometra, Isometra, and certain Charitometridae. The ventral pits on either side of the dorsoventral crest for the reception of the muscle bundles are usually least distinct in the articulation between the second and third pinnulars, where they may even be entirely lacking; but in the articulations A MONOGRAPH OF THE EXISTING CRINOIDS 43 following they are usually well circumscribed. In the not uncommon autotomy of the pinnules the fracture therefore always occurs between the second and third pinnulars—the articulation that comes nearest the ligamentary articulation of the arms. The articulation between the third and fourth pinnulars, like all those following, is similar to that between the second and third, but the pits for the muscular fossae are better marked. The development of the muscle bundles shows great variability. In the distal parts of the proximal pinnules the movement thus takes place laterally. In the distal articulations of the distal pinnules there is usually a pit in the dorsal portion of the dorsoventral crest. This pit harbors a little dorsal ligament. Here a possibility is present for slight flexibility also in a dorsoventral direction, most strongly developed in certain Comasteridae, as for instance Comatulella brachiolata. Hyocrinidae.—In Ptilocrinus pinnatus the transverse crest of the face of the pinnular forms an angle with the brachial articular face of over 150°. The calcareous lamella over the muscle insertion projects forward as a large spout-shaped process. The proximal joint face of the first pinnular is peculiarly twisted in relation to the distal face on the same ossicle. This last face is a straight muscular articulation without interarticular ligaments. The second and following articulations are very close; they possess a dorsal ligament pit, but except for this lack nearly all sculpture. As is evidenced by sections these unions are in reality synostoses and therefore lack muscles. Pentacrinidae.—The two muscle attachments on the pinnular face of the IBr. in Metacrinus are very small and close together. They project rather a long way into the brachial, and are overshadowed by a projecting calcareous lamella. The proximal face of the first pinnular makes an angle between the transverse ridge and the inter- muscular furrow of about 50°, this angle facing towatd the median portion of the brachial. The distal end of the first pinnular shows a typical straight muscular articulation with the adoral parts most strongly developed. The distal end of the second pinnular shows only indistinct traces of muscular attachments. As in the immediately following ossicles the adoral side is the one most strongly developed. The synarthroid type of articulation is very little accentuated here, the union ap- proaching the synostotic. There is, however, a dorsal ligament pit. The joint faces, which follow rectangularly in transverse section, have small ventral muscle pits and retain the dorsal ligament pit; the ligament pits are shallow, becoming very insignifi- cant distally. The dorsoventral crest is bounded dorsally by the dorsal pit. The articulations of the distal pinnules show the same type. The dorsal pit is retained, and the dorsal ligament fossa is more distinctly accentuated. The distal segments of the distal pinnules in transverse section are either rectangular (Meta- erinus) or triangular (Diplocrinus) and have only a very thin calcareous bridge sepa- rating the furrow for the ventral soft parts from the dorsal nerve. This thin bridge is often sievelike (?for the ciliated pits) and the joint therefore acquires—as seen from the ventral side—a certain likeness to the Silurian forms transitional between the canaliculate and noncanaliculate types. As seen in transverse section the pinnulars of the Pentacrinidae, by their rectan- gular or triangular form, remind one rather strongly of those found in the Thalassome- tridae. At BULLETIN 82, UNITED STATES NATIONAL MUSEUM Bathycrinidae.—In this family pinnules are absent from a number of the proximal brachials. The pinnular joint face on the more distal brachials reminds one rather strongly of that in Ptilocrinus. The articulation between the first and second pinnu- lars is a muscular articulation of the usual brachial type. The dorsal ligament is, however, narrow and inconsiderable. The distal pinnule segments are distinguished chiefly by the large and extraordinarily thin lateral processes arising from the body of the pinnule segment itself and embracing the ventral soft parts. Comasteridae.—Aside from the development of combs, the most characteristic feature of the pinnules in this family is the enormously strong development of the muscle bundles, which is especially remarkable in view of the weak development of the muscular unions between the brachials. The angle that the transverse ridge of the brachial face and that of the pinnular face make with each other varies from about 130° in Comaster to 180° in Comissia and Capillaster. The cavity in which the muscles of the first pinnular are inserted on the brachial is usually not very deep. In the proximal comb-bearing pinnules very strong muscle attachments appear, which occupy the greater part of the surface of the facets of the articulations beyond that between the second and third pinnulars. This speaks in favor of these pinnules being specially adapted for making rapid and powerful movements. In the distal parts of the pinnules, where the combs occur, muscles are developed which are enormously more powerful on the side of the dorsal crest on which the teeth are developed. In the distal combless pinnules the muscles are more moderately developed. In the articulation between the third and fourth pinnulars and beyond there appears as usual a small dorsal ligament fossa. Zygometridae, Himerometridae, Mariametridae, Colobometridae, and Tropiometri- dae—Except in unusual cases the muscular attachments are everywhere small. The pits in the brachials into which the muscles from the first pinnular are inserted are usually fairly deep, but there are some exceptions, as in Himerometra. Between the second and third pinnulars scarcely any muscular fossae are found in the large proximal pinnules of Himerometra. In the very long and flexible P; in Pontiometra andersoni the articulations beyond the one between the second and third pinnulars are provided with large lateral muscu- lar fossae. In the more distal parts of the pinnule, as in the distal pinnules, the dorsal part of the dorsoventral crest disappears, and we therefore get a more or less dis- tinctly marked dorsal ligament fossa. It must be observed that the lumen in this pinnule is unusually large. We may compare the conditions here with the size of the jumen in the immovable proximal pinnules in Stephanometra. Gislén remarked that in this connection it may be pointed out that the Macro- phreata, provided with large brachial muscle attachments, have in the same way large nerve lumina in the brachials, and a large central organ for these nerves in the centro- dorsal. Stout dorsal nerves are thus correlated with large muscle bundles and a strong capacity for movement or swimming. In Stephanometra the stiff and styliform proximal pinnules differ considerably from the rest of the pinnules in regard to their articulations. The articulation be- tween the first and second pinnulars is, as usual, a muscular articulation, and that between the second and third pinnulars is a pinnular articulation in which, however, A MONOGRAPH OF THE EXISTING CRINOIDS 45 only the muscle of the aboral side is noticeably developed. In the articulation be- tween the third and fourth pinnulars there is a great reduction of the sculpture of the joint face, and from the articulation between the fourth and fifth pinnulars onward we get typical synostoses. This gives the styliform appearance and the stiffness to the pinnule. In certain of the Colobometridae there is a similar, though usually not so extreme, stiffening of the proximal pinnules. In Austrometra the genital pinnules are provided with winglike processes, as is the case in the Bathymetridae. Here, however, the processes are extended almost horizontally so that the ossicle acquires a more disk-shaped appearance. Calometridae, Thalassometridae, and Charitometridae.—In all these families the pinnule segments after the third or fourth are decidedly triangular, though the more proximal segments—the second-fourth—in the proximal pinnules are sometimes rectangular. The triangular form of the pinnulars may sometimes be indistinct in the proximal pinnules, these being more or less rounded, as in Neometra and Stylo- metra. In these families the genital pinnules often have flattened segments with small lateral processes that protect the gonads. In the family Calometridae the proximal pinnules have enormously enlarged first and second pinnulars. The pinnule face on the brachial occupies a great part of its ventral surface. The muscle attachment pushes its way rather deeply into the brachial segment, as is usually the case also in the two other families. The transverse crest of the pinnular face is twisted to an angle of at least 180° in relation to that of the brachial face. The articular fossae between the proximal pinnulars are extraordinarily well developed in all the proximal pinnules. The interarticular ligaments are narrow or rudimentary. The union between the second and third pinnulars soon, however, becomes much closer than those between the other segments, and after about P, acquires a synostotic character that is especially evident in Neometra. Processes are also found on the earlier ossicles of the proximal pinnules, though less localized or marked, in some of the Charitometridae, and outside of this group in Catoptometra and in certain comasterids, as for instance Comanthina belli. Notocrinidae and Macrophreata.—Gislén remarked that in regard to these forms there is not much to be said. Within this group the muscle attachments are very well developed; otherwise the pinnule articulations are fully typical, except for the cases mentioned below. The distal segments of the distal pinnules are more or less rounded, and their joint fossae are provided with the usual small dorsal fossa. The more important exceptions to the typical structure are the following. The Heliometrinae have long proximal pinnules with very short segments. Here the pinnular muscle fossae are very strongly developed and the hypertrophy culminates between the distal pinnulars where, as in the Comasteridae, the most powerful muscle ‘s found on that side of the ossicle where the rudimentary comb is situated—the aboral side. Another exception to the normal structure is found in [sometra. In this genus the third and fourth segments of the pinnules from about P, are greatly expanded, evidently to serve as a protection for the gonad. The hypertrophy is most strongly pronounced on the aboral side. Gislén had no opportunity for studying the pinnules of the Thysanometrinae or Atelecrinidae. 46 BULLETIN 82, UNITED STATES NATIONAL MUSEUM THE PINNULE GAP The earliest pinnules, appearing at the ends of the arms, are not formed until the arms have developed a certain number of brachials. The gap between these and the arm bases is filled by the successive appearance of P;, then usually P3, and finally P2 and Py. Gislén pointed out that in the Macrophreata the first pinnules appear on the tenth-twelfth brachials when about 15 brachials have developed. Inother comatulids the pinnule gap is possibly in certain cases somewhat smaller, and possibly the form- ation of the proximal pinnules does not always take place as regularly as was assumed by Mortensen. In full-grown comatulids the pinnule gap is retained in Atelecrinus, in which up to 15 brachials lack pinnules, and in a reduced degree in Comatilia, Hypalo- metra, Clarkometra, Balanometra, and Pentametrocrinus, as well as in certain Colo- bometridae and Perometrinae. In the two last-mentioned groups, however, only P, and in rare special cases P, are lacking. Gislén said that the author has attempted to explain the cause of this by assuming that the comatulids, before the appearance of the first pinnules, might be considered “a fairly typical representative of the Flexibilia Impinnata.’’ He remarked that the author also mentioned several other similarities to this group that support the assump- tion of a closer relationship. Regarding that portion of the arm proximal to the second syzygy these assumed similarities are: (1) Short, oblong, wedge-shaped brachials. (2) Absence of syzygies after the first syzygial pair—a corollary, according to Gislén, of the preceding. Gislén noted that the author said the interval between the first and second syzygies is the greatest in the whole arm, but that in reality this concerns only those comatulids that have very short intersyzygial intervals distally, and the second syzygy at about the tenth brachial. In certain Mariametridae, in Ptilometra, and in the Notocrinidae, Pentametrocrinidae, and Atelecrinidae, the interval is often very short, and the second syzygy appears as early as about the seventh brachial. (3) Before the second syzygy the pinnules are said to ‘‘be absent until after the brachials are completely formed;’ Gislén says that this sentence becomes explicable only if understood as meaning “until after the brachials that form the second syzygial pair are completely formed.”’ This assertion, according to Gislén, does not concern forms in which the second syzygy first appears far out on the arm—certain Oligophreata. Gislén said that according to the author there often appear among the young of comatulids these similarities to the Flexibilia Impinnata: (1) All the brachial articula- tions are of the same nature. This, he says, is not correct, for when the earliest pinnules begin to be developed muscular and nonmuscular articulations can be distinguished histologically. (2) The flexibility of the arm is caused by a continuous ventral muscle band which is not subdivided into interbrachial muscles. Gislén remarked it is easy to prove by dissecting or sectioning young comatulids that interbrachial muscles really occur. There are also other similarities found there, but they are of a too general nature to point directly toward the Flexibilia Impinnata—Gislén remarked that, except for the paragraphs concerning the pinnule gap and the formation of a centrodor- sal, they might just as well be mentioned as similarities with the Inadunata. Gislén said that for a long time the pinnule gap was a puzzle to him, until at last through comparison with the condition in Phrynocrinus he began to see the matter more clearly. In Phrynocrinus the disk extends outward at least as far as the twen- tieth brachial, half burying a number of pinnules in the perisome. The first pinnules A MONOGRAPH OF THE EXISTING CRINOIDS 47 appear at about the eighth brachial. Proximal to this they are suppressed completely, evidently because, even if they had been present, they would have been far from reach- ing up to the surface of the disk and therefore would have been of no use for obtaining nourishment here. It seemed to Gislén that if the pinnule gap were attributable to historical factors something similar should be found in the fossil ancestors of the com- atulids. In all the smallest specimens of recent pentacrinites known—species of Metacrinus and Neocrinus decorus—all the pinnules are developed, and the proximal pinnules are the largest. There is no gap in the pinnulation. Gislén said it was not until he turned to the fossil species of Pentacrinus—the ge- nus that, according to him, most closely approaches the comatulids—that he found the solution of the problem. In Pentacrinus the ventral perisome rises to a height of half the length of the arms, to the fifth or sixth arm division, or to at least the fortieth brachial if all the series are reckoned consecutively. In Heliometra glacialis, which has an unusually voluminous disk, this reaches, when the arms are folded and seen in profile, up to the fifteenth to the eighteenth bra- chial, and the anal tube in one case was observed to reach to the twenty-first brachial; when the arms are extended the disk reaches only to the eighth brachial, or to the height at which it is joined to thearms. In Notocrinus virilis the disk with the arms folded reaches to the eighth or ninth brachial, and joins the arms at the fifth or sixth brachial. In full-grown individuals of Hathrometra tenella var. sarsii the disk is at- tached at the third or fourth brachial, and when the arms are folded to the fifth to sev- enth brachials; in this species the anal tube is very long and reaches to the twelfth bra- chial. In the pentacrinoids of this species the disk reaches about as high as it does in full-grown individuals, or to the sixth brachial, the anal tube reaching to the tenth brachial, though the brachials are relatively longer in the young than in the fully grown. Instill smaller pentacrinoids in which the budding of the pinnules in the distal part of the arms has only recently begun the disk reaches only to the first brachial, and the anal cone to the fourth; but in these pentacrinoids the reduction of the basals and radials has not as yet set in to any appreciable degree. According to Gislén the difference in the height of the disk with the arms extended or brought together is hardly noticeable in forms having thin disks. For instance in Heterometra crenulata the disk is attached as far as the sixth or seventh brachial but does not reach noticeably higher when the arms are folded. In Asterometra anthus the disk reaches to the third brachial. If the proximal pinnules are very long and slender they may, in spite of a fairly high disk, be of use in taking up nourishment or in functioning as tactile organs about the mouth, as in Heliometra glacialis. With an enlarged disk, however, the point may be reached where the proximal pinnules cannot reach to the ventral side of the disk; they are then excluded from playing any part in conveying nourishment, or from functioning as tactile oral pinnules. If the perisome then bulged out around them and they became imbedded in it, evidently their reduction and disappearance would not be far off. Gislén noted that this has often been the case in a number of species of Penta- crinus, and that the subangularis group of species shows evident reduction of the proximal pinnules. Chiefly on the basis of characters connected with the stem, but 48 BULLETIN 82, UNITED STATES NATIONAL MUSEUM also bearing in mind the more or less strong reduction of the proximal pinnules, Gislén proposed the new generic name Seirocrinus for Pentacrinus subangularis and its imme- diate relatives. Gislén said it is evident that the hypothetical ancestor of the comatulids unites in itself characters belonging to both Seirocrinus and Pentacrinus, sensu stricto. It approaches Pentacrinus in its short stem consisting, after the development of the first nodal, of nodals only, as well as by the strong cirri. He said that among known forms Pentacrinus dargienesi comes nearest to the comatulids. The stem did not reach more than 60-70 mm. in length, while the cirri were 100 mm. long. The comat- ulids approach Seiroerinus through the defective pinnulation that appears in their ontogeny, which probably, in Gislén’s belief, is caused by the perisome once having swelled out between the arms. Gislén noted that it is not only among the Articulata discussed above that a gap in the proximal pinnulation occurs. Among recent forms pinnule gaps appear in the adults of various stalked types. They are met with throughout the Bathycrinidae. In Bythocrinus the first pinnule appears on from the fourth to the twelfth brachial (in Bythocrinus ef. braueri on the fourth to the sixth brachials; in B. braueri on the eighth; and in B. chuni on the tenth to twelfth) ; in Democrinus (rawsonit and weberi) the lowest pinnule is on the sixth brachial; in Rhizocrinus (lofotensis) it is on the eighth; in Bathycrinus it is found on from the eighth to the tenth (in Bathycrinus pacificus on the eighth; in B. aldrichianus on the ninth; and in B. serratus on the ?tenth); in Ily- crinus it is found on from the ninth to the ?twelfth (in Ilycrinus australis, ninth to eleventh; J. carpenterii eleventh or twelfth; and in J. complanatus, tenth to twelfth) ; and in Monachocrinus (sexradiatus) it is found on from the tenth to fourteenth. In the Phrynocrinidae only the arms of Phrynocrinus nudus are known. The first pinnule appears here on the eighth brachial. In the Hyocrinidae also a proximal pinnule gap is found, as one pinnule to the right and one to the left is lacking. The first pinnule therefore appears on from the fourth to the sixth brachial, the variation being due to the difference in distribution of the syzygies. Gislén said that if his idea that the bathycrinids are derived from the Bourgue- ticrinidae and the latter are descended from a form nearly related to the comatulids and thiolliericrinids is correct the explanation of the pinnule gap in the bathycrinids will be the same as that given in the case of the comatulids. The disk in the bathy- crinids of the recent seas does not reach far up on the arms—in Rhizocrinus, with 5 arms, to the second brachial, and in Jlyerinus with 10 arms also to the second brachial (fourth postradial ossicle). Gislén said it seemed to him in many cases as though this may be explained partly by the disk having sunken in between the greatly elon- gated basals (in Bythocrinus, Democrinus, and some species of Monachocrinus). In Phrynocrinus the very large disk reaches up to about the twentieth brachial. The disk and arms of Nawmachocrinus are not known; in this genus the radials are greatly elongated. A high disk is also found in certain of the Hyocrinidae. In Ptilocrinus, in which it extends to the sixth brachial, the arms are grown fast together as far as the fifth brachial. In Calamocrinus, in which the disk is connected with the arms up to the ninth or tenth brachials, the anal tube reaches to the I1Br, (the axillary tenth post- A MONOGRAPH OF THE EXISTING CRINOIDS 49 radial ossicle). On the other hand, Thalassocrinus and Hyocrinus have low disks. The position of Gephyrocrinus is intermediate. Gislén concluded that a voluminous disk is of very common occurrence among the primitively organized Articulata. Sometimes this high disk is combined with loss of the proximal pinnules. In the comatulids, according to him, there is evidence that defective pinnulation was once present in the occurrence of a proximal pinnule gap during the juvenile stages. According to Gislén the genital glands of the pinnules are to be considered as originally processes from the primitive gonad, to which the present axial gland corre- sponds, losing their connection with it only at a later stage. There is no definite correlation between the pinnule and the gonad, as has been assumed, for in some recent forms, as Metacrinus, Notocrinus, and Comatula, the gonads do not appear in the pinnules, but in the arms. All the ontogenetic indications point to a late migra- tion of the gonads into the arms and pinnules. We may therefore suppose, according to Gislén, that the ventral sack originally enclosed, besides a part of the intestine, at least a part of the gonad. When the ventral sack later absorbed the whole of the disk, we may assume that the gonad extended under all the interadii. The perisome then rose between the arms and caused, at least in a number of forms, a reduction of the proximal pinnules. When the disk once more dwindled in size the migration of the gonads into the arms began. In a number of forms there still remains, as a re- minder of the high disk and pinnule reduction, a proximal pinnule gap in the adults, best marked in the Bathycrinidae and Atelecrinidae, and appearing during ontogeny in all comatulids. Gislén said it is noticeable that the Flexibilia, which never developed pinnules and had no ventral sack, always had a perisome extending high up between the arms. Gislén remarked there seemed to him to be a certain amount of interest in trying to ascertain the total volume of the gonads in some of the comatulids. He therefore detached all the gonads from an arm in some specimens with greatly swollen genital pinnules, weighed the mass thus obtained, multiplied this by the number of arms, and divided the result by the specific gravity (1.1), and in this way obtained the total volume of the gonad. From this he calculated the height to which the disk would be raised were the gonads included in it, obtaining the following results: Heliometra glacialis (with the genital pinnules relatively inconspicuously distended), weight of gonads, 7.43 gr.; volume, 6.76 cc.; radius of the disk 10.5 mm.; raising of disk by inclusion of the gonads, 19.5 mm., corresponding to 14 brachials. As the disk reached to the tenth brachial, a disk with the gonads included within it would reach to about the twenty-fourth brachial. Antedon petasus; weight of gonads, 0.53 gr.; volume, 0.48 ec.; radius of the disk, 3.3 mm.; raising of disk by inclusion of the gonads, 14.1 mm., corresponding to 18 brachials. As the disk with the arms folded reaches to the seventh brachial, a disk with the gonads included would reach to about the twenty- fifth brachial. Compsometra serrata; weight of gonads 0.36 gr.; volume, 0.33 cc.; radius of the disk, 3.4 mm.; raising of the disk by inclusion of the gonads, 9.1 mm., corresponding to 16-17 brachials. As the disk reaches to the sixth brachial, it would reach, with the gonads included, up to about the twenty-second brachial. Gislén said that it seemed as if the enlargement of the disk under the conditions given above was rather considerable and fully sufficient to account for the reduction of the proximal 50 BULLETIN 82, UNITED STATES NATIONAL MUSEUM pinnules. He said further that this assumption of the extension of the gonads under the interradii being the chief cause of the enlargement of the perisome makes it clear also why in young comatulids we do not get a high disk in connection with the pinnule gap, for at this stage the gonads are wholly rudimentary. Dr. Gislén’s explanation of the pinnule gap is most interesting and ingenious and certainly is worthy of serious consideration. But the author still maintains his belief in his own interpretation of the conditions in the early pentacrinoids. There is, of course, the possibility that both of us are in error in ascribing the pinnule gaps in the recent crinoids to factors having their origin in phylogenetical antecedents. These gaps may well be nothing more than a response to the conditions of environment with no phylogenetical background at all. In the ease of the developing young of the comatulids the prime requisite is to secure food as early as possible. The larger the diameter of the circle combed by the food-collecting organs, the larger is the potential food supply. So pinnules first appear at the tips of the growing arms, where it is obvious they will be most effective. With the commencement of concentration of food upon the disk, resulting from the action of the pinnules at the arm tips, the protection of the disk becomes increasingly important, so that it would be natural to expect the tactile pinnules at the edge of the disk to put in an appearance. The need for the intermediate pinnules, which are either genital or supplementary tactile pinnules, or both, does not arise until later, so that their development is delayed in favor of that of the more important pinnules. The absence of a varying number of basal pinnules in such types as the Atele- crinidae and Bathycrinidae is probably due to the fact that, as a result of the crowding of the arm bases, there is no room for them. If present they would be without value to the animal, and so they are not developed. Such an explanation is easy to com- prehend if pinnules are regarded in the light of articulated appendages, but is less easy of comprehension if they are considered as reduced arm ramifications. INTERRELATIONSHIPS OF THE COMATULIDS Gislén’s rearrangement of the comatulid groups was given in detail in Part 3, pages 60-64. His notes on the evolution and interrelationships of the comatulids are of much interest, for his was the first serious attempt to harmonize in detail the systematic background of the recent and the fossil forms. The evolution of certain comatulid characters—Gislén said that originally cirri were arranged in five radial columns. It was not until the suppression of the inter- nodals in the pentacrinite stem and the increased shortening of the nodals took place that two alternating columns occurred, to conserve space, in each radius. Since after the development of the centrodorsal the number of the cirri continued to increase a third column appeared, and finally the arrangement became a series of closely crowded irregular alternating rows. This developmental course may be traced more or less distinetly in the ontogeny of certain comatulids. Centrodorsals with closely crowded alternating rows of cirri are therefore the most highly specialized. Originally the cirrus sockets had a distinct sculpture, but this has disappeared more or less completely among the younger forms. The cavity within the centrodorsal was originally, as all the fossil comatulids show, from fairly small to very inconspicuous, never larger than in the corresponding A MONOGRAPH OF THE EXISTING CRINOIDS oi recent forms. Gislén said it has been supposed by the author that from the consider- ably larger size of the centrodorsal cavity in the young we might conclude that a small centrodorsal cavity represents a more specialized stage. This feature according to Gislén has just as little phyletical significance as the longer segments in the young, and must be attributed to an as yet unfinished calcareous deposition in the walls around the chambered organ. Small and weakly calcified species always have, in comparison with the usual type in their family, an unusually large centrodorsal cavity. The stronger development of the basals is a more primitive feature, noticeable both during the phylogeny and during the ontogeny of the comatulids. The articular faces of the radials seem originally to have been moderately de- veloped and to have had an inwardly broader muscular fossa of about the same size as the interarticular ligament. Later this muscular fossa was reduced to a low broad band which may even disappear entirely, as in the Comasterida or Mariametrida, or may become enormously enlarged in comparison with the interarticular ligament, as in the Macrophreata. This expresses itself in the ontogeny by the young of the group first mentioned having relatively larger muscular attachments, while the young in the latter group have relatively smaller ones. A small radial cavity is found only in the later comatulids, from the Upper Cretaceous onward. A radial cavity filled with spongy calcareous tissue is never found in the earlier comatulids, but occurs in the recent Comasterida and Mariametrida. So far as we are acquainted with the fossil comatulids, syzygies appear in the older forms, but synarthries not until the younger ones. In the great majority of the older forms the syzygies seem to have been few in number, and their septa few. The arms were usually 10 in number, sometimes 5; if greater, the number of the component parts in the division series was very variable in older times. The evolution of the comasterid type ——The oldest comatulids we know are found in the lowest Jurassic and belong to the comasterid type. Even then this type had a highly developed centrodorsal with small, closely set, alternating cirrus sockets, usually indistinctly or not at all sculptured, and rather strongly reduced basals. It may therefore be assumed that the comasterid type even at that remote period had a fairly long developmental course behind it. The primary characteristics of the older forms were the extremely small centrodorsal cavity and the inwardly broader fairly large muscular fossae on the radial articular faces. The characteristics which even today distinguish the comasterids were developed early. The first form without cirri is met with in the Lower Cretaceous. The centrodorsal is always more or less flattened. The radial fossae are from the very first vertical, and the radial cavity large, its central depression broad and deep, in recent forms filled with spongy calcare- ous tissue. The muscular fossae in the younger forms are low and broad. The arms in fossil forms are very little known, and so far as we are acquainted with them they are 5 or 10 in number and uniserial. Two unique features which in recent times especially distinguish this group are the exocyclic mouth and the pear-shaped organs instead of sacculi. As primitive features which are retained in forms now living may be reckoned the clumsiness, the often undeveloped synarthries (as in Comatula and Comaster), the usually variable number of division series and of their component ossicles, the oc- 52 BULLETIN 82, UNITED STATES NATIONAL MUSEUM casional plating of the disk, the covering plates in Nemaster and Comatilia, the oc- casional appearance of the gonads (best developed in the posterior radii) in the arms, and the creeping mode of locomotion. Specialized features are the reduction of the cirri, the numerous arms and syzygial septa, the powerful hooks on the distal pin- nules, and the comb on the proximal pinnules and sometimes also on the distal ones. The evolution of the solanocrinid type——The solanocrinids appear first in the Middle Jurassic and become commoner in the Upper. Nevertheless they have centrodorsals that are more primitive than in Palaeocomaster, and the appearance of the cirrus sockets approaches more closely the pentacrinid type. Moreover, the basals in the oldest forms are strongly developed. It may be supposed, therefore, according to Gislén, that the solanocrinids developed into comatulids later than Palaeocomaster. It is also possible that the solanocrinids are derived from a different branch of the pentacrinids. In Palaeocomaster, and in its descendants the recent comasterids, there is found a strong tendency toward reduction of the cirri, as in Seirocrinus. There is never any reduction of the cirri in the Solanocrinidae, which always have the cirrus sockets large and powerful, as in Pentacrinus. Possibly, therefore, the type first mentioned may have descended from a form more nearly related to the former penta- crinid genus, and the second type from a form more nearly related to the later one. Among the solanocrinids Archaeometra is the most primitive in regard to the basals, which are very powerful. The cirrus sockets are still rather few and large. In Solanocrinus the number of cirrus rows and columns is increased, and in the Lower Cretaceous 15 columns of cirrus sockets become usual. In typical solanocrinids a centrodorsal with cirri in alternating rows is never attained. On the other hand, the genus Solanocrinus seems to show more primitive features in respect to the size and shallowness of the radial cavity and the sculpture of the cirrus sockets. Gislén remarked it may now be asked whether the solanocrinids have completely died out, or whether they survive among the recent forms. The former has probably been the case with a number of clumsy forms from the Upper Jurassic with biserial arms. In addition to these there are in the Upper Jurassic, as well as in the Lower Cretaceous, quite typical solanocrinids with monoserial arms. Among the younger of these we find columns of large cirri on the flattened centrodorsal; but the sculpture of the cirrus sockets has almost disappeared, and the muscular fossae are low and broad and are possibly sometimes absent. The centrodorsal cavity still continues small and shallow; the radial articular faces lean only inconspicuously inward, and the radial cavity is therefore large—possibly it is beginning to be filled by spongy cal- careous tissue. It seemed to Gislén very probable that in the recent suborder Mariametrida we have the recent descendants of the solanocrinids. The difference is that the sculpture of the cirrus sockets in the younger forms has been still further reduced and therefore at its highest consists only of an areola, and that the cirri, ‘generally at least, alternate distinctly. There are further the following specialized features—the reduction of the muscular fossae, distinetly indicated already in the Solanocrinidae, is often very advanced and sometimes complete (Pontiometra and Stephanometra); the radial cavity is generally filled up with spongy calcareous tissue; the syzygial septa have increased A MONOGRAPH OF THE EXISTING CRINOIDS 53 in number; and the basals are usually strongly reduced except in certain Himero- metridae and in the family Tropiometridae of the Tropiometrida. Gislén said that most of the primitive features are shown in the Zygometridae, certain Himerometridae, and the Mariametridae. Such features are: A syzygy between the elements of the [Br series in the Zygometridae; an indefinite number of components in the division series (in the Zygometridae and certain Himerometridae) ; few connecting fibers between the parietal and visceral layer of the coelome sacks in the disk so that the disk is easily thrown off (especially marked in the Zygometridae) ; the disk covered by calcareous granules; few syzygial septa (Mariametra and the Zygometridae); and undifferentiated proximal pinnules (Zygometridae and certain Himerometridae and Mariametridae). The whole of the suborder Mariametrida, with the exception of the small genera in the Colobometridae, has a rather coarse and clumsy structure, a small and shallow centrodorsal cavity, and sparse syzygies. Among these features the two first men- tioned and probably also the last are primitive. Besides, there are solanocrinid characters in the discoidal centrodorsal with rather stout cirri, as well as in the shallow and often meandering branched furrows that lodged the dorsal coelome between the centrodorsal and the radial pentagon. In regard to the structure of the centrodorsal, and in the Comasterida and Maria- metrida of the radial pentagon too, there seems, on the whole, to be evident a striving toward a similar ideal type. The further the comatulids progressed in their develop- ment, the closer they approached this goal as a rule, and it is therefore often more difficult to distinguish later forms of different groups than is the case with earlier ones. Gislén said that this explains also why the centrodorsal and radial ring in certain recent forms of Comasterida and Mariametrida, although in many instances easy to distinguish from each other, in some cases can not be diagnosed with certainty. The same is true with a number of younger fossil forms. There are types among the forms from the Upper Cretaceous known only from the centrodorsal that can only doubtfully be referred to the Comasteridae, or to the descendants of the solanocrinids. It seemed to Gislén that the parvicavus group of the genus Glenotremites may be as- sumed to belong to these latter because of the size of the cirrus facets. Presumably the angelini group may be referred here too. Gislén believed that in Cypelometra there is found a specialized descendant of the solanocrinids which as regards the appearance of the cirrus facets has reached the mariametrid stage. The evolution of the notocrinid and conometrid types.—Gislén said that these two form series may be treated most suitably in connection with each other. In the notocrinid series the most primitive type is Loriolometra, represented among the species belonging to Glenotremites by G. arnaudi and the essensis group. Loriolometra is continued by Sphaerometra, represented among the species of Glenotre- mites by the paradorus and possibly the rotundus groups. With the genus Semio- metra the type has acquired so many new features that Gislén believed it necessary to place it in the family Paleantedonidae. The series had originally rather few large cirrus sockets arranged in columns, with a weak transverse crest, often peripherally striated, protruding basals, broad free margin of the dorsal side of the radials, rela- tively large radial cavity, a more or less distinct dorsal star, and, most important of all, the specially characteristic deep radial pits in the centrodorsal for the reception 54. BULLETIN 82, UNITED STATES NATIONAL MUSEUM of the dorsal coelome. The youngest representatives have numerous small cirrus sockets without sculpture, in alternating rows, concealed basals, no free dorsal margin of the radials, and a fairly small radial cavity; the radial pits are shallow or obliterated. Gislén observed a rosette in Semiometra impressa. He said that the recent form Notoervnus may properly be included in this series, as it is undoubtedly nearly related to Loriolometra, from which it differs in its shallower and wider radial coelome pits, by having a relatively larger centrodorsal cavity and, in full grown individuals, by the absence of sculpture on its cirrus sockets. The later forms of this series advance to- ward group a of the Macrophreata (see Part 3, page 63, under Macrophreata). The lettensis group with its fine radial pores corresponds as regards this character to the family Asterometridae in which the radial pits are deep, but very narrow. The conometrid series, relatively poorly represented among species in which the radial pentagon is known by the genera Amphorometra, Placometra, Conometra, and Jaekelometra, is more numerously represented within the genus Glenotremites by the anglesensis and pellati groups. This series is distinguished from the preceding one by the absence of radial coelomic pits. Among recent forms it has its analogy in the Thalassometrida, sensu stricto. Gislén said it is very remarkable what a large number of species of which only the centrodorsal is known are found in both these series, contrasting with the conditions in the solanocrinid series. In the genus Glenotremites of the solanocrinid type only five species are found (the ezilis group). But there are 30 species that can certainly be included in the two series now under discussion. This is due in the case last men- tioned to the fact that the elements of the radial pentagon are loosely connected both with each other and with the centrodorsal. In the recent Thalassometrida the corre- sponding ossicles very soon fall apart when boiled in caustic potash. A feature to be noticed is that in the recent Mariametrida the centrodorsal and the radial pentagon are very firmly united with each other, exactly as appears to have been the case in the fossil solanocrinids. It seemed scarcely probable to Gislén that the notocrinids could have descended from any of the comasterids or solanocrinids. In the last named, to be sure, we some- times find diverticula from the dorsal coelome between the centrodorsal and the radial pentagon, indicated by shallow branched meandering furrows. These, how- ever, are entirely different from the deep and broad perpendicularly penetrating pits which occur in Loriolometra. Because of their depth—they traverse almost the whole of the centrodorsal even in full-grown individuals—it is easy to assume their pro- longation into the stem in stalked young. Although there are a good many similarities in regard to general features between Archaeometra and Amphorometra, Gislén said it is not very probable that the cono- metrids are derived from either of the two families first mentioned. This is made still less likely by the fact that in the Upper Jurassic we meet with a form which it seems might fairly well be considered a primitive ancestral form of the Thalassometrida and Conometridae in general, and of Placometra in particular. Gislén said we therefore reach the conclusion in these two cases also that these two families possibly represent one, or perhaps two, special offshoots from the pentacrinid stock, which in that case have attained the comatulid type by independent paths. A MONOGRAPH OF THE EXISTING CRINOIDS 55 In Pterocoma pennata from the Upper Jurassic we see a form exhibiting many features in common with certain comatulids from the Upper Cretaecous, as well as with the Thalassometrida. Gislén said that this early form does not show the least approach toward the contemporary solanocrinids. Possibly this species represents a type standing near the ancestral form of one or, perhaps, both the series mentioned above—Thalassometrida and Notocrinida. The slender habitus, the long undiffer- entiated rather scanty cirri, the numerous syzygies, and the prismatic pinnules with strongly developed side plates and covering plates all indicate that Pterocoma pennata should be placed very near the younger comatulid types and that it has nothing in common with the solanocrinids. Although masses of this species, otherwise almost perfectly preserved, are known, it is remarkable how very seldom the centrodorsal has been noticed. This is presumably due to the fact that the centrodorsal very easily becomes detached, and possibly also that at the animal’s death it was left remaining upon a still existent stem. In the only specimen among 81 examined in which he found a centrodorsal preserved Walther observed and figured a large circular opening in the center of the dorsal side. Gislén said he had observed a similar pore on the dorsal side of a centrodorsal in a specimen he had examined. Gislén said that a circumstance speaking in favor of the Thalassometrida and Notocrinida having reached the eleutherozoic mode of life at different times is the very different point of time at which their pentacrinoids become detached from the stem—in the former very early, in the latter type at a very late stage. The Thalassometrida and the Notocrinida, according to Gislén, are the groups that in the recent seas have retained most of the features belonging to the stalked forms. These are especially numerous in the Thalassometrida—long cirri, some- what rhombic in cross section; centrodorsal with the cirri in columns, sometimes in an almost simple radial column; basals relatively well developed; the free dorsal surface of the radials generally broad; the arm bases closely set and ‘‘wall-sided,”’ the arms ending abruptly with an abortive portion bearing rudimentary pinnules; the pinnules prismatic to triangular in cross section; the side plates and covering plates strongly developed; and the disk often closely studded with calcareous granules or plates. The evolution of the palaeantedonids.—Gislén said that Semiometra is a final type in the notocrinid series and constitutes a form transitional to the palaeantedonids, of which the genus Palaeantedon is typical. Gislén did not believe it possible to doubt that this genus represents a type corresponding to the recent antedonins, or in other words to group @ among the Macrophreata. This is evidenced by the centrodorsal being closely set with numerous unsculptured cirrus sockets in alternating rows, by the large centrodorsal cavity, by the lack of any radial pores or pits on the ventral face of the centrodorsal, by the obliterated basals, by the large muscular fossae on the articular faces of the radials, by the inward sloping of the radial articular faces, by the slender oblique brachials, ete. The genus Hertha may be considered as a predecessor of the palaeantedonid type, specialized and characterized by the develop- ment of high narrow muscular fossae and a small radial cavity. The recent genus Coccometra typologically corresponds to the genus Hertha. The genus Discometra also comes very near the macrophreate type, although this type may prove to be a heterogeneous one. Among the recent Macrophreata belonging to group a the Ante- 56 BULLETIN 82, UNITED STATES NATIONAL MUSEUM doninae exhibit a number of more primitive characteristics, such as a relatively small centrodorsal cavity, rather inconspicuous muscular fossae, and weak synarthrial backward projections. It is interesting to note that we sometimes get here radial prolongations of the dorsal coelome that reach down to the ventral side of the centrodorsal and there form small and shallow pits. It seemed very plausible to Gislén to assume that the second group of Macro- phreata proceeded from a type nearly allied to the Thalassometridae, as well as to the fossil genera Amphorometra and Conometra. If we imagine a reduction of the side plates and covering plates in the thalassometrid type, we get a form extremely similar to the more primitive types within group b—the subfamily Zenometrinae. The size of the centrodorsal cavity for which this group 6 among more advanced forms is remarkable shows, as the synarthrial backward projection also does, a more moderate development in certain Zenometrinae. If it were a question of placing the genus Conometra in one of the recent groups it would be difficult to decide whether it should be referred to the Thalassometridae or to the Zenometrinae. In the same way it would be almost impossible, in certain cases, to distinguish between recent Thalassometrida and Zenometrinae if the centrodorsal and radial pentagon only were known. On the other hand, the genus Amphorometra shows, like Placometra, distinct points of con- tact with thalassometrid forms—with the Thalassometridae and Charitometridae. It seems hardly possible, according to Gislén, that the genus Jaekelometra can have been the ancestral form of the Atelecrinidae. There are certainly similiarites in the size of the basal ring, but the relative smoothness of the cirrus facets in Jaekelometra, the small size of the cavity in the centrodorsal, and the steep inward inclination of the radial articular faces which are provided with relatively insignificant muscular fossae differ radically from the conditions in the Atelecrinidae, in which the cirrus sockets have a strong crest and the centrodorsal cavity and radial muscular fossae are extra- ordinarily large. It appears, too, as if in Jaekelometra we are able to find an indication of the development of a rosette through thin centripetal continuations of the basals. No such formation seems to be present in Atelecrinus. Gislén therefore concluded that Jaekelometra is a type standing nearer certain primitive Thalassometrida. Conclusions on the evolution of the comatulids.—Gislén said that his investi- gation of the mutual relations between the comatulids led him to assume the possi- bility of three, or perhaps four, different lines of stalked crinoids of the pentacrinid type having reached the eleutherozoic mode of life and become comatulids. The first and oldest branch of the comatulids is represented by Palaeocomaster, to which the recent comasterids correspond. At the present time they still show their unique nature by the strong tendency toward reduction of the cirri, by the frequent occurrence of exocyclic forms, by the absence of sacculi, which are replaced by the pear-shaped organs, by the combs, and by the dorsal hooks. In contrast to all the other comatulids they have never acquired the capacity for swimming [not quite correct; see Part 2, page 602, fourth paragraph; page 606, second and last paragraphs], but still continue to employ a creeping mode of locomotion. The second somewhat younger branch consists of the solanocrinids. We find among these in the Upper Jurassic forms having strongly biserial arms and very clumsily built. Gislén assumed that these were very probably creeping forms. Pre- sumably this type has died out, while less specialized solanocrinids with monoserial A MONOGRAPH OF THE EXISTING CRINOIDS 57 arms have given rise to the recent Mariametrida. This group is more specialized than the solanocrinids in the appearance and arrangement of the cirrus sockets, the far advanced reduction of the basals, the development of the calcareous plug, the numerous arms, and the many syzygial septa. The relatively stout arms, the small centrodorsal cavity, and probably the relatively few syzygies may be considered as primitive characters that have been retained. In addition to these, the Zygometridae, Mariametridae, and certain Himerometridae show a number of primitive features. The third group, which possibly has a double origin, is represented by the Noto- crinidae and Conometridae. It is divided into two distinct series of which one, the Notocrinidae, leads toward the Palaeantedonidae and group a of the Macrophreata, while the other leads to the Thalassometrida and group 6 of the Macrophreata. A form that possibly may be interpreted as a prototype of one or possibly of both of these series is found in Pterocoma pennata. The Thalassometrida have preserved a number of pentacrinid characters in the long cirri, somewhat prismatic in cross section and arranged in radial columns, in the prismatic or triangular brachials and pinnule segments, in the relatively well developed basals, in the broad free dorsal surface of the radials, in the strong development of the side plates and covering plates, and in the granular covering of the disk. The nearer one comes to the present time the more difficult it may become in critical cases to distinguish the types of these different groups from each other. If one excepts the comasterids, which have followed their own special course of develop- ment, and certain Mariametrida in which the reduction of the radial muscular fossae is strongly pronounced, the comatulid groups tend toward an ideal type of comatulid. For this reason it becomes extremely difficult to keep apart the converging forms that by different paths have approached this ideal type. Gislén considered the subfamily Perometrinae as representing such a series, lead- ing from colobometrid forms toward the antedonid type. He said it might have been desirable to divide the Macrophreata into more groups that could later have been placed at the apices of the different series of the older comatulid types. He did not choose to do this, however, as it seemed to him to be systematically impracticable and unfeasible. So he contented himself with depicting instead as nearly as possible, guided by the known facts, the different paths of development through which the distinctly emphasized polyphyletic character of the macrophreate group becomes explicable. NOTES Dr. Gislén regards the approach toward an idealized—perhaps more properly speaking generalized—type of comatulid that is evident in the several groups as con- vergence from originally distinct lines of different ultimate origin. To the author it seems more logical to reverse the picture and to assume a gradual divergence of the several groups from a number of types originally very similar and of an ultimate common origin. The comatulids have too many fundamental features in common that are not shared with other groups, even with the pentacrinites, to be considered polyphyletic. From his generalizations Gislén omits the Comasteridae, which he said have followed their own particular course of development. He did not mention the genus 208244—40——5 58 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Comatulides (see Part 3, page 400), which is a very generalized type with little resem- blance to other comasterids, and Comatonia (see Part 3, page 288), still more general- ized, with a rounded conical centrodorsal, cirri in 3-4 irregular rows, strong synarthrial backward projections, and sacculi, he relegated to the Macrophreata, in spite of its perfectly typical and enormous comasterid combs. It is quite true that some of the small species of the Colobometridae present a marked superficial resemblance to the species included in the subfamily Perometrinae. Nevertheless this resemblance is purely superficial, and the members of the two groups are always easily distinguished. The cirri are always thoroughly distinctive and show no tendency to intergrade. This confusion seems to have arisen in part from his disposition of the genus Clarkometra, which he assigned to the Perometrinae, where- as in reality it is a member of the Colobometridae, with perfectly typical colobometrid cirri. The small species in the family Colobometridae bear approximately the same relation to the larger that the species of the genera Comatonia, Microcomatula, and Comatulides do to the larger comasterids. Dr. Gislén’s work on the fossil comatulids is of outstanding importance, and for the first time these puzzling fossils have been brought into a semblance of order. With all due respect for his excellent work, and for him personally, the author believes that far more detailed knowledge of the recent forms and their comparative structure, growth changes, life histories, etc., is necessary before their relationships with the fossils as a whole can be satisfactorily elucidated. FOOD AND FEEDING HABITS Gislén studied in great detail the feeding habits and the food of several recent crinoids. At the Kristineberg Zoological Station he studied Antedon petasus brought in from the outer parts of the Gullmar Fjord, where the species is found on a shell and gravel bottom at a depth of 30-40 meters. At the biological station of Haegdalen, Trondhjem, he studied Hathrometra tenella var. sarsii and Rhizocrinus lofotensis taken between Leksviken and Tautra on muddy bottom at a depth of 250 meters. He found Antedon petasus, which he had had opportunities for studying in aquaria for a long time, at the beginning rather difficult to keep alive. It turned out that the temperature of the tank water supplying the aquaria with water was too high (16- 18° C.). It was not practicable to use water brought up from the bottom directly, for as this cold water warmed up gas bubbles adhered to the pinnules of the crinoids, which floated to the surface and soon died. This was prevented by setting the ves- sel containing the animals on ice. Afterward by setting the aquarium, filled with water of very high salinity from the outer parts of the archipelago, in a tub filled with ice water he was able to keep Antedon alive and vigorous for weeks. The water did not need changing more than once every, or every other, week. The salinity of the sea water varied between 26 and 28 per thousand. The necessary supply of oxygen was obtained by means of an aerating apparatus of the usual type used at the station. The water was kept at a temperature of 10-12° C. — occasionally it was a couple of degrees higher or lower. It was soon found that it was not advisable to let the temperature rise higher than 14° C. With a higher salinity, however, the animals seemed able to stand a somewhat higher temperature. Thus they got on A MONOGRAPH OF THE EXISTING CRINOIDS 59 fairly well at a temperature of 15° C. when the sea water had a salinity of 30 per thousand. If the animals are to do well Gislén says it is absolutely necessary to sup- ply the bottom of the aquarium with small pebbles or shells to which they can cling with their cirri, otherwise they fall over and soon die. In the crinoids, as in other echinoderms, there occur on the surface of the body ciliary currents that serve the purpose of keeping the animal clean. These are most pronounced in the interradial and interbrachial areas of the disk. On the median parts of the interradial areas run centripetal (adoral) currents. These are checked at the oral angle by the upright adambulacral fold on the margin of the ambulacral furrow. Circular currents, therefore, running vertically, often arise in the oral angles. Along the sides of this median centripetally directed current the grains of carmine are carried radipetally, and in the area nearest the ambulacral groove centrifugally. As a result of this all the grains in the interradial area are gradually carried away along the margin of the ambulacral furrow. Where the ossicles in this upright margin are rather low, or where there is a gap between them, the grains are sucked down into the centripetal current of the ambulacral groove, and in this way most of the parti- cles, even those falling upon the interradial areas, at last reach the mouth. Some of them, however, are carried out to the sides of the body and fall off the animal. The interbrachial parts of the perisome have ciliary currents going in an oropetal or radipetal direction. The oral pinnules are ciliated slightly toward their tips, and the grains of carmine carried thither thus sink down on to the surface of the disk and usually sooner or later reach the mouth. The anal cone is ciliated only at the base, where weak currents may be observed running upward. The excrement con- sists of the undigested particles from the intestine held together by a jellylike substance and formed into large yellow mucus balls. These balls are too large to be transported by the ciliary currents. They are removed from the surface of the disk, where they usually fall, by the movements of the animal when moving to another place, or by stronger movements of the water caused by the animal flexing its arms, etc. The cirri and the dorsal side of the animal are not ciliated but are kept clean by the move- ments of the animal. Gislén said that the currents described above refer to Antedon petasus, but those of Hathrometra tenella var. sarsii are similar except that the marginal currents directed centrifugally may be followed further out on the sides of the arms. In Rhizocrinus the interradial currents are centripetal only. When not irritated a hungry Antedon sits immovable with the arms outstretched and slightly curved upward. The pinnules extend out from the arm almost at right angles, and the tentacles, arranged in groups of three on’ the pinnules, are stiffly extended. If a plankton sample or a few grains of carmine with some crab liver be ptt into the water of the aquarium the arms and pinnules at once become very active. The tentacles all beat rapidly inward toward the ambulacral furrow. This, the margins of which usually lie tightly pressed together, opens as soon as any grains fall upon it, with a wavy movement proceeding toward or away from the mouth. The mouth, previously only a narrow slit, opens to its widest extent and becomes circular. After some time the tentacles straighten themselves out again, but now and then when tiny grains fasten on the papillae the tentacles are whipped with rapid nervous move- 60 BULLETIN 82, UNITED STATES NATIONAL MUSEUM ments toward the ambulacral groove, and then again straighten out as quick as light- ning. Some of the groups opposite each other usually whip inward toward the fur- row at the same time, but no flexing inward of the pinnules proceeding along the arms can be observed. Small particles are fastened to the tentacles by the secretion of unicellular mucous glands opening on the papillae, and with rapid motions of the tentacles are cast into the ambulacral groove where the ciliary current catches them and carries them to the mouth. Sensory hairs are also found on the tentacles, and it has been supposed that these serve as sensory organs. The function of the tentacles, according to Gislén, is prob- ably in reality manifold. First, they are certainly of considerable importance in respiration. Second, the mucus secretion, as pointed out above, serves to fasten the falling grains of plankton or detritus upon the tentacles temporarily. Third, the sensory cilia serve presumably in some way as an organ for taste and feeling. When Gislén dropped picric acid upon them the arms were flexed and relaxed violent- ly; when he added a few drops of sublimate the arms were also rapidly bent and twisted, the pinnules were stretched out, and the animal made swimming motions as if in an endeavor to escape. On a third occasion quinine sulphate with"a" few grains of car- mine was added. The arms were waved actively to and fro, the pinnules were flexed in to the sides of the arms or rubbed against each other, or laid over the place where the drops had fallen, the arms were turned upside down, with the evident intention of getting rid of the unpleasant matter. When old putrefying crab liver was added to the fluid containing the grains of carmine the pinnules were flexed in toward the sides of the arms. This seemed to Gislén to indicate that a sense of taste is present which presumably may be considered as localized in the sensory hairs of the papillae of the tentacles. Gislén was unable to determine definitely whether, as maintained by Reichen- sperger, the secretion of the papillae of the tentacles is poisonous. He allowed living pinnules to lie in contact with freshly caught plankton (Ephyrae, peridineans, diatoms, and Plutei) but was unable to ascertain that any rapid unconsciousness or toxic action took place. A small Cladonema seemed to make motions of escaping when touched by a tentacle, but usually the tentacle bent away from the irritating object. In the intestine of Heliometra glacialis Gislén has found quantities of copepods. The copepods have a tolerably strong power of motion of their own, and he said it is very probable that in this case the secretion of the papillae of the tentacles in some way paralyzes the prey as otherwise it is difficult to imagine how the relatively weak ciliary currents can transport the captured prey to the mouth. The most proximal of the tentacles in each of the groups of three is the shortest and is directed straight upward, the middle one is longer and extends obliquely out- ward, and the most distal is the longest and is directed straight outward. On the arms this differentiation is indistinct, or is not marked at all. In the proximal parts of the arms and on the disk the tentacles are unbranched. The tentacles are often more or less contracted, but at the maximum extension the longest tentacle reaches from 0.5 to 0.7 mm. As the pinnules are placed at a distance of from 1.0 to 1.4 mm. from each other, with the maximum extension of the tentacles the region between the pinnules can be swept completely clean A MONOGRAPH OF THE EXISTING CRINOIDS 61 The ciliary epithelium, characteristic of the ambulacral groove, begins at the base of every group of tentacles. The ambulacral furrow itself is slightly concave and ciliates strongly toward the mouth. The mucus is secreted in the ambulacral groove by the mucus glands, and by it the grains are fastened together into larger lumps, which are transported to the mouth, the edges of which are often pressed together like a pair of lips during movements similar to swallowing. The centripetal ciliary currents were also observed in Hathrometra and in Rhizocrinus. In a healthy animal that does not receive any great amount of nourishment at one time the mouth is shaped like a slit. This slit in Antedon petasus is always oriented from the anterior radius toward the posterior interradius, or in the anterior half twisted a little to the left of the same line. Gislén observed a slit-shaped mouth with the same orientation in a specimen of Thawmatometra comaster preserved in spirit. In Heliometra glacialis the slit has the same orientation as in Antedon petasus but is directed a little more obliquely, from the right anterior interradius toward the left posterior radius, and the same condition is found in specimens of Poliometra proliza, Asterometra anthus, and Oligometrides adeonae. The mouth is twisted still more in the same direction in Tropiometra afra macrodiscus, in which it looks as though it might become almost transverse in position, as in the Comasteridae. Most of the crinoids preserved in alcohol that Gislén examined have a rounded mouth opening, like sickly or dying individuals of Antedon petasus. All kinds of small particles eddying up from the bottom owing to the movements of the animal or by currents in the water—mostly the latter—are carried to the mouth; in other words detritus and benthonic organisms, as well as some part of the plankton that rains down. Gislén’s experience seemed to show that the benthonic nourishment is the most important. The following are analyses of samples of excrement from living animals examined: Antedon petasus.—All the samples were brought home from Smedjan, Kristine- berg, Sweden, where the animals were dredged from a depth of 30-40 meters. 1. A rough sample. The recently dredged animals were put into a vessel of sea water where they were allowed to stay for some hours, and the abundant excrement evacuated was collected. This contained detritus, small grains of sand, threads of algae, diatoms, some peridineans, crustacean larvae (0.3-0.5 mm. long), small cope- pods (0.09 by 0.18 mm.), macerated pieces of plants (one piece 0.40 by 0.60 mm.), and a half-digested myzostomid. 2. A lump of excrement taken directly from the anal opening contained detritus, diatoms (Coscinodiscus, Navicula, Nitschia, etc.), copepods, an ostracod, a veliger larva, a Ceratium, and a few living infusoria. 3. Some food sucked from the mouth by means of a pipette included detritus, diatoms, copepods (empty casings), some ostracods, one veliger larva, nauplius larvae, empty podiae of small crustaceans, a peridinean, and a few sponge spicules. The following are analyses of samples of excrement from preserved specimens: Metacrinus rotundus, Sagami Bay, Japan, 180 meters. The excrement was collected from the bottom of the vessel in which the animal had been laid. It con- sisted of detritus, diatoms, tissues of more highly organized plants, solitary crusta- cean podiae, crustaceans, and Foraminifera. 62 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Comantheria grandicalyx, Bonin Islands, 54-72 meters. The sample was obtained in the same way as the preceding. It consisted of detritus, threads of algae, small grains of sand, diatoms, sponge spicules, bits of leaves (the largest 4.5 by 1.25 mm.), bryozoans, solitary holothurian anchors, a syllid (2.1 by 2 mm.) and a syllid larva, and a piece of a hydroid colony (length 5 mm., breadth of branches 0.15 mm.). Comaster delicata grandis, Bonin Islands, 72 meters. ‘Thesample was taken directly from the intestine. It contained algal threads, numerous radiolarians, tintinnoids, crustacean podiae, solitary diatoms, grains of sand, Foraminifera, bits of bryozoans, and sponge spicules. In another sample from another specimen of the same species taken at the Bonin Islands in 144 meters the diatoms (Planktoniella, Thalassiothriz, and Nitschia) were more abundant; otherwise this sample was like the other. Zygometra microdiscus, Broome, Western Australia, from the shore. The sample was taken from the anal opening. It contained grains of sand, diatoms, small cope- pods, ostracods, Foraminifera, and veliger larvae. Tropiometra afra macrodiscus, Misaki, Sagami Bay, Japan, 2-3 meters. Two specimens were dissected. The intestine was almost empty. There were incon- siderable remains of grains of sand, crustacean podiae, and diatoms. In the gullet were large parasitic myzostomes. Asterometra anthus, Bonin Islands, 180 meters. The sample was taken from the intestine and contained detritus, radiolarians, solitary bits of crustacean podiae, Foraminifera, and one Ceratium. Heliometra glacialis, west of Spitzbergen (lat. 79°30’ N., long. 10°30’ E.), 100 meters. The sample was from the intestine and contained numerous copepods, solitary ostracods, polynoid bristles, and one tintinnoid. Dr. Astrid Cleve-Euler identified the diatoms in some of the samples given above. A sample of the excrement of Antedon petasus contained Actinocyclus crassus, Actino- ptychus undulatus, Biddulphia aurita, Campylodiscus thuretii, Cerataulus smithii, Coscinodiscus excentricus, C. oculus iridis, C. sp. (with large meshes), Cocconeis scu- tellum, Diploneis nitescens, Grammatophora marina, Hyalodiscus scoticus, H. stelliger, Navicula cyprinus, Paralia sulcata, Nitschia longa, N. seriata, Rhabdonema arcuatum, R. minutum, Rhizosolenia setigera, Trachyneis aspera, and the following Silicoflagellata: Distephanes speculum and Ebria tripartita. The excrement of Comantheria grandicalyx contained Achnanthes longipes, Am- phora sp., Actinoptychus undulatus, Coscinodiscus sp. (with large meshes), Diploneis didyma, D. litoralis, Epithemis (Rhopalodia) gibba, Grunoviella marina, Grammatophora marina, Melosira nummuloides, Orthosira splendida, Paralia sulcata, Trachyneis aspera; there were many sponge spicules and star-shaped hairs. In the excrement of Comaster delicata grandis there were few diatoms, Asterolampa marylandica var. major, Nitschia marina, Navicula lyra, all solitary, the silicoflagellate Dictyocha fibula, and many radiolarians. The excrement of Metacrinus rotundus contained Achnanthes mammalis, Actinocy- clus ehrenbergi, Actinoptychus undulatus, Amphora costata, A. macilenta, Climacospnenia elongata, Cocconeis reticulata, C. dirupta, C. finmarchica, C. pellucida, Coscinodiscus radiatus, C. excentricus, O. curvatulus, C. oculus iridis, Cyclotella striata, Diploneis sp., Grammatophora oceanica var. macilenta, G. angulosa var., Isthmia nervosa fragments, A MONOGRAPH OF THE EXISTING CRINOIDS 63 Navicula libellus, Nitschia panduriformis var. minor, Synedria affinis, Trachyneis aspera, Thalassiosira gelatinosa, and Thalassiothrix nitschioides. Mrs. Cleve-Euler’s opinion on the diatoms identified was that the major part of the species represented are of benthonic origin. The species of Cocconeis are epiphytes on algae and similar material. Some of the others, as Actinocyclus, Actinoptychus, Coscin- odiscus, Cyclotella, Hyalodiscus stelliger, and Paralia, are tychopelagic (=semipelagic), mostly heavy types that keep to deeper water. Thalassiothriz is pelagic. Mrs. Cleve-Euler was unable to find any pronouncedly pelagic genus, such as Chaetoceras, in either of the first two samples. She said of the last two, ‘‘Possibly setae of true pelagic plankton types, such as Chaetoceras, are present, but they could not be de- termined,” and added, “It does not seem probable that they [true pelagic forms] are devoured to any extent worth mentioning by these animals.” Basing his opinion on the evidence afforded by the samples of excrement, and in the light of Mrs. Cleve-Euler’s statements, Gislén considered it clearly manifest that the crinoids live upon a mixed diet of detritus, benthonic microorganisms, and plank- ton, but that the detritus plays a very important part, at least for some of them. FORMATION OF THE AMBULACRAL GROOVE AND FEEDING IN THE COMASTERIDS Gislén noted that the ambulacral groove is very variable in its formation and offers a great deal that is interesting if one follows its varying extension along the arms and pinnules. The ambulacral groove originally extended to all the arms and appeared on all the pinnules. This primitive condition is still found in the stalked crinoids, for instance in Metacrinus and Rhizocrinus. In the comatulids, however, a varying num- ber of proximal pinnules have developed into tactile organs and have been trans- formed into the so-called oral pinnules, which have no ambulacral groove or gonad. In most of the Bathymetrinae there is only one pair of these on each arm, Antedon petasus has two pairs, and A. mediterranea three. Simultaneously with the change of function their appearance also changes, and they generally become long, flagellate, and exceedingly flexible. They are very inconsiderably transformed in the Himerome- tridae, Stephanometridae, Charitometridae, and Thalassometridae, while a number of the Mariametridae, Calometridae, and Colobometridae, and most of the Macroph- reata and Comasteridae show very advanced specialization. Gislén said another reason why the extension of the ambulacral grooves differs from the usual type on certain pinnules is found in the development of the gonads. In most of the comatulids the first gonad appears on the first postoral pinnule, and they appear on all the pinnules following until far out on the arms. In the distal pinnules the gonads are not so strongly developed, and they gradually become partly, or com- pletely, obliterated. In certain types, however, the gonads are restricted to certain pinnules, the genital pinnules, and in correlation with their exclusive function as egg or sperm producers, these have their ambulacral groove and their hyrocoele more or less strongly reduced and become stout and clumsy. The distal pinnules, on-the other hand, begin to play the role of respiratory organs and to become specialized for food collecting. They often become very much longer and more slender, and have their tentacles and ambulacral furrow very strongly developed. Thus at least certain species of the family Tropiometridae have no ambulacral groove on the first 8-12 pairs 64 BULLETIN 82, UNITED STATES NATIONAL MUSEUM of pinnules, which are developed as genital pinnules. Similarly two representatives of the family Charitometridae, Poecilometra acoela and Pachylometra angusticalyz, sometimes lack the ambulacral groove on as many as 20 pairs of proximal pinnules, and Gislén found that in Diodontometra the first 9 pairs of pinnules, which have become transformed into broad joints protecting the gonads, have no ambulacral groove. In all these cases the reduction of the ambulacral groove is the same on all the arms. The comasterids, which show far reaching reduction in the apparatus for obtaining food, have gone in another direction. Gislén said that these animals are in a state of transition—from a radiul type they are changing to a pronouncedly bilateral type. The mouth moves forward in the anterior radius, or toward the right anterior interradius, and often takes a marginal position on the disk. The anal funnel assumes the central position of the mouth, in consequence of a lengthening of the intestine. The anterior arms lengthen and their pinnules become long and slender, like the distal pinnules of Tropiometra. The posterior (aboral) arms are shortened and have limited longitudinal growth. They become thicker, stouter, and clumsier, according to Gislén perhaps, though not very likely, on account of the abundant formation of sex products in them. A cooperating factor in the enlargement of the joints is certainly to be found in the strengthened muscular and ligamentary connections between the brachials. As a result of this division of labor between the anterior and posterior arms the ambulacral furrows lose their radially symmetrical arrangement and are often present on the anterior radii only. It is to be observed that the ambulacral furrow - really disappears, and is not found remaining “‘in a closed groove.” Great variation prevails even in the same species. In a number of 10-armed specimens of Comatula pectinata and its variety purpurea from Cape Jaubert, Western Australia, there is no ambulacral furrow on from none to four of the posterior arms; ina Comanthus parvicirra from Sagami Bay, Japan, out of 33 arms 13 in the posterior radii had no ambulacral groove, while 4 neighboring arms had it developed distally of the fifteenth or twentieth pair of pinnules. _ Gislén remarked that the reduction of the ambulacral grooves goes so far in the Comasteridae that one is tempted to ask whether the apparatus for gathering food can really be sufficiently large to collect nourishment enough for these relatively stout and clumsy animals. In order to form an opinion of the relations existing between the apparatus for collecting food and the size of the animal Gislén estimated the total length of the ambulacral groove, measured the volume of the animal, and then by comparing the figures determined their relation to each other. The total length of the ambulacral grooves was obtained by measuring the length of the arms, multiplying by their number, and to the figure thus obtained add- ing the product of the number of pinnules per arm, the average length of the pinnules, and the number of arms. The final figure gives the approximate length of the ambulacral grooves. Among the crinoids measured by Gislén Rhizocrinus lofotensis, which has only four or five pairs of pinnules per arm, gave the lowest total figure. The length of the ambulacral grooves in a 6-armed specimen was only 0.14—0.22 meters. Among the comatulids measured a specimen of Comissia ignota minuta with 10 arms 20 mm. long showed the shortest ambulacral grooves, 1.30 meters. It is about the same length A MONOGRAPH OF THE EXISTING CRINOIDS 65 in a specimen of Compsometra parviflora with 10 arms 25 mm. long, in which the grooves measure 1.37 meters. It is only inconsiderably longer in a specimen of Pentametrocrinus diomedeae with 5 arms 55 mm. long (1.85 meters) and in a specimen of Thaumatometra comaster with 10 arms 45 mm. long (2.98 meters). Oligometra chinensis with 10 arms 55 mm. long has the ambulacral furrows 7.15 meters in length, while a large Antedon petasus with the 10 arms 110 mm. long has them 16.28 meters long. A specimen of Stephanometra spicata with 26 arms 130 mm. long has the ambu- lacral grooves 38.48 meters long; a specimen of Liparometra grandis with 26 arms 120 mm. long has them 44.72 meters long; a specimen of Heliometra glacialis with the 10 arms 200 mm. long has them 54.5 meters long; and in a specimen of the multibrachiate Himerometra magnipinna with 49 arms 90 mm. long they reach 61.23 meters in length. A large individual of Metacrinus rotundus with 56 arms 210 mm. in length from the radials has the ambulacral furrows 71.68 meters long. The highest numbers, how- ever, are found in the Comasteridae, a specimen of Comantheria grandicalyx with 68 arms 125 mm. long having the ambulacral furrows 102.68 meters in length. Gislén mentioned a few facts that it seemed to him might influence the relative length of the ambulacral furrow. At great depths the food supply is more scanty than higher up, partly because the bottom life is not so abundant, and partly because the plankton does not rain down so plentifully, whereas in the littoral zone the food is very abundant. Deep-sea forms ought therefore to have a relatively long ambu- lacral groove in proportion to their size, while shallow water forms ought to be able to manage with a shorter one. The degree of calcification probably plays some part also. A strongly calcified animal has relatively less organic substance than one which is inconsiderably calcified. The former has less substance to support, and ought to be able to obtain the necessary amount of nourishment with a proportionately shorter ambulacral furrow. Two objects of similar shape but differing in size stand in relation to each other as the square of the surface and the cube of the volume. If we take a 10-armed comatulid of a certain size and compare it with a similar but much smaller one the lesser one ought to have a relatively greater surface in proportion to the volume; or if we compare it with a larger one, the larger one in proportion to the volume ought to have a smaller surface. The figure represented by the total length of the ambulacral groove is a function of the surface supplying nourishment and varies with it. The ratio between the ambulacral grooves and the volume in very small comatulids gives an abnormally high figure, and in very large comatulids an abnormally low one. The following figures show the relationship between the length of the ambulacral furrows and the bulk in various comatulids. Comissia parvula, Sagami Bay, Japan, 720 meters (10 arms 35 mm. long): Length of ambulacral furrows, 1.95 meters; weight 0.89 grams; volume 0.74 cc.; specific weight 1.20; length of ambulacral furrows divided by volume 2.64. Comatula pectinata, Cape Jaubert, Western Australia, 20 meters (10 arms, 6 grooved, 90 mm. long): Length of ambulacral furrows 4.37 meters; weight 4.83 grams; volume 3.46 cc.; specific weight 1.39; length of ambulacral furrows divided by volume 1.26. 66 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Comantheria grandicalyx, Bonin Islands, 54-72 meters (68 arms 125 mm. long): Length of ambulacral furrows 102.68 meters; weight 89.6 grams; volume 73.6 cc.; specific weight 1.22; length of ambulacral furrows divided by volume 1.39. Comaster delicata grandis, Bonin Islands, 72 meters (46 arms 115 mm. long): Length of ambulacral furrows 58.92 meters; weight 39.1 grams; volume 30.0 cc.; specific weight 1.30; length of ambulacral furrows divided by volume 1.96. Comanthus parvicirra, Misaki, Japan, 0-3 meters (33 arms, 20 grooved, 50 mm. long): Length of ambulacral furrows 7.16 meters; weight 6.65 grams; volume 5.40 cc. ; specific weight 1.23; length of ambulacral furrows divided by volume 1.33. Catoptometra magnifica minor, Kiu Shiu, Japan, 900 meters (27 arms 95 mm. long): Length of ambulacral furrows 34.50 meters; weight 6.70 grams; volume 5.42 cc.; specific weight 1.24; length of ambulacral furrows divided by volume 6.37. Eudiocrinus loveni, Bonin Islands, 144 meters (5 arms 43 mm. long): Length of ambulacral furrows 2.47 meters; weight 0.65 grams; volume 0.47 cc.; specific weight 1.38; length of ambulacral furrows divided by volume 5.26. Heterometra crenulata, Cape Jaubert, Western Australia, 21 meters (30 arms 75 mm. long): Length of ambulacral furrows 43.05 meters; weight 11.55 grams; volume 8.28 ce.; specific weight 1.39; length of ambulacral furrows divided by volume 5.20. Lamprometra palmata, Trincomalee, Ceylon, 7 meters (43 arms 50 mm. long): Length of ambulacral furrows 23.44 meters; weight 7.15 grams; volume 5.75 cc.; specific weight 1.24; length of ambulacral furrows divided by volume 4.08. Cyllometra pulchella (=manca), Kiu Shiu, Japan, about 200 meters (14 arms 40 mm. long): Length of ambulacral furrows 6.02 meters; weight 0.72 grams; volume 0.60 ec.; specific weight 1.20; length of ambulacral furrows divided by volume 10.02. Tropiometra afra macrodiscus, Misaki, Japan, 0-3 meters (10 arms 255 mm. long): Length of ambulacral furrows 47.69 meters; weight 39.8 grams; volume 34.4 ce.; specific weight 1.16; length of ambulacral furrows divided by volume 1.39. Asterometra anthus, Bonin Islands, 207 meters (10 arms 70 mm. long): Length of ambulacral furrows 10.60 meters; weight 3.12 grams; volume 2.34 cc.; specific weight 1.33; length of ambulacral furrows divided by volume 4.52. Pectinometra flavopurpurea, Sagami Bay, Japan, 720 meters (20 arms 80 mm. long): Length of ambulacral furrows 17.84 meters; weight 4.60 grams; volume 3.55 ce.; specific weight 1.30; length of ambulacral furrows divided by volume 5.02. Perissometra aranea, Kiu Shiu, Japan, 198 meters (11 arms 100 mm. long): Length of ambulacral furrows 7.04 meters; weight 1.93 grams; volume 1.46 cc.; specific weight 1.32; length of ambulacral furrows divided by volume 4.83. Antedon petasus, Kristineberg, Sweden, 30-40 meters (10 arms 110 mm. long): Length of ambulacral furrows 16.28 meters; weight 2.63 grams; volume 2.29 cc.; specific weight 1.15; length of ambulacral furrows divided by volume 7.10. Heliometra glacialis, East Greenland, 80-100 meters (10 arms 200 mm. long): Length of ambulacral furrows 54.5 meters; weight 50.0 grams; volume 45.0 cc.; specific weight 1.11; length of ambulacral furrows divided by volume 1.21. Gislén said that a glance at these figures shows at once that in the ratio between the length of the ambulacral groove and the volume the average is much smaller in the Comasteridae than in the other comatulids—about 1.7 as against 5.0. A MONOGRAPH OF THE EXISTING CRINOIDS 67 Comissia parvula shows for a comasterid an abnormally high figure, partly because it is a small species, partly because it is endocyclic, and partly because it is a deep- sea species. For the last-mentioned reason the figure for Catoptometra magnifica minor is also unusually high. Lamprometra palmata is a littoral type and therefore shows a low figure. Species having a high specific weight, that is, strongly calcified species, usually show low figures also. In Pectinometra flavopurpurea the decrease is counterbalanced by the great depth at which the animal is found. On the other hand, the slightly calcified Antedon petasus and Cyllometra pulchella—the latter also very small—show very high figures. Gislén said it is more difficult to explain the deviation in Tropiometra afra macro- discus and Heliometra glacialis. Both are, however, exceptionally large types, and the former is at the same time a littoral species. The body of Heliometra glacialis is not very strongly calcified, which should perhaps counterbalance its gigantic size, and so also should the fact that it is not a littoral form. An analysis of the intestinal contents of Heliometra glacialis showed a great abundance of copepods, these being present almost exclusively—very nourishing food. This comatulid lives outside the great glaciers of Spitsbergen and Greenland, where an abundance of plankton is found. Gislén suggests that possibly this may explain to some degree the low figure. The appearance of “rudimentary combs” on the proximal pinnules must, however, be noticed. Gislén noted that he had chosen for hisinvestigations the most divergent typesfrom the most diverse localities he was able to obtain, and he said that therefore one should not feel too much surprised if a few divergencies from the general rule appear. It is possible that in the two cases of great divergency described some unknown facts may play a part. Taken as a whole, however, it seemed to Gislén that the difference between the low figure of the Comasteridae and the high figure for the rest of the comatulids is very remarkable. It is all the more noticeable that the figures given for the Comasteridae are low throughout, as three of the five comasterids examined have ambulacral furrows on all the arms and most of the pinnules. FORMATION AND CONTENTS OF THE INTESTINE IN THE COMASTERIDAE Gislén noted that although in the Comasteridae the ambulacral grooves are abnormally short in proportion to the size of the animals, the great length of the intestine suggests that the food is richer and at the same time more difficult to digest. He said that the author’s explanation of the elongated digestive tube in the Coma- steridae (Part 1, p. 343, fourth paragraph) is only in part correct. Certainly the Comasteridae in general are shallow-water types, particularly the exocyclic ones; but so also are certain families with central mouths, as the Himerometridae, Maria- metridae, Stephanometridae, and Tropiometridae. That the Comasteridae are confined to muddy bottoms is not correct, as may be seen by an examination of the localities given in the Siboga report. Comasterids were collected at 65 localities at which the character of the bottom was recorded; of these localities 54 also yielded exocyclic comatulids. In 41 of these localities they were taken from a bottom consisting of stone, sand, “coral,” and Lithothamnion; mud was found in only three cases. Gislén’s examination of the intestinal contents in the Comasteridae showed that the vegetable 68 BULLETIN 82, UNITED STATES NATIONAL MUSEUM food is certainly very plentiful, consisting for the most part of Chlorophyceae, but that inorganic matter is inconsiderable or wholly absent. MORPHOLOGY OF THE COMBS AND THEIR BIOLOGICAL FUNCTION Gislén noted that the Comasteridae differ from all the other comatulids except the Heliometrinae in the possession of the so-called combs, which usually appear on the proximal pinnules but sometimes are found far out on the distal parts of the arm in Comaster and in occasional species [or individuals] of Comanthus (Vania), Comanthina, and Capillaster. In the Heliometrinae rudimentary or fairly well developed combs appear on one or more of the proximal pinnules. The genus Coma- tonia (see Part 3, page 288) is supposed by Gislén to belong to this subfamily of the Antedonidae. It has been observed (see Part 2, page 95, fourth paragraph) that in Heliometra glacialis the terminal segments of the arms and pinnules are devoid of ambulacral grooves. Gislén points out that we thus find the remarkable coincidence that in a group not belonging to the Comasteridae, in which the ambulacral furrows are unusually short, combs reappear, and that a change sets in suggesting a reduction of the ambulacral furrows. He recalled that the intestinal contents in Heliometra show remarkable differences from those of the rest of the comatulids catching plank- ton or detritus. The combs are developed at the ends of the pinnules. Only occasionally, as in Comantheria or Comanthus, are the most distal segments of the combed pinnules smooth. In Comatonia teeth appear on P, and P, almost from their bases. The teeth are usually developed on one side only, and are ranged in a single row on the side away from the mouth. In Nemaster, Comanthina, and Comanthus two rows of teeth occur, an inner row of large teeth, and an outer row of small ones. Transverse sections through the proximal combed pinnules and the distal pin- nules of Comantheria delicata grandis show that, at least in this form, a small interior tooth occurs inside the large exterior tooth, as in Nemaster, etc., although it is not discernible to the naked eye. The sections make it evident that these teeth are derived from the dorsal hooks that occur in most of the comatulids and are especially well developed in the Comasteridae; these have increased in size and have made their way up on one side of the pinnule. Usually on a pinnule that has not been transformed three dorsal hooks stand in a transverse row on the distal segments. When enlarged to form a comb the hook situated most adorally gains the preponderance and surpasses the others, which become more or less suppressed, in size. In regard to its combs Nemaster shows its primitive character in the occurrence of two rows of teeth in the comb. ‘The inner row is less evident in the more specialized genus Comaster, although quite distinct in sections. In Comanthus pinguis, in which the comb shows a pre- sumably secondary tendency to reduction, the inner row of teeth is very indistinct. In contrast to the distal ends of the other pinnules, those of the combed pinnules may be rolled into a tight spiral, as one of the interpinnular muscles, always the aboral, is very strongly developed. This muscle continues through a large part of the seg- ment, penetrating deeper on the proximal side than on the distal, and so locating the thin dividing wall between the muscles of the articulations somewhat distally of the center of the ossicles. As a consequence of this exceedingly one sided muscular development, the muscular side of the pinnule is much enlarged. The intruding A MONOGRAPH OF THE EXISTING CRINOIDS 69 hypertrophied muscle has pressed the dorsal coelome toward the adoral side. Pre- sumably migration of the dorsal hooks toward the aboral side is explained by the stronger calcification and development of the side turned away from the mouth. Gislén said that I endeavored to explain the biological function of the combs by assuming that they pick away foreign particles from the arms and pinnules, especially parasitic myzostomes. But he pointed out that the Comasteridae are almost more abundantly provided with these than any other family. He himself found an ento- parasitic ?Protomyzostomum on Comantheria delicata grandis, and ectoparasitic myzo- stomes on Comanthus pinguis and Comissia parvula. So he did not think the function of the combs could to any essential extent be the one given by the author. The dif- ferent species of myzostomes are in general each confined to its own particular coma- tulid or to a few species. If the principal task of the combs had been to free the animal from parasitic myzostomes, either these parasites would soon have been rooted out in the Comasteridae, or they would have to become endoparasitic to escape the combs. According to Gislén, in either case the combs as organs for exterminating myzostomes would have become superfluous at about the same time as they appeared. He said that according to this theory it is difficult to understand why certain genera should acquire combs far out on the distal pinnules, for the myzostomes practically always occur on the disk, or on the arms in its immediate neighborhood. For the particles of food become more and more concentrated toward the mouth. Still, Gislén said, there is a possibility that in certain forms the combs have a general cleansing function. Gislén maintained that the dorsal hooks in comasterids with reduced cirri act as anchoring organs offsetting the loss of the cirri. He noted that this seems to be the case also with the grooveless pinnules in Comatulella brachiolata and remarked that in these the coiling takes place more toward the laterodistal side than toward the dorsal side. He found this theory supported partly by the occurrence of the dorsal hooks, so generally and exceptionally strongly developed in the Comasteridae, and partly by the reduction and disappearance of the cirri in certain species. In Comanthus pinguis the cirri are strong and the dorsal hooks weak, and in Comanthus timorensis the conditions are reversed. His idea of the development of the teeth of the combs from the dorsal hooks is as follows: Originally all the distal segments of the pinnules possessed small dorsal hooks. Dorsal hooks appear in a more or less rudimentary form on the outermost segments of the more distal pinnules in most of the comatulids. On the other hand, the oral pinnules are smooth distally except in the subfamily Heliometrinae of the family Antedonidae. The combs must be considered to have appeared first on the oral pinnules. In forms in which the combs extend far out on the distal pinnules this feature appears only at an ontogenetically very late stage. The oral pinnules in most of the comatulids are developed into tactile organs and are very sensitive. In a number of forms, thanks to short segments, they are also exceedingly flexible. Because of the formation of the facets of the segments this flexibility can function only in a lateral direction. Thanks to the dorsal hooks and the great flexibility of the oral pinnules, the latter were able to catch hold of objects falling upon the disk or outside the ambulacral grooves that were too large to be 70 BULLETIN 82, UNITED STATES NATIONAL MUSEUM carried away by the ciliary currents. In order to do this, the pinnule tip has been greatly contracted aborally. Between the usual distal pinnule segments the transversely striated muscle bundles are very small, almost rudimentary, and the greater part of the facets of the segments is occupied by the ligamentary connections. This structure does not permit more than rather weak and slow contraction. Between the pinnule segments in the region of the combs, on the other hand, the transversely striated muscle bundles, especially those of the aboral part, are enormously developed, which allows of strong and rapid contraction. Gislén believes that the aboral muscle, by stronger growth gradually forced the weaker aboral muscle toward the side and thus secured a more medioventral position for itself. Owing to this the pinnule tip was no longer able to curve in laterally when contracted, but was rolled up ventrally, and the comb must, in order not to lose its effect, proceed up the aboral side of the pinnule. Gislén said that in this way we may explain the greatly one-sided strengthening of the muscles, the torsion of the pinnule tip (twisted at right angles to the longitudinal direction), and the progress of the dorsal spines up the lateral side turned from the mouth. In the Comasteridae, as pointed out by Gislén, branches from the dorsal nerve run to the bases of the teeth, and these may possibly be of some importance in con- nection with the senses of touch or of taste. As a consequence of the rich innervation the pinnule tip should be able to determine whether an object seized is edible or inedible. It is often difficult to draw a sharp line between the ciliary currents for cleansing purposes and those carrying food. Here, too, the possibility of an alterna- tion of function may easily be presumed. When the combs have once begun to grasp small particles that have fallen upon the animal, it may easily be imagined that instead of letting the object drop down between the arms they may put it into the ambulacral groove or into the mouth. From this it is not a very long step to begin picking up or pinching off small particles—algal threads, bits of bryozoans, pieces of leaves, ete.—which do not fall directly down upon the animal, but which are found in its immediate neighborhood. In this way, according to Gislén, the formation of the combs upon the oral pinnules may be explained. The dorsal hooks on the distal pinnules were used as anchoring organs. These would sometimes catch hold of some soft loose object from which small pieces could be severed. By curving in the tip of the pinnule over the ambulacral groove of the arm or of the pinnule itself the combs were able to pass on the particles of food to the ciliary current to be transported to the mouth. In this way a torsion must also have been developed in the formation of the pinnule, owing to the one sided development of the muscles. The combs upon the distal pinnules might have arisen in this way. The animal by means of these structures was supplied with an extra (and more heavily digested) contribution of food, the intestine therefore becoming lengthened and the ambulacral grooves reduced as being unnecessary. In certain comasterids, as in the species of Comissia, Comatilia, and some forms of Comactinia echinoptera, the combs are limited to one or two pairs of the proximal pinnules. These forms are endocyclic, except possibly one form of Comactinia echinoptera. Gislén believes that in these and other endocyclic comasterids the combs are of predominating importance for cleansing and anchoring purposes. A MONOGRAPH OF THE EXISTING CRINOIDS 71 In a specimen of the almost perfectly endocyclic Comaster novaeguineae with 67 fully developed and 9 short regenerating arms averaging 63 mm. in length the weight was 11.6 grams and the volume 8.25 ce. If all the arms and pinnules had possessed functional ambulacral grooves the total length of these grooves would have been 38.1 meters. The ratio between the ambulacral grooves and the volume would have been 4.6. But the two posterior radii bearing 34 arms show only a rudimentary ambulacral furrow on the arms, while a large number of the pinnules have no ambulacral grooves. The animal’s effective ambulacral grooves are therefore probably at most not more than from 25 to 30 meters long, the ratio thus being between 3.0 and 3.6—unusually small for an ordinary comatulid, but unusually large for a comasterid. Most of the comasterids are exocyclic. The most advanced types, as Comaster, Vania, and Capillaster, have combs even upon some of the distal pinnules, and are in most cases exocyclic, except various very slender species of Comaster, as C. minimus, C. novaeguineae, and C. sibogae. Among these, too, the combs are most likely of importance for gathering food. The combs appear in the Comaster type upon every other, or every third, pinnule. About half the genital pinnules are provided with combs. As a result of the development of sex products these pinnules are greatly swollen, clumsy, and presumably not very flexible. It is easy to imagine that this somewhat impedes the combs when engaged in passing the collected food to the ambulacral groove, and that instead it is probably easier to pass the food from one comb to another along the arm toward the mouth. As the combs occur on the arms with no ambulacral groove as well as on the others, an explanation of this sort becomes a necessity, if the combs are to be regarded as of any importance in supplying food. In a specimen of Comanthus parvicirra form vaniipinna the most distal combed pinnule on a grooveless arm was 7 mm. long, while the distance to the nearest ambulacral groove was 18 mm. Gislén said there is an interesting parallel here to the similar method of obtaining food found in the ophiurans. The evolutionary plan, however, is realized in two different ways in the two cases. In the ophiurans it is the tube-feet, the equivalent of the tentacles of the crinoids, that take over the transportation of food to the mouth. The ambulacral groove, which orginally was undoubtedly ciliated, and used for transporting food to the mouth, became unnecessary and was closed by the margins growing together, finally coming to lie in the epineural canal. Ontogenetical evidence of this, according to Gislén, is found in the ambulacral groove still being open in very young individuals. In the Comasteridae it is the combs—the transformed dorsal hooks on the ends of the pinnules—that have begun to take over the transportation of nourishment. The process of evolution has not gone so far here, but there is an evident tendency toward the elimination of the ambulacral furrow. The process, however, takes place in a way different from that seen in the ophiurans. The epi- thelium of the ambulacral groove degenerates so that the difference between it and the surrounding epithelium disappears. In both cases there results a nonciliated arm. In the ophiurans the ciliary groove becomes enclosed in an epineural canal and is retained as a cord, now exclusively nervous, while in the comasterids the ambu- lacral folds are smoothed out and the ambulacral groove obliterated, the complete disappearance in this case evidently being possible because there are already nervous systems developed in the arms. 72 BULLETIN 82, UNITED STATES NATIONAL MUSEUM DETAILS OF ARM STRUCTURE In his report upon a mass of dissociated comatulid ossicles from the Eocene (Middle Jacksonian) at Baldock, Barnwell County, S. C., Professor Gislén took occasion to classify the types of arm ramification found in the comatulids, both recent and fossil. He distinguished 85 types, which he analyzed in detail on the basis of available data taken from previously recorded specimens. He also reviewed with great care such of the structural features of the Himerometridae as would be of importance in the determination of dissociated fragments. From these dissociated ossicles he reconstructed Himerometra bassleri, sp. nov., most closely related to the recent H. persica, which is interesting in being the only representative of the family Himerometridae that has been found in America. From the same material he also segregated fragments that he referred to Micro- crinus ef. conoideus Emmons and to Glenotremites (Paleantedon) carolinianus, sp. =i TYPES OF ARM BRANCHING IN THE COMATULIDS Dr. Gislén has divided the arm branching in the comatulids into 85 types, which are (omitting those found only in fossil forms) as follows: (2) Eudiocrinus type (Hudiocrinus only). (3) Pentametrocrinus type (Pentametro- crinus and Atopocrinus). (4) Thaumatocrinus type (same as the preceding, but with 10 radials). (7) Comatula pectinata type (10-armed species of Comatula, and Coma- tulella). (8) Zygometra pristina type (Zygometra pristina and 10-armed young of Zygometra and Catoptometra). (9) Antedon type (the majority of comatulids). (10) Monachometra fragilis type I. (11) Promachocrinus type (Promachocrinus only). (15) Comatula rotalaria type (Comatula rotalaria only). (16) Comatella maculata type. (17) Neocomatella pulchella type I. (18) Neocomatella alata type. (19) Comatella brachycirra type. (20) Asterometra type. (21) Stenometra diadema type. (22) Monachometra fragilis type Il. (23) Diodontometra type. (25) Catoptometra rubroflava type. (26) Catoptometra hartlaubi type I. (27) Perissometra macilenta type. (28) Pachylometra angusticalyzr type. (29) Heterometra bengalensis type. (80) Comanthus solaster type. (31) Adelometra angustiradia type. (32) Crinometra imbri- cata type. (32a) The abnormal Comanthoides spanoschistum type. (33) Capillaster coccodistoma type. (34) Nemaster discoidea type. (35) Heterometra compta type. (37) Comatella stelligera type. (38) Catoptometra magnifica minor type. (39) Neo- comatella pulchella type Il. (40) Crinometra coneinna type. (41) Stylometra spinifera type. (42) Stephanometra spicata type. (43) Mariametra subcarinata type. (44) Comanthus parvicirra type 1. (45) Comaster serrata type. (46) Crossometra sep- tentrionalis type. (47) Zygometra comata type. (48) Heterometra crenulata type. (49) Comantheria rotula type. (50) Oceanometra magna type. (51) Comanthus japonica type. (52) Comanthus timorensis type 1. (53) Himerometra martensi type. (54) Crinometra transversa type. (55) Catoptometra hartlaubi type Il. (56) Coman- thus parvicirra type II. (57) Comanthus timorensis type II. (58) Crinometra in- sculpta type. (59) Heterometra nematodon type. (60) The irregular Comanthus timorensis type III. (61) Nemaster grandis type. (62) Capillaster multiradiata type. (63) Capillaster macrobrachius type. (64) Capillaster mariae type. (65) Comaster gracilis type I. (66) Comatella nigra type. (67) Catoptometra magnifica type. (68) Oxymetra erinacea type. (69) Pontiometra andersoni type. (70) Coman- A MONOGRAPH OF THE EXISTING CRINOIDS 73 thus timorensis type IV. (71) Comaster gracilis type II. (72) Comaster multibrachiata type. (73) Comantheria delicata type. (74) Zygometra microdiscus type I. (75) Comanthus pinguis type. (76) Comanthus bennetti type. (77) Comantheria polycnemis type. (78) Comantheria intermedia type. (79) Comantheria alternans type. (80) Comanthina belli type. (81) Zygometra microdiscus (elegans) type Il. (82) Coman- theria briareus type. (83) Himerometra magnipinna type. (84) Comanthina schlegelii type. (85) Capillaster sentosa type. EXTENT OF THE DISK Gislén says that the disk in Clarkometra elegans reaches to brachials 3+-4; in Tropiometra afra macrodiscus and in Antedon petasus to brachials 4 or 5; in Leptone- master venustus to brachial 5; in Promachocrinus to brachial 6; in Notocrinus to brachial 7; in Pentametrocrinus diomedeae to brachials (5-) 8; in Heterometra crenulata and Amphimetra jacquinoti to brachial 6; in Heliometra and Anthometra to brachials 7-8; in Atopocrinus to brachial 9; in Comanthus pinguis only to the second or fourth element in the IIIBr series; and in Comaster novaeguineae only to the third or fourth element in the [Br series. 208244406 Order COMATULIDA A. H. Clark (continued) The general account of this order, with the diagnosis, synonymy, range, and keys to the included families and higher groups, will be found in Part 3, pp. 69-74. Suborder OLIGOPHREATA A, H. Clark (continued) The synonymy, diagnosis, range, and history of this suborder will be found in Part 3, pp. 74-76. Superfamily MARIAMETRIDA Gislén Mariametrida Gist6N, Zool. Bidrag Uppsala, vol. 9, 1924, pp. 230, 233, 235, 237, 240; Arkiv for Zool., vol. 19, No. 32, 1928, p. 4.—A. H. Ciark, Rec. Indian Mus., vol. 34, pt. 4, 1982, p. 554; Treubia, vol. 14, livr. 2, 1933, p. 211.—GisLfn, Kungl. Fysiogr. Sillsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 7, 49.—A. H. Ciarx, John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1936, p. 88.—H. L. Crarx, Mem. Mus. Comp. Zool., vol. 55, 1938, p. 42. Diagnosis.—A superfamily of the suborder Oligophreata (see Part 3, p. 74) in which there is no comblike structure on the distal portion of the oral pinnules; the distal pinnules are never prismatic; the genital pinnules are never prismatic, though occasionally a few of the basal segments may be carinate; the oral pinnules, which may be long or short, slender or stout, or stout basally and slender distally, very flexible to stiff and spinelike, composed of long or short segments, usually show some distinct trace of carination on more or fewer of the basal segments, and are some- times sharply prismatic throughout; and the mouth is always central or subcentral, the anal tube being more or less marginal. Sacculi are always present, though often in small numbers. The arms vary from 5 to 100 in number but are usually between 15 and 40, though not infrequently 10. Remarks.—The superfamily Mariametrida as herein understood includes the families Zygometridae, Eudiocrinidae, Himerometridae, Mariametridae, and Colobo- metridae. In addition to these Gislén included also the family Tropiometridae, which seems to me to be quite out of place here, its prismatic pinnules, very broad division series and first two brachials, ambulacral deposits, and cirri indicating a rather close relationship to the Thalassometridae and Charitometridae. The family Zygometridae as understood by Gislén included the genus Eudiocri- nus, Which is herein set apart as a distinct family. The family Stephanometridae, which was included in the Mariametrida by Gislén, is here merged with the Maria- metridae. Within the superfamily Mariametrida the families Zygometridae and Eudiocrini- dae agree, but differ from the other families, in having a syzygy between the elements of the [Br series, and also in having a more or less completely plated disk, although 74 A MONOGRAPH OF THE EXISTING CRINOIDS 75 adambulacral deposits in the perisome of the pinnules are much reduced or absent (see Part 2, p. 236). These two families together thus form a distinct group, includ- ing two rather strikingly different types, within the superfamily Mariametrida, this group contrasting rather sharply with the group including the very closely related Himerometridae, Mariametridae, and Colobometridae. Comparisons.—The members of the superfamily Mariametrida are very easily dis- tinguished from the species of the superfamily Comasterida (see Part 3, p. 76) by the complete absence of a comblike structure on the distal portion of the oral pinnules. In most cases they are also readily distinguished by the central or subcentral mouth; but it must be remembered that in several of the species in the Comasterida the mouth is always, sometimes, or occasionally central. The species of Mariametrida are readily distinguished from the species in the super- family Tropiometrida by the middle and distal pinnules, which in the species of Tropiometrida always show a prominent sharp dorsal carination that becomes very conspicuous under a low power when the pinnule is dried. Furthermore, the outer pinnules in the species of Tropiometrida almost invariably are provided with side- plates and covering plates (see Part 2, figures 819-823, p. 392; figures 831-848, p. 405), which are easily visible in dried pinnules under a hand lens. It should be remembered, however, that in the genus Tropiometra these are greatly reduced (see Part 2, fig. 797, p. 372) and are not visible under a hand lens. From the species included in the suborder Macrophreata the species of Maria- metrida are distinguished by the relatively robust and short-segmented distal pin- nules in which the first two segments are not appreciably broadened; by the irregular and relatively wide spacing of the syzygies; by the presence of some trace of carination on at least the basal segments of the oral pinnules; by the more or less abrupt differ- entiation of one or more of the oral pinnules, which are not simply elongated; and by the stronger and less easily detached cirri. Most of the species of Mariametrida have more than 10 arms, while in the Macro- phreata very nearly all the included species have 10 arms only, though one has 20 owing to the occurrence of 10 instead of 5 radials, and in one or two there may be a few more than 10. The 10-armed species of Mariametrida may be distinguished from the species of Macrophreata by the enlarged proximal pinnules, which are either stout and composed of relatively short segments, or if slender are more or less obviously prismatic with a group of spines, or a conspicuous process, at the distal end of the dorsal ridge on each segment. In the 10-armed species of Mariametrida also the brachials are almost always shorter, and the cirri are stronger and far less likely to be detached from the centrodorsal. The very few species of Macrophreata in which the brachials are short have especially fragile cirri, and P, is composed of a great number of exceedingly short segments, looking more or less like a string of little beads. Both the Mariametrida and the Macrophreata include genera containing 5-armed species. In the 5-armed species of Mariametrida (included in the genus Hudiocrinus) the arms are short, the oral pinnules are much enlarged and sharply prismatic, and the second pinnule is on the second ossicle beyond the first (which is on the second post- radial ossicle—see fig. 127, p. 79, of Part 2) instead of on the immediately succeeding ossicle as would be expected. In the 5-armed species of Macrophreata (included in the genera Pentametrocrinus and Atopocrinus) the arms are very long with greatly elongated 76 BULLETIN 82, UNITED STATES NATIONAL MUSEUM brachials and exceedingly fragile, the proximal pinnules are either absent or do not differ appreciably from those succeeding, and there is no gap in the succession of pinnules. In both the Mariametrida and the Macrophreata there are species in which one or more of the proximal pinnules, most commonly the first inner pinnule (P,), are absent. Most of these species in the Mariametrida have 10 arms only. In the Mariametrida the species with deficient pinnulation are all included in the single family Colobometri- dae. They are very easily distinguished from all the species in the Macrophreata by having on the dorsal surface of the cirrus segments either a serrate transverse ridge, or two tubercles or spines situated side by side, or a transverse row of 3 or 4 spines or tubercles; the dorsal surface of the cirrus segments in the Macrophreata is either broadly or more or less sharply rounded, or else is carinate in the middorsal line. As the great majority of the 10-armed species of Mariametrida belong to the family Colobometridae, in the case of doubtful 10-armed forms the dorsal surface of the cirrus segments should always be the first feature examined. Family ZYGOMETRIDAE A. H. Clark Series I of Antedon P. H. Carpenter, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 94; Journ. Linn. Soc. (Zool.), vol. 21, 1889, p. 304. Elegans group fof Antedon] P. H. Carpenter, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 266; Journ. Linn. Soe. (Zool.), vol. 21, 1889, pp. 304, 307.—Harriavs, Bull. Mus. Comp. Zool., vol. 27, 1895, p. 1386 (discussion), p. 137 (structural position).—A. H. Cuarx, Proc. U. S. Nat. Mus., vol. 33, 1907, pp. 127, 130 (descriptions of new species); Smithsonian Misc. Coll., vol. 50, 1907, p. 348 (the equivalent of Zygometra, gen. nov.). Zygometridae A. H. Ciark, Proce. Biol. Soc. Washington, vol. 21, 1908, p. 135 (includes Eudiocrinus and Zygometra); Proc. U. 8. Nat. Mus., vol. 34, 1908, p. 210 (in key), p. 211 (includes Zygo- metra; range; not represented in the Hawaiian Islands), p. 212 (range of included genus), p. 276 (revised description; range; includes Zygometra and Eudiocrinus); vol. 35, 1908, p. 118, figs. 14, 15, p. 119, fig. 16 (arm structure); Amer. Nat., vol. 42, 1908, p. 541 (characteristic of Indo- Pacific-Japanese fauna), p. 722 (ecology), p. 725 (color); Geogr. Journ., vol. 32, 1908, p. 602 (characteristic of Indo-Pacific-Japanese region), p. 605 (ecology); Proc. Biol. Soc. Washington, vol. 22, 1909, p. 2 (modification of cirri and lower pinnules as in the Thalassometridae) ; Amer. Nat., vol. 43, 1909, p. 256 (absence from the Red Sea noteworthy); Proc. U. S. Nat. Mus., vol. 36, 1909, p. 361 (part of Antedon as used by P. H. Carpenter); Proc. Biol. Soc. Washington, vol. 22, 1909, p. 174 (only family witn a syzygy in the IBr series; referred to Comatulida Oligo- phreata); Amer. Nat., vol. 43, 1909, p. 581 (arm structure); Vid. Medd. Nat. Foren. Kgben- havn, 1909, p. 119 (occurs at Singapore; has large eggs that presumably develop rapidly; ecology); Amer. Journ. Sci., vol. 29, 1909, p. 43 (articular facet of a radial, showing develop- ment); Proc. U. S. Nat. Mus., vol. 39, 1911, p. 538 (articulation between elements of IBr series really a remarkably perfect pseudosyzygy, as demonstrated by Zygometra pristina, sp. noy); vol. 40, 1911, p. 649 (referred to Oligophreata); Die Fauna Siidwest Australiens, vol. 3, Lief. 13, 1911, p. 438 (1 genus and 3 species in Australia); Mem. Australian Mus., vol. 4, 1911, p. 717 (proportion of the species of this family in the fauna of Australia as known to P. H. Car- penter), p. 720 (proportion of the species of this genus in the Australian fauna), p. 725 (Eudio- crinus and Catoplometra absent from Australia; this family disproportionately developed in Australia), p. 728 (closely related to the Himerometridae), p. 729 (in key), p. 731 (1 genus in Australia); Crinoids of the Indian Ocean, 1912, p. 3 (Hyponome sarsii the detached visceral mass of some species of this family), p. 6 (confined to the East Indian region; number of included genera and species), p. 11 (occurs west to the west coast of the Malay Peninsula and the Andamans), p. 14 (very restricted distribution; littoral), p. 21 (distribution in detail), p. 42 (closely related to the Himerometridae), pp. 44, 45, 47, 48, 50, 52 (in keys), p. 56 (key to the A MONOGRAPH OF THE EXISTING CRINOIDS 77 included genera); Die Fauna Siidwest Australiens, vol. 4, Lief. 6, 1913, p. 311 (discussion and comparison with the Himerometridae).—SprinGEeR and Cuark, Zittel-Eastman’s Paleontology, 1913, p. 236 (in Oligophreata).—A. H. Cuarx, Bull. Inst. Océanogr., Monaco, No. 294, 1914, pp. 7, 8 (relation to temperature); Journ. Washington Acad. Sci., vol. 4, 1914, pp. 559-563 (correlation of geographic and bathymetrical ranges), p. 582 (relation to temperature) .— ALEXANDER, Rec. Western Australian Mus., vol. 1, pt. 3, 1914, p. 108—A. H. Cuarx, Rec. Western Australian Mus., vol. 1, pt. 3, 1914, p. 114 (species collected by the Endeavour in Western Australia); Journ. Washington Acad. Sci., vol. 5, 1915, pp. 126-134 (bathymetric range; phylogenetic and paleontological significance); Amer. Journ. Sci., vol. 40, 1915, p. 63 (recent genera with fossil representatives, number of included species, recent and fossil, and their geological, geographical, bathymetrical and thermal ranges), p. 67 (detailed discussion of the bathymetricalrange); Int. Rev. gesamt. Hydrobiol. und Hydrogr., 1915, p. 224 ff. (detailed account of the distribution of Australian species); Smithsonian Misc. Coll., vol. 65, No. 10, 1915, p. 25 ff. (phylogenetic study) —H. L. Cuark, Biol. Results Fishing Exper. F. I. S. En- deavour, 1909-14, vol. 4, pt. 1, 1916, p. 21 (Oreometra a connecting link between this family and the Calometridae, if not actually a member of the Zygometridae).—A. H. Cuarx, Unstalked crinoids of the Siboga Exped., 1918, p. 59 (discussion of genera and species).—GIsL&£N, Kungl. Svenska Vet. Akad. Handl., vol. 59, No. 4, 1919, p. 19—A. H. Cruarx, Univ. Iowa Studies in Nat Hist., vol. 9, No. 5, 1921, p. 13 (includes fossil and recent species), p. 14 (not represented in the West Indies); Smithsonian Misc. Coll., vol. 72, No. 7, 1921, pp. 3, 17.—Guis.in, Nova Acta Reg. Soc. Sci. Upsaliensis, ser. 4, vol. 5, No. 6, 1922, pp. 9, 64; Zool. Bidrag Uppsala, vol. 9, 1924, pp. 19, 30, 31, 35, 42, 77, 84, 89, 99, 230, 231, 235, 240; Vid. Medd. Dansk Nat. Foren., vol. 83, 1927, p. 20.—A. H. Cuarx, Journ. Linn. Soc. (Zool.), vol. 36, 1929, p. 637; Treubia, vol. 14, livr. 2, 1933, p. 211.—Exman, Zoogeographica, vol. 2, Heft 3, 1934, p. 364 (zoogeo- graphic and paleontological relationships).—A. H. Ctarx, Temminckia, vol. 1, 1936, p. 300.— Gistin, Kungl. Fysiogr. Sallsk. Lund Férh., vol. 7, No. 1, 1936, p. 8—A. H. Cuark, John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1936, p. 103—H. L. Cuarx, Mem. Mus. Comp. Zool., vol. 55, 1938, p. 29. Diagnosis.—A family of the superfamily Mariametrida in which the elements of the IBr series are united by syzygy and the arms are 10 or more in number. The perisome of the disk is more or less completely covered by rounded plates, which do not, however, form a solid pavement (see Part 2, p. 226, fourth paragraph); the lateral perisome of the pinnules may contain a few spicules but is usually quite without cal- careous deposits (see Part 2, p. 236). Geographical range.-—From southern Japan, the Bonin and Philippine Islands, and Hongkong southward to Sandon Bluffs, New South Wales, and to between Fremantle and Geraldton, Western Australia, and westward to Ceylon. Bathymetrical range-—From the shore line down to 914 meters; the species are mainly inhabitants of shallow water. Remarks.—The occurrence of a syzygy between the elements of the IBr series easily distinguishes the species of this family from all other comatulids except only those belonging to the family Eudiocrinidae. The latter, however, possess but five arms, a character that gives them a very distinctive appearance. In the genus Catoptometra the syzygy between the elements of the IBr series is very brittle, and specimens of all the species of this genus when captured in the usual way by the dredge always have at least one of the postradial series, and not infrequently all of them, broken away at this point so that the syzygy is very evident. Further- more, the species of Catoptometra have a very distinctive appearance owing to the stout, smooth, strongly recurved, and usually short cirri, and the unusual develop- ment of the spines on the distal borders of the brachials and usually also on the borders 78 BULLETIN 82, UNITED STATES NATIONAL MUSEUM of the elements of the division series. In color most of them are unique in being scarlet and bright yellow in usually broad alternating bands; but some of them are, and others may be, olive-green, brown, or dull yellow. In the genus Zygometra the syzygy between the elements of the [Br series is tough and is almost never broken through during capture. Their numerous arms give these species a striking similarity to species of Himerometridae with which they are easily confused. In identifying any multibrachiate comatulid, therefore, the first step after the determination of the presence or absence of a comb on the oral pinnules (the presence of a comb places the specimen at once in the Comasteridae) should always be the criti- cal examination of the articulation between the elements of the IBr series. If this is a fine straight dotted line, indicating the presence of a syzygy, the specimen belongs to some species of Zygometra. It is not long before one becomes very adept at recognizing the species of Zygometra at sight, for they have a very characteristic appearance difficult to define. In the smaller species the usually long sharp spines on the outer cirrus segments distinguish them at once from any of the species of Himero- metridae with which they might be confused, while the large stout cirri of the large Australian species, coupled with the characteristic oral pinnules, are equally diagnostic. Nevertheless it is always possible to confuse the small species with species of Hetero- metra and the large species with species of Himerometra, so that no matter how familiar one may be with these creatures the articulation between the elements of the IBr series should invariably be examined. History.—In his report upon the comatulids collected by the Challenger published in 1888 Dr. P. H. Carpenter placed the three forms of Antedon, as understood by him, in which the elements of the [Br series are united by syzygy instead of synarthry (Antedon fluctuans, A. multiradiata, and A. microdiscus) under the heading Antedon, Series I. Although these three forms were the only ones formally included by him under this heading, he said that it is probable Miller’s Comatula tessellata (= Amphi- metra tessellata) belongs also to Series I, though he had never been able to get a sight of the type specimen, it being the only one of all the described species of Antedon—if, indeed, it be an Antedon—that he had not personally examined. As contemplated by Carpenter, Series I of Antedon, including species in which the elements of the IBr series are united by syzygy, was the equivalent of Series II, including 10-armed species, Series ITI, including species with I Br 2 series, and Series IV, including species in which the I[Br series are 4(3-+4). In a note written after his account of the forms included under Series I had been printed, Carpenter said that these forms may be considered conveniently as belonging to the Elegans group, named for Antedon elegans, which had been described by Prof. F. Jeffrey Bell in 1884, but in which the occurrence of a syzygy between the elements of the IBr series had escaped notice until after his discus- sion of Series I. In his account of the comatulids collected by Dr. John Anderson in the Mergui Archipelago published in 1889 Carpenter further discussed the Elegans group in con- nection with Antedon elegans (Antedon elegans is Zygometra elegans, but the specimens from the Mergui Archipelago represent Z. comata) and described a new species, Antedon (Pontiometra) andersoni, which he said may be for the present referred to the Elegans group. A MONOGRAPH OF THE EXISTING CRINOIDS 79 Dr. Clemens Hartlaub in 1895 divided the comatulids into two series, Series I, with plated ambulacra, divided into two sections, a, with the elements of the IBr series united by synarthry, including the Basicurva, Acoela, Spinifera, and Granulifera groups, and 6, with the elements of the IBr series united by syzygy, including the Elegans group only; and Series II, with unplated ambulacra, including the Eschrichti, Milberti, Tenella, Palmata, and Savignyi groups. In my paper on new genera of recent crinoids published on April 11, 1908, the new family Zygometridae was proposed, though not formally defined, and in it were included the genera Eudiocrinus and Zygometra. In a paper published on May 14, 1908, though written before the one just noticed, the family Zygometridae is given as including the genus Zygometra only, and the range of this genus (which at that time included Catoptometra) is added. The characters of the family are made evident 1n a key to the families of the Comatulida having recent representatives. The genus Eudiocrinus, together with Decametro- crinus, is included in the family Eudiocrinidae. At the time this paper was written my idea of Hudiocrinus was based upon the species assigned to that genus and described in detail and figured in the Challenger report, of some of which I had abun- dant material collected by me in southern Japan. After the paper had been submitted for publication, I received from Japan a specimen of a new species of Hudiocrinus (variegatus) allied to the type species, E. indivisus, which showed that these forms are allied to the species of Zygometra, whereas those described in the Challenger report are of a wholly different nature. For the latter in the preceding paper the genus Pentametrocrinus had been proposed and the new family Pentametrocrinidae had been suggested to include this and the genus Decametrocrinus. This was explained in detail, the families Zygometridae and Pentametrocrinidae were diagnosed, and the genus Eudiocrinus was redefined and the included species listed, in a paper published on June 20, 1908. In 1909 the family Zygometridae was assigned by the author to the Comatulida Oligophreata, together with the families Comasteridae, Himerometridae, Colobo- metridae, Thalassometridae, and Tropiometridae. The status of the family Zygo- metridae has since remained unchanged; but in the present work the genus Hudiocrinus has been removed from it, the special family Eudiocrinidae (redefined) being created for its reception. KEY TO THE GENERA OF THE FAMILY ZYGOMETRIDAE a1, Segments in outer portion of cirri with prominent sharp dorsal spines (Philippine Islands and Hongkong south to Sandon Bluffs, New South Wales, and to between Fremantle and Geraldton, Western Australia, and westward to Ceylon; 0-128 (?135] meters) _.---_-- Zygometra, p. 79. a?. No dorsal spines on distal cirrus segments (southern Japan, Bonin and Philippine Islands, Hongkong, and southward to Lesser Sunda Islands; 25-914 meters) __Catoptometra, p. 125. Genus ZYGOMETRA A. H. Clark Hyponome Lov&n, Forh. Skand. Naturf. Mgde Christiania, vol. 10, 1868 (1869), p. LIv (description; genotype, H. sarsii [= detached visceral mass of Antedon microdiscus Bell, 1882]); Ann. Mag. Nat. Hist., ser. 4, vol. 4, 1869, p. 159 (same); Can. Nat., new ser., vol. 4, 1869, p. 265 (same) .— Lirxeyn, Vid. Medd. naturh. Foren. Kjgbenhavn, 1869 (1870), pp. 185, 186, footnote, 187, 188, footnote—P. H. Carprnter, Proc. Roy. Soc., vol. 28, 1879, p. 388 (identity with Antedon) ; Nature, vol. 19, 1879, p. 450.— PERRIER, Compt. Rend. Acad. Sci., vol. 96, 1883, p. 450.— P. H. Carpenter, Ann. Mag. Nat. Hist., ser. 5, vol. 11, 1883, p. 336 (identity); Challenger 80 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Reports, Zoology, vol. 26, pt. 60, 1888, pp. 87, 89, 97.—Barumr, in Lankester, A treatise on zoology, pt. 3, Echinoderma, 1900, p. 195.—A. H. Cuarx, Smithsonian Mise. Coll., vol. 50, 1907, p. 348 (not available as a substitute for Zygometra); Proc. U. 8. Nat. Mus., vol. 34, 1908, p. 524 (type unidentifiable; not available as a substitute for Zygometra). Comatula (part) P. H. Carpenter, Trans. Linn. Soc. (Zool.), ser. 2, vol. 2, 1879, p. 23, footnote. Hypomene Wacusmuru AND Sprinaer, Proc. Acad. Nat. Sci. Philadelphia, 1881, p. 202 (p. 28 of separate). Antedon (part) Beu, Proc. Zool. Soc. London, 1882, p. 534, and following authors. Hyponeme Rouieston and Jackson, Forms of animal life, 1888, p. 573. Zygometra A. H. Crark, Smithsonian Mise. Coll., vol. 50, 1907, p. 339 (new species described as Zygometra koehleri [=Catoptometra hartlaubi]), pp. 345, 347 (diagnosis; genotype, Antedon microdiscus Bell, 1884), Bull. Mus. Comp. Zool., vol. 51, 1908, p. 246 (characters; genotype) ; Proc. Biol. Soe. Washington, vol. 21, 1908, p. 134 (closely related to Eudiocrinus, restricted; arm structure), p. 135 (assigned to the family Zygometridae); Proc. U. S. Nat. Mus., vol. 34, 1908, p. 211 (referred to the Zygometridae), p. 212 (occurs in Japan [refers to Catoptometra] ), p. 271 (relation to Eudiocrinus), p. 276 (referred to the Zygometridae), p. 542 (listed); vol. 35, 1908, p. 118, fig. 15 (arm structure); Proc. Biol. Soc. Washington, vol. 22, 1909, p. 2 (cirri and lower pinnules comparable to those of Thalassometridae), p. 174; Die Fauna Siidwest-Austra- liens, vol. 8, Lief. 13, 1911, p. 439 (8 peculiar Australian species); Mem. Australian Mus., vol. 4, 1911, p. 715 (recent species first described in 1882; fossil representatives), p. 724 (Australian species), p. 732 (in key), p. 734 (key to Australian species), p. 760 (reference; characters; range) ; Rec. Indian Mus., vol. 7, pt. 3, 1912, p. 267 (new species discovered related to Z. comata but occurring farther west); Smithsonian Misc. Coll., vol. 60, No. 10, 1912, p. 1 (Hyponome sarsi the detached visceral mass of some species of this genus); Crinoids of the Indian Ocean, 1912, p. 9 (occurs in shallow water in a very limited area; also in the European Jura; possesses additional species in Australia), p. 10 (does not extend north of Hongkong; absent from Japan; reason), p. 11 (occurs in Mergui Archipelago), p. 21 (range; this closely parallels that of Comatula solaris and C. pectinata), p. 56 (in key), p. 103 (synonymy; type); Die Fauna Siidwest-Australiens, vol. 4, Lief. 6, 1913, p. 311 (discussion of the small species and comparison with those of Heterometra, and of the large species and comparison with those of Himerometra); Journ. Washington Acad. Sci., vol. 5, 1915, p. 8 (very highly developed in the Australian fauna).—F. W. Cuarke and Wueeter, U.S. Geol. Surv. Prof. Pap. 90-L, 1915, p. 195 (inorganic constituents of the skeleton) ; Prof. Pap. 102, 1917, pp. 23 ff. (same).—A. H. Cuarx, Unstalked crinoids of the Siboga-Exped., 1918, p. 59 (key to the included species).— F. W. Cuarke and Wueg.er, U. S. Geol. Surv. Prof. Pap. 124, 1922, p. 20 (analysis of skeleton).— Gist&n, Zool. Bidrag Uppsala, vol. 9, 1924, p. 35 (brachial homologies), p. 84 (distal intersyzygial interval); Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, p. 18. Diagnosis.—A genus of Zygometridae in which the outer cirrus segments are much shorter than the earlier and bear prominent dorsal spines. Geographical range-—From Hongkong and the Philippine Islands southward to Sandon Bluffs, New South Wales, and to between Fremantle and Geraldton, Western Australia, and westward to Ceylon. Bathymetrical range.—From the shore line down to 128 (?135) meters. The species are especially inhabitants of the sublittoral zone. Remarks.—of the six species included in this genus four are closely allied, while the other two are rather widely different from these and from each other. When typically developed, Zygometra elegans is easily distinguished from Z. microdiscus, but some specimens seem to be more or less intermediate, so that the former has been considered as simply a form of the latter. The relation between Zygometra elegans and Z. comata is essentially the same as that between Z. microdiscus and Z. elegans. In most cases Z. comata is readily distinguishable from Z. elegans, but specimens may A MONOGRAPH OF THE EXISTING CRINOIDS 81 be found that are intermediate in practically all their characters. Zygometra andro- meda is simply a highly ornate from of Z. comata. The range of the small and generalized Z. comata includes the whole Indo- Malayan region, and it also reaches northwestern Australia. Zygometra microdiscus and Z. elegans are confined to the Australian region; Z. microdiscus occurs in the Aru Islands from which Z. elegans is not known, but Z. elegans extends much farther south on both the east and the west coasts than Z. microdiscus. So far as we know at present Z. punctata is confined to northeastern and northern Australia and the Aru Islands, while Z. pristina has been reported only from the Philippines. History.—Prof. Sven Lovén’s so-called recent cystidean Hyponome sarsi, des- cribed in 1869, was based upon a detached visceral mass of some species of this genus. There can be little doubt that this species was the one later (1884) described by Bell as Antedon microdiscus. Though in all probability Antedon microdiscus, the type of the genus Zygometra, is identical with Hyponome sarsi, the type of the much earlier genus Hyponome (see page 95), no useful purpose would be served by substituting Hyponome for Zygometra. Technically the slight doubt that may be considered to exist in regard to the identity of Hyponome sarsi and Antedon microdiscus may be regarded as sufficient to justify the continued use of Bell’s name microdiscus in place of sarsi, and of the generic name Zygometra instead of Hyponome. In his report upon the Challenger comatulids published in 1888 Dr. P. H. Car- penter set apart as Antedon, Series I, those species in which the elements of the IBr series are united by syzygy. In this group he included fluctuans, multiradiata, and microdiscus. Later in the same volume he added Bell’s elegans, under which he placed fluctuans as a synonym, and also an unnamed species [= Pontiometra andersoni| from the Mergui Archipelago. Except for the inclusion of the last named, Carpenter’s Series I of Antedon coincides exactly with Zygometra as now understood. The genus Zygometra was diagnosed and the type species designated as Antedon microdiscus Bell, 1884, in a paper published on October 29, 1907. But in the article just preceding this in the same volume, which was published on the same day, a new species was described under the heading “Zygometra koehleri, sp. nov.’ In this volume the description of Zygometra koehleri is found on page 339, while the genus Zygometra is diagnosed on page 347. Zygometra koehleri is merely the young of the species now known as Catoptometra hartlaubi, and it might be argued that since the generic name Zygometra was first mentioned in the combination Zygometra koehleri, this form is really the type of Zygom- etra and consequently Zygometra should replace the later Catoptometra as used herein. As nothing would be gained by such substitution, which would certainly result in endless confusion, the generic names Zygometra and Catoptometra are in the following pages used in the sense in which they bave been uniformly employed during the past 30 years. As first proposed, the genus Zygometra included all the species of comatulids with the elements of the IBr series united by syzygy, but in 1908 those species in which the outer cirrus segments are smooth dorsally were removed from Zygometra and placed in the new genus Catoptometra, leaving Zygomeira as understood herein. 82 BULLETIN 82, UNITED STATES NATIONAL MUSEUM KEY TO THE SPECIES IN THE GENUS ZYGOMETRA at. More than 40 cirrus segments; more than 35 (usually 40 or more) arms. b!, More than 40 (65-100) arms; all, or most, of the IJIBr and following division series 4(3+ 4); usually 45-55 cirrus segments; proximal pinnules very stout and very long (20-32 mm.), tapering to a slender and flagellate tip; cirri very long and stout (Aru Islands and northern Australia south to Shark Bay, Western Australia, and Port Curtis, Queensland; 0-128 (Pheblimeters) eset soe eee ee cas eet nee ee So See se ae microdiscus (p. 82) b?. About 40 (usually 35-45) arms; most of the IIIBr and following division series 2; usually 35- 45 cirrus segments; proximal pinnules of moderate length (14-18 mm.) and not especially stout; cirri not especially long or stout (northern Australia south to between Fremantle and Geraldton, Western Australia, and Sandon Bluffs, New South Wales; Amboina; 0-89 TYISCOTIS) arr ee ere ee ese eh RE AES oe Aa eee eee elegans (p. 98) a. Less than 40 cirrus segments; less than 40 arms. b!. More than 10 arms; syzygy between the elements of the IBr series perfect. c!. Cirri moderately long, with usually 25-30 segments; usually 25-35 arms. d', Distal edges of radials and proximal and distal borders of the elements of the division series and first 2 brachials plain and unmodified (Mergui Archipelago to Lesser Sunda Islands, Holothuria Bank, the Moluccas, Salawatti, Philippines, and Hongkong; 0-73 meters) 22st Ades feces Se eee comata (p. 110) d. Distal edges of radials, proximal and distal borders of the elements of the division series, and to a lesser extent the proximal and distal ends of the first 2 brachials thickened and everted,this eversion being finely scalloped or tuberculated (?Ceylon)_ andromeda (p.110) c. Cirri short and rather stout, strongly recurved in distal half, with not more than 21 segments; not more than 21 arms (Aru Islands and northern Australia, south to Port Curtis, Queens- Jand$+0=50 (meters) oie Sek 3. 2S ceo. 5 2 eee a ee 2a punctata (p. 120) &. Only 10 arms; syzygy between the elements of the IBr series imperfect, the radiating ridges being developed only on peripheral half of joint faces; arms 50 mm. long; cirri XII, 20-21, 12 mm. long (Philippines; 33 meters) .--_. _.-..-2-2=----=----~=--=_-- pristina (p. 123) ZYGOMETRA MICRODISCUS (Bell) Puate 1, Figures 1, 2; Prats 2, Fiaures 3, 4, 6 (See also vol. 1, pt. 1, fig. 332 (cirrus), p. 283; pt. 2, fig. 186 (arm base), p. 79; pl. 13, figs. 1045, 1052 (pinnules) ; pl. 14, fig. 1060 (pinnule); pl. 25, figs. 1155-1157 (disk).] Hyponome sarsi Lovén, Forh. Skand. Naturf. Mgde Christiania, July 1868 (1869), p. L1v (descrip- tion and affinities; Cape York, Torres Strait); Ofv. Vet.-Akad. Forh., 1869, pp. 729-731; Compt. Rend. Acad. Sci. vol. 69, Sept. 1869, pp. 711, 712.—Bruu1nas, Can. Nat., new ser., vol. 4, 1869, p. 270.—Lovén, Can. Nat., new ser., vol. 4, 1869, p. 265 (detailed description; Cape York); Ann. Mag. Nat. Hist., ser. 4, vol. 4, 1869, p. 159.—Lirxen, Vid. Medd. Naturh. Foren. K¢b- enhayn, 1869, pp. 184, 185, 186, footnote (discussion) ; Can. Nat., new ser., vol. 4, 1869, p. 267.— Wrvitite Tuomson, Proc. Roy. Soc. Edinburgh, vol. 7, 1872, p. 417—P. H. CarPenTer, Proc. Roy. Soc., vol. 28, 1879, p. 888.—pn Lortot, Paléontologie frangaise, Terraine jurassiques, vol. 11, pt. 2, 1889, p. 436.—Gist&n, Zool. Bidrag Uppsala, vol. 9, 1924, p. 271 (food). Hyponome sarsii Lovin, Amer. Journ. Sci., ser. 2, vol. 48, 1869, p. 429——Wryvitte Tuomson, Proc. Roy. Soc. Edinburgh, vol. 7, 1871, p. 417— Scurttrer, Zeitschr. deutschen geol. Ges., vol. 30, Heft 1, 1878, p. 29.—Prrrier, Compt. Rend. Acad., Sci., vol. 96, 1883, p. 450; Ann. Mag. Nat. Hist., ser. 5, vol. 11, 1883, p. 223.—P. H. Carpenter, Ann. Mag. Nat. Hist., ser. 5, vol. 11, 1883, p. 336; Proc. Roy. Soc. Edinburgh, vol. 12, 1884, p. 369.—LocxkrnaTon, Standard Nat. Hist., vol. 1, 1884, p. 140.—P. H. Carpenter, Challenger Reports, Zoology, vol. 11, pt. 32, 1884, pp. 67, 85, 166; vol. 26, pt. 60, 1888, p. 97 (identity); Journ. Linn. Soc. (Zool.), vol. 24, 1891, p. 45 (identity) —Hartiaus, Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 25.— Neviant, Riy. Ital. Sci. Nat., ann. 11, 1891, fase. 2, p. 6—A. H. Cuark, Smithsonian Mise. Coll., vol. 50, 1907, p. 348 (identity); Mem. Australian Mus., vol. 4, 1911, p. 761 (same); Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p. 1 (history; identity), p. 11 (specific identity) ; A MONOGRAPH OF THE EXISTING CRINOIDS 83 Crinoids of the Indian Ocean, 1912, p. 3 (identity); Smithsonian Misc. Coll., vol. 72, No. 7, 1921, p. 22 (food). Comatula sp. P. H. Carpenter, Trans. Linn. Soc. (Zool.), ser. 2, vol. 2, 1879, p. 23, footnote (species with first and second distichals and first and second brachials united by syzygies, like second and third radials). Hypomene sarsi WACHSMUTH and SPRINGER, Proc. Acad. Nat. Sci. Philadelphia, 1881, p. 202 (p. 28 of separate), footnote. Antedon microdiscus Brut, Proc. Zool. Soc. London, 1882, p. 534 (specific formula only) —P. H. CarpEentTeER, Proc. Zool. Soe. London, 1882 (1888), p. 747 (specific formula) —Br.u, Report Zool. Coll. H. M.S. Alert, 1884, p. 155 (specific formula), p. 163 (description; Port Molle, 12 fathoms; Nicol Bay), pl. 15——P. H. Carpenter, Quart. Journ. Micr. Sci., vol. 27, 1887, pp. 384, 386 (distribution of the sacculi)—von Grarr, Challenger Reports, Zoology, vol. 20, pt. 61, 1887, p. 6 (Challenger station 186; myzostomes)—P. H. Carpenter, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 97 (description; Challenger station 186; Port Molle; Nicol Bay), pl. 37, figs. 4-6; Journ. Linn. Soe. (Zool.), vol. 21, 1889, p. 807 (disk compared with that of [Pontiometra] andersoni)—WattuER, Einleitung in die Geologie als historische Wiss., 1894, p. 298 (from Carpenter).—D6prEriEIn, Denkschr. med.-nat. Ges. Jena, vol. 8, pt. 5, 1898, p. 476 (Thursday Island); pl. 36, figs. 2, 2a—Mrtnckert, Archiv fiir Naturg., Jahrg. 71, vol. 1, 1905, Heft 1, pp. 219 ff. (regeneration) Hamann, Bronns Klassen und Ordnungen des Tier- Reichs, vol. 2, Abt. 3, 1907, p. 1577 (listed) —A. H. Cuarx, Mem. Australian Mus., vol. 4, 1911, pp. 714, 716, 720 (identity; discussion); Smithsonian Misc. Coll., vol. 60, No. 10, 1912 (includes Antedon multiradiata P. H. Carpenter); Crinoids of the Indian Ocean, 1912, pp. 31, 33, 39 (identity). Antedon sp. P. H. Carpenter, Proc. Zool. Soc. London, 1882 (1883), p. 743 (with 2 distichals the axillary not a syzygy but the first 2 brachials united by syzygy); Journ. Linn. Soc. (Zool.), vol. 24, 1891, p. 45 (Hyponome sarsii a detached disk). Antedon multiradiata voN Grarr, Challenger Reports, Zoology, vol. 10, pt. 27, 1884, pp. 18, 56 (Challenger station 187; myzostomes).—P. H. Carpenter, Challenger Reports, Zoology, vol. 11, pt. 32, 1884, pp. 84, 85, pl. 55, figs. 3, 4 (disk); Quart. Journ. Mier. Sci., vol. 27, 1887, p. 386 (sacculi enormously abundant on pinnules; Cape York); Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 96 (description; Challenger station 187; Hyponome sarsii not unlikely to have been disk of this species), pl. 9, figs. 1-6.—Braun, Centralbl. fiir Bakt. und Parasitenk., vol. 3, 1888, p. 186 (myzostomes).—P. H. Carpenter, Journ. Linn. Soc. (Zool.), vol. 21, 1889, p. 307 (disk compared with that of [Pontiometra] andersoni), p. 309 (position of second syzygy).— HarTLaAvB, Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 25—Mzunckert, Archiv fiir Naturg., Jahrg. 71, 1905, vol. 1, Heft 1, pp. 221 ff. (regeneration), fig. N, p. 229, fig. O, p. 231—Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1577 (listed) —A. H. Cuarxk, Mem. Australian Mus., vol. 4, 1911, p. 716 (identity; known from Australia); Smith- sonian Misc. Coll., vol. 60, No. 10, 1912, p. 11 (synonym of microdiscus) ; Crinoids of the Indian Ocean, 1912, p. 33 (identity) —H. L. Cuarx, Biol. Results Fishing Exper. F. I. 8. Endeavour, 1909-14, vol. 4, pt. 1, 1916, p. 19 (synonym of microdiscus). Hyponeme sarsit ROLLESTON and Jackson, Forms of animal life, 1888, p. 573. Zygometra microdiscus A. H. CuarKx, Smithsonian Misc. Coll., vol. 50, 1907, p. 348 (listed); Zool. Anz., vol. 34, 1909, p. 367 (Mermaid Strait); Vid. Medd. Nat. Foren. Kgbenhavn, 1909, p. 167 (type specimen has a regenerating disk); Amer. Journ. Sci., ser. 4, vol. 32, 1911, p. 130 (sig- nificance of distinctive characters); Die Fauna Siidwest-Australiens, vol. 3, Lief. 13, 1911, p. 436 (Nicol Bay), p. 441 (Australian tropical species; range), p. 443 (range on east coast), p. 444 (range on west coast), p. 458 (new localities; summary of previous localities; notes), p. 465 (association with other species); Mem. Australian Mus., vol. 4, 1911, p. 717 (known from Australia), p. 721 (occurs south to Port Molle), p. 723 (Nicol Bay; Mermaid), p. 724 peculiar to north Australia), p. 734 (in key), p. 760 (synonymy; characters; summary of Aus- tralian records; range; type specimen has a regenerating disk which is responsible for the spe- cific name) ; Smithsonian Misc. Coll., vol. 60, No. 10, 1912, p. 11 (Cape York; detached visceral mass the type of Hyponome sarsi of which a complete synonymy is given; includes multiradiata; South Passage, Shark Bay, 9 m.; description); Proc. U. S. Nat. Mus., vol. 43, 1912, pp. 385, 84 BULLETIN 82, UNITED STATES NATIONAL MUSEUM 393 (discussion; localities); Crinoids of the Indian Ocean, 1912, pp. 31, 38, 39 (identity); p. 103 (synonymy; summary of previous records); Die Fauna Siidwest-Australiens, vol. 4, Lief. 6, 1913, p. 310 (South Passage, Shark Bay, 9 m.; description); Smithsonian Misc. Coll., vol. 61, No. 15, 1913, p. 20 (published references to specimens in British Museum; localities; characters of the specimens).—ReicuensrerGEr, Abh. Senck. Naturf. Ges., vol. 35, Heft 1, 1913, p. 82 (Aru Islands), p. 90 (Aru Islands, west of Ngaiguli, 14 m.; characters of the 2 specimens).— A. H. Crarx, Rec. Western Australian Mus., vol. 1, pt. 3, 1914, p. 123 (detailed comparison with Z. elegans; specimens from northwest Australia and from Lewis Island positively identified as this species); Internat. Rev. gesamt. Hydrobiol. und Hydrogr., 1915, p. 224 ff. (detailed account of distribution in Australia) —F. W. CruarKe and Wuee er, U. S. Geol. Surv. Prof. Pap. 90-L, 1915, p. 194 (inorganic constituents of skeleton).—H. L. Crank, Biol. Results Fish- ing Exper. F. I. S. Endeavour, 1909-14, vol. 4, pt. 1, 1916, p. 19 (new locality; notes).—Harr- MEYER, Mitt. zool. Mus. Berlin, vol. 8, Heft 2, 1916, p. 234 (W-Australien, Nos. 5966, 6136; Mermaidstrasse, No. 5607).—F. W. Cuarke and Wuperter, U.S. Geol. Surv. Prof. Pap. 102, 1917, p. 21 (Aru Islands; inorganic constituents of skeleton).—A. H. Cuarx, Unstalked crinoids of the Siboga-Exped., 1918, p. 59 (in key; range), p. 60 (synonymy; notes; station 273), p. 275 (listed), pp. 277, 278 (inorganic constituents of skeleton).—Gis~fin, Kungl. Svenska Vet. Akad. Handl., vol. 59, No. 4, 1919, p. 3 (listed), p. 5 (station 13), pp. 19, 22, 28.—A. H. Cuarx, Smithsonian! Misc. Coll., vol. 72, No. 7, 1921, p. 22 (food).—F. W. Cuarke and WHEELER, U. S. Geol. Surv. Prof. Pap. 124, 1922, p. 17 (inorganic constituents of skeleton).—GrsLin, Zool. Bidrag Uppsala, vol. 9, 1924, p. 278 (Broome, Western Australia; shore; food), figs. 142, 143, p. 95 (pinnule articulations).—A. H. Cuarx, Journ. Linn. Soe. (Zool.), vol. 36, 1929, p. 637 (Baudin Island, 8-15 fathoms).—Gis.tin, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 36, 37, 40.—H. L. Cuarx, Mem. Mus. Comp. Zool., vol. 55, 1938, p. 32 (Roebuck Bay; Broome; Lagrange Bay, 5-8 fathoms; notes). Zygometra multiradiata A. H. Cuark, Smithsonian Misc. Coll., vol. 50, 1907, p. 348 (listed); Amer. Journ. Sci., ser. 4, vol. 32, 1911, p. 130 (significance of distinctive characters); Die Fauna Siidwest-Australiens, vol. 3, Lief. 13, 1911, p. 441 (Australian tropical species occurring south to Dampier Archipelago and Cape York), p. 443 (range on east coast), p. 444 (range on west coast), p. 446 (summary of west Australian records); Mem. Australian Mus., vol. 4, 1911, p. 712 (identity with Hyponome sarsi), p. 717 (known from Australia), p. 721 (occurs south to Cape York), p. 724 (peculiar to north Australia); p. 734 (in key), p. 761 (synonymy; charac- ters; Australian records; range); Smithsonian Misc. Coll., vol. 60, No. 10, 1912, p. 11 (synonym of Z. microdiscus); Crinoids of the Indian Ocean, 1912, p. 3 (identity of Hyponome sarsi), p. 33 (synonym of microdiscus), p. 104 (synonymy; summary of known records); Rec. Western Australian Mus., vol. 1, pt. 3, 1914, p. 123 (synonym of microdiscus). Zygometra miciodiscus A. H. Cuark, Amer. Journ. Sci., ser. 4, vol. 32, p. 130 (significance of dis- tinctive characters). Zygometra mertoni ReicHENSPERGER, Abh. Senck. naturf. Ges., vo). 35, Heft 1, 1913, p. 82 (Aru Islands), p. 90 (description and comparisons; localities), fig. 6, p. 91 (proximal pinnules).— A. H. Crark, Unstalked crinoids of the Siboga-Exped., 1918, p. 59, footnote 1 (synonym of microdiscus). Zygometra microdisca H. L. Cuark, Echinoderm fauna of Torres Strait, 1921, pp. 5, 7 (history), p. 20 (range). Zygometra cf. microdiscus GisL£N, Zool. Bidrag Uppsala, vol. 9, 1924, p. 271 (Hyponome sarsi; food). Diagnostic features.—A large species with usually 65-100 arms from 85 to 145 mm. long, and the longest cirri with usually 45-70 segments; all, or nearly all, the division series beyond the IBr series are 4(3-++4); and the proximal pinnules are stout at the base and very long, 20-32 mm. in length, with 25-60 segments. Description—The centrodorsal is large and prominent, varying from thick dis- coidal with a slightly convex dorsal pole about 3 mm. in diameter to more or less hemispherical. The cirrus sockets are arranged in two to four (usually two or three) irregular marginal rows. A MONOGRAPH OF THE EXISTING CRINOIDS 85 The cirri are XVII-L (usually XXXV-XLII), 30-70 (usually 45-55), 25-55 (usually 35-45) mm. long, long and stout. The longest segments, in the proximal half of the cirri, vary from slightly broader than long to slightly longer than broad, being usually about as long as broad, and those beyond the twelfth-twentieth, which is usually a more or less marked transition segment, are much broader than long. The segments beyond the transition segment bear dorsal spines with glassy tips, which are commonly blunt, though well marked, but may be either small or prominent and conspicuous. The opposing spine is strong and sharp, directed obliquely distally, usually, but not always, more conspicuous than the dorsal spines preceding; its height is equal to about half the width of the penultimate segment. The terminal claw is usually half again as long as the penultimate segment and is moderately curved. The distal edge of the radials may be just visible beyond the rim of the centro- dorsal, or the radials may be almost completely concealed, only their anterolateral angles being visible in the interradial angles. The IBr,; are very short, six to eight times as broad as long, and are either entirely free laterally or just in contact basally. The IIBr series, IIIBr series, and IVBr series are 4(3+-4), rarely or exceptionally 2, most frequently the IIIBr series. ‘The IVBr series are usually developed internally in reference to the II[Br series. The postradial series are strongly convex dorsally and are more or less compressed laterally, often having a fairly distinct ventrolateral rim. The arms are 40-100 (usually 50-90) in number and are 60-150 (averaging about 110) mm. long. They are smooth, strongly convex dorsally, and laterally com- pressed. The brachials are at first short and oblong, then short and almost or quite triangular, and later short wedge-shaped, becoming squarish toward the arm tips. Syzygies occur between brachials 3-++4, again from between brachials 234-24 to between brachials 39+40 and from between brachials 42-++-43 to between brachials 56+57, and distally at intervals of 8-20 (usually 11-15) muscular articulations. Sometimes there is an additional syzygy from between brachials 8+9 to between brachials 13-+-14. The proximal pinnules are usually exceedingly long with broad and slightly keeled basal segments, more rarely of moderate length. P, and P, are long and stout, P, being often rather longer than P,. P, may reach 32 mm. in length. P, and P, are composed of 40-60 (rarely as few as 25) segments of which the basal are broad, but not especially distinguished, and those following diminish in size and gradually develop a projection of the dorsal edge at the distal end that disappears in the smaller terminal segments. PP, is more or less smaller than P,, 18-25 mm. long with 25-56 (usually 30-50) segments of which the basal are broad, about as long as broad, and the distal are half again as long as broad. The proximal portion of the pinnule up to the fifth segment tapers slightly, the pinnule from this point onward becoming more rapidly flagellate. P, is usually more or less considerably smaller than P;, but it may be of the same size or even slightly larger. It is 20-25 mm. long, with 37-56 segments of which the basal are a little flattened. P; resembles P; and P, but is shorter and smaller, up to 20 mm. long with 25-38 seg- ments. P, resembles P3; but is smaller and shorter, up to 13 mm. long with 26 seg- ments. P; is 10 mm. long with 24 segments, and Py) is 8 mm. long with 15 segments. The distal pinnules are 8 mm. long with about 20 segments. 86 BULLETIN 82, UNITED STATES NATIONAL MUSEUM The disk is 12-15 mm. in diameter and is strongly incised. Usually the disk ambulacra are protected by a well-developed calcareous plating that ceases at the arm bases. The anal tube is more or less completely covered with coarse calcareous concretions. The other interambulacral areas may contain more or less numerous plates, or they may be naked. Color in life —Dr. H. L. Clark says that the color is primarily a reddish purple or even deep claret with the dorsal surface of the arms more or less yellow, but the amount of yellow varies enormously; at one extreme are individuals that are practi- cally all purple, really unicolor, and they are not rare, while at the other are those in which the yellow predominates to such an extent that only the tips of the cirri and more or less of each pinnule distally are purple. The most beautiful specimens are yellow or nearly white, finely speckled more or less profusely with purple. In preserved material, even that which has been very carefully prepared, the yellow shades become buff or light brown, and the purple becomes dingy and often distinctly brown. Natu- rally such museum material fails to give any adequate idea of the beauty of the living animal. The largest specimen taken by Dr. Clark, from Roebuck Bay, was deep claret with the dorsal side of the arms yellow; some of the young arms were tipped with white. As preserved it is a deep purple, and only distally do the arms show that they were dorsally yellow in life; they are now a light brown dorsally near the tips. The cirri are dark purple, a few of them cream color near the base. Notes.—One of the two specimens from northwest of Heirisson Prong, Shark Bay, is the finest example I have ever seen of this species. It has about 100 arms, which are 130 mm. in length. The IIIBr series are mostly 2, but all the other divi- sion series are 4 (3+4). The cirri are about XX XV, 47—54, from 50 mm. to 55 mm. long; dorsal spines begin to develop from the seventeenth to the twentieth segment. The color is deep purple. The other specimen is rather small, with about 65 arms which are about 85 mm. long. The IVBr series are usually developed internally in reference to the I1IBr series. The cirri are composed of 37-45 segments and are from 25 mm. to 30 mm. long. The specimen from South Passage, Shark Bay, has 50 arms 140 mm. long; the cirri are about XXXYV, 30-31, stout, 25 mm. in length. One IIBr, in a IIBr 4(3+4) series bears instead of a pinnule a slightly undersized arm the first division series of which consists of 5 ossicles all apparently united by synarthry and none bearing pinnules. This division series carries two IVBr series, both of which are 4(3+4). The arms have a knotty and irregular appearance which is probably due to parasitiza- tion. P, is from 23 mm. to 25 mm. long, very stout basally and tapering gradually to a delicate tip, and is composed of from 25 to 27 segments. Dr. H. L. Clark collected 20 specimens of this species at Broome and 5 at Lagrange Bay of which, he said, the most interesting are the largest and the smallest. The smallest, from Lagrange Bay, has 19 arms 20-25 mm. long; the 12 division series beyond the IBr series (which would give 22 arms) are all 4(3+4); the cirri are XIII, 25-29, relatively very long, more than half as long as the arms. The color is uni- formly deep purple. Dr. Clark says that in spite of its small size it is quite a typical microdiscus. The largest specimen, in its present dry condition very nearly 300 A MONOGRAPH OF THE EXISTING CRINOIDS 87 mm. across, was taken in Roebuck Bay. Dr. Clark says it has about 110 arms, but it is impossible to count the exact number without serious damage to the specimen. The cirri are XLIII, 52-70, from 50 to 55 mm. long; one of the cirri is forked at the tip. In the specimen from Mjéberg’s station 13, as described by Gislén, the centro- dorsal is flattened with the bare dorsal pole 3 mm. in diameter. The cirrus sockets are arranged in three or four irregular rows. There are 42 cirrus sockets of which 17 carry large cirri, 16 carry regenerating young cirri, and the rest are empty. The cirri are XVII, 48-51, 40-50 mm. long. From the sixteenth to the twenty- fourth segment onward there is a small distal dorsal tubercle or blunt spine. The first segment is very short, about half as long as the second, and those following increase in length to about the twelfth, which is somewhat longer than broad, and then decrease in length. The blunt dorsal spines, which are directed obliquely for- ward, are in height equal to about one-fifth the width of the segments that bear them. The opposing spine is in length equal to about half the width of the pen- ultimate segment, and is directly obliquely distally. The terminal claw is half again as long as the penultimate segment, slightly curved, and white like the dorsal spines. The radials are almost completely concealed, being visible only in the interradial angles. The IBr, are very short, seven times as broad as long, and are in contact basally. The IBr, (axillaries) are broadly pentagonal, about three times as broad as long. The IIBr series are 4(3+-4), except for a single one which is 2. Of the 18 IIIBr series present 11 are 4(3+-4) and 7 are 2. The nine IVBr series are all 4(3+-4). The division series and first two brachials are laterally flattened with a fairly distinct ventrolateral edge. The arms are between 49 and 60 in number and are about 130 mm. long. They are a little compressed laterally. On the undivided arms the first syzygy is between brachials 3-++4, the second is from brachials 30-++31 to brachials 39-+40, the third is from brachials 49+-50 to brachials 56-++57, and the fourth is from brachials 61-+62 to brachials 64-65. On one arm the syzygies are between brachials 3+4, 8+9, 26+27 and 39+40, and on another between brachials 3++-4, 13+-14, 30+31 and 49-50. But in general the animal shows fairly definite zones of syzygies crossing all the arms. The anal tube is encrusted with coarse calcarous concretions. P, is 20 mm. to 22 mm. long and is composed of 39-48 segments. The basal segments are about as long as broad, and the distal are half again as long as broad. P, is 20 mm. to 23 mm. long with 36-43 segments; as in P, the basal segments are a little flattened. P,; is 20 mm. long with 38 segments. P, is 13 mm. long with 26 segments. P; is 10 mm. long with 24 segments. Py) is 8 mm. long with 15 segments. The distal pinnules are 8 mm. long with about 20 segments. The color in life, ac- cording to a note by Dr. Mjéberg, was yellowish white both dorsally and ventrally, the pinnules dark with small rectangular whitish spots. The proximal segments of P, have lateral black spots. In alcohol the specimen is yellow-brown, the arms being chocolate-brown ventrally. The segments of the distal pinnules have dark central bands and white ends so that the pinnules as a whole suggest a string of beads. From Holothuria Bank I have seen one fine specimen and fragments of others. Bell mentioned three specimens from Nicol Bay, which he found in the British Museum collection and which he said were smaller than the type specimen from Port Molle. He noted that the smallest of these has not more than XXX cirri, which 88 BULLETIN 82, UNITED STATES NATIONAL MUSEUM do not have more than 40 segments. The dorsal spines, especially the opposing spine, are better developed than in the type. There are about 50 arms, and in some cases the IIIBr series are 2. The ground color is purplish, marked with yellow bands. I have examined two of these specimens. The specimen collected by the Gazelle in Mermaid Strait may be described as follows: The centrodorsal is thick discoidal with a rather small flat bare polar area. The cirrus sockets are arranged in three closely crowded alternating marginal rows. The cirri are XXV, 51-60, 40 mm. long. The first segment is very short, four or more times as broad as long, those following increasing very gradually in length to the sixth or seventh, after which they remain uniform, usually slightly broader than long, sometimes about as long as broad, to the seventeenth to twenty-fourth. They then decrease in length, soon becoming about twice as broad as long and so remaining to the end of the cirrus. The penultimate segment is rather longer than those pre- ceding, about as long as broad. Up to a point between the seventeenth and twenty- fourth segments the cirri are without dorsal processes, and the segments have a dull surface; there is an indication of a median constriction on each segment. Beyond this point the segments become laterally compressed and develop subterminal dorsal tubercles or small spines which do not, however, become very prominent. The opposing spine is much larger than the dorsal processes preceding. It arises from the entire dorsal surface of the penultimate segment, is about as long as the distal width of that segment, and is directed slightly forward. The terminal claw is about half again as long as the penultimate segment and is rather slender and rather slightly curved. The disk and the brachial perisome to the last axillary are completely covered with a pavement of small regular plates. The radials are even with the edge of the centrodorsal. The IBr, are exceedingly short, regularly oblong and bandlike, basally united interiorly. The IBr, (axillaries) are very short, three to four times as broad as long, almost triangular, with the lateral edges not quite so long as those of the IBr,, laterally rounded and widely free. The IIBr series are 4(3-++-4), rounded later- ally, without marginal processes, and widely free; the IIBr, are united interiorly for about the proximal half or rather more, the free distal portion of their interior sides diverging at rather more than a right angle. The IIIBr series are always 4(3+-4) exteriorly and usually the same interiorly, though here not infrequently 2, thus suggesting the arrangement characteristic of Himerometra. The IVBr and VBr series, when present, are 4(3-+-4). There are about 85 arms approximately 120 mm. long. The first eight brachials are oblong, slightly over twice as broad as long, those following becoming slightly wedge-shaped, about twice as broad as long, gradually changing to oblong, or very nearly so, about two and one-half times as broad as long. Pp is 30 mm. long, very stout basally but tapering distally and becoming very slender and flagellate in the distal third. It is composed of about 50 segments, which at first are about as long as broad, then broader than long, distally again about as broad as long, and terminally longer than broad. After the end of the proximal third the distal ends of the segments on the dorsal side become prominent, soon developing into rather long dorsal processes which disappear at the beginning of the distal third. P» is similar to Pp, but rather larger. Ppp and P, are progressively smaller and shorter, the latter being only about 9 mm. long and very slender, without A MONOGRAPH OF THE EXISTING CRINOIDS 89 dorsal projections on the segments. The pinnules in the proximal portion of the arm distal to the oral pinnules are 4 mm. long, small and weak, with 13 segments of which the first three are about as long as broad and the remainder are somewhat longer than broad, becoming about half again as long as broad distally. The distal pinnules are 7 mm. long, slender, with about 20 segments of which the first is short, not quite twice as broad as long, the second is trapezoidal, about as long as the distal width, and the remainder are about half again as long as broad. The middle and distal pinnules are slightly flattened and subcarinate. The color is dark purple, with the cirri and large lower pinnules lighter. The details of the 14 specimens collected by the Srboga at the Aru Islands are as follows: (1) 65 arms. Color (in alcohol) yellow, with small circular spots of red-brown on the division series and arm bases; cirri yellow in the proximal half, each segment with a narrow dorsal purple saddle, in the distal half light purple. (2) 60 arms. Similar to the preceding; all the division series are 4(3+-4). Cirri 40 mm. long, with 51-52 segments; the transition segment is about the twentieth. Color purple, darkest on the cirri, with a broad mediodorsal line of yellow on each arm. (3) About 80 arms; one of the division series is 6(3+-4). Color yellow-brown. (4) About 75 arms, about 110 mm. in length. Cirri 37 mm. long. Color yellow- brown. (5) 85 arms. Color yellow-brown. (6) About 100 arms. Color yellow-brown. (7) About 80 arms. Color yellow-brown. (8) About 65 arms, about 85 mm. long. Cirri 35 mm. long and composed of 49 segments. Color light yellow, concentrically banded with narrow bands of purple; cirri deep purple with occasional large yellow blotches. (9) Similar to No. 8. (10) 85 arms, 145mm. long. Cirri 40 mm. long and composed of 52-55 segments. Color yellow-brown. (11) 80 arms, about 100 mm. long. Cirri 35 mm. to 37 mm. long and composed of 48-51 segments. Pp 32 mm. long, with about 60 segments. Color nearly white, with the cirri and the sides of the division series and arms deep purple. (12) About 85 arms. Color yellow-brown dorsally, purple ventrally. The preceding Svboga specimens are all very uniform. Division series of 2 are very rare. The proximal pinnules are typically very large and stout basally, tapering gradually and becoming flagellate distally. The distal edges of the segments in the outer half are prominent. (13) A young individual with 17 arms about 40 mm. in length. Two of the postradial series bear 2 arms each; two others bear 4 arms each, there being one IIBr 4(3+4) series bearing internally a IIIBr 2 series; the fifth postradial series bears one IIBr 4(3+4) series carrying 2 [11 Br 2 series. (14) A young specimen with 14 arms 25 mm. long. One IIBr series is present, which bears 2 IIIBr 4(3+4) series. The synarthry between the elements of the IBr series is just beginning to transform into a syzygy. 208244—40-—7 90 BULLETIN 82, UNITED STATES NATIONAL MUSEUM In the larger specimen collected by Dr. H. Merton west of Ngaiguli, Aru Islands, there are almost 90 arms with an estimated length of about 100 mm. The centro- dorsal is thick and hemispherical in form. The cirrus sockets are arranged in 2 or 3 rows. The cirri are XLII, about 50, and reach 45 mm. in length. The 20 distal cirrus segments bear prominent dorsal spines. All the division series without excep- tion are 4(3+4). P, is somewhat longer than P,, reaching 25 mm. in length; the proximal portion to the fifth segment tapers slightly, but from that point onward the pinnule becomes more rapidly flagellate. P, and P, are composed of 53-56 segments. P, and P, resemble the first two pinnules in structure but decrease in length and in the number of their component segments. The color in alcohol is whitish; on the dorsal side of the arms there is a broad deep-violet stripe that runs to the centrodorsal ; this is narrowly interrupted on almost every axillary. The smaller specimen has 54 arms. As in the preceding, the division series are without exception 4(3+4). The centrodorsal is as in the other, and bears 45 cirrus sockets. The cirri are composed of about 50 segments. P, and P, are up to 18 mm. long. As in the case of the other specimen, the disk is lacking. Dr. Merton’s other specimens from the Aru Islands were considered by Reichens- perger to represent a new species that he described under the name of Zygometra mertoni in the following terms: The centrodorsal is discoidal with a relatively large, quite flat dorsal pole about 3 mm. in diameter. The dorsal pole sometimes shows small rudiments of cirrus sockets. The cirri are arranged in one closely crowded marginal row, and there is rarely the beginnings of a second row. The normally developed cirri are XVI-XVIII, 21-23, from 10 mm. to 12 mm. long. The first and second segments are twice as broad as long, the third is slightly longer than the second, and the fourth is about as long as broad. The fifth—ninth segments are markedly longer than broad, the seventh being relatively the longest. From the tenth onward the segments slowly become more squarish. The sixteenth—nineteenth seg- ments are somewhat broader than long, but the penultimate is again as long as broad. From the eighth onward the segments bear a prominent dorsal spine. The opposing spine is stout and reaches about two-thirds the width of the penultimate segment in length. The terminal claw is strong and strongly curved. The middle cirrus seg- ments have somewhat flaring distal ends. The radials are rarely visible in the interradial angles. The IBr, are very short, about six times as broad as long, sometimes partially concealed by the centrodorsal. The IBr, (axillaries) are very short and almost triangular. The elements of the IBr series are united by pseudosyzygy. The I[Br series, which are present in most of the specimens, are strikingly irregular, being sometimes 2 and sometimes 4 (3+4). In one example two of the I/Br series are 2, and three are 4 (3-+4). A specimen with 13 arms has 2 II Br 4 (3+-4) series, and one IIBr series 5 (3+-4) in which the syzygial pair is regularly formed and is followed by a fifth ossicle as the true axillary. In a IIBr 4 (3+-4) series the third ossicle (hypozygal) is very short, half the length of the second, and the axillary is scarcely longer, almost triangular. The 11-16 arms are 45 mm. long. The first and second brachials are similar, moderately thick, and almost dis- coidal. The first syzygial pair (composed of brachials 34-4) is somewhat longer than the second brachial. The following five or six brachials are approximately oblong, three times as broad as long. Up to this point the ossicles of the division series and A MONOGRAPH OF THE EXISTING CRINOIDS 91 the lower brachials are somewhat compressed laterally. The arm now becomes more rounded, and the brachials become triangular, distally again becoming rectangular. From about the twentieth brachial onward the distal ends are rather strongly pro- duced. The first three pinnules are similar in structure, stout, almost always more strongly keeled than those succeeding, tapering evenly distally and running to a sharp point. P, and P, scarcely differ. In the largest specimen P, measures 6.5 mm. in length and is composed of 16-21 segments, which to about the seventh are broader than long, the third-sixth being about twice as broad as long; the distal segments are very slender, and are up to 3 times as long as broad. P, is 7 mm. long and is composed of 18-20 segments, of which the six lowest are from broader than long to about as long as broad, and those following become longer than broad. Ps is 5mm. long, with 15-17 segments, of which the relations are as in Py. P, is shorter, stiff, and straight, with broader proximal segments and rapidly tapering distal segments; it measures 3.3 mm. in length and is composed of 14-16 segments. PP, is the shortest pinnule on the arm; it measures from 3 mm. to 3.1 mm. in length and is composed of 10-13 segments. P, is about 3.5 mm. long, and the pinnules following gradually increase to a length of 5mm. The disk is thickly studded with calcareous granules, which vary from wart- like to pointed. The high cylindrical anal tube and the ambulacra of the disk are prominently plated. The disk measures 5-7 mm. in diameter. In Déderlein’s specimen from Thursday Island there are 51 arms. The cirri are XLII, about 44; the segments in the distal third of the cirri have dorsal spines. Syzygies occur between brachials 3+-4, again from between brachials 27-+-28 to be- tween brachials 29+30, between brachials 43+44, 55+56, and so on. P, and P» are about equal in length. P, has about 25 and P, about 37 segments. Ps is shorter than the preceding pinnules and is composed of about 25 segments. There is no plating on the disk. The color is whitish, each segment of the division series, arms, cirri, and proximal pinnules bearing a transverse stripe often broken up into spots. Only a few cirri are dark brown to the tip. Déderlein remarked that although the number of arms is less than in the specimens described by Bell and Carpenter, and the fifth forking of the postradial series mentioned by Carpenter as a specific character is not present, he has no doubt that this specimen really belongs to the present species. The detached disk from Cape York in the Hamburg Museum, which was described by Lovén under the name of Hyponome sarsi, presents no characters by which it may be distinguished from a detached disk of this species, in which, as in all the species of the family, the disk is very easily lost. On the other hand, there is no way by which a detached disk could be definitely determined as of this species. In the Challenger specimen from station 186, as described by Dr. P. H. Carpenter, the centrodorsal is relatively large with the dorsal surface slightly convex and free of cirri. The cirrus sockets are arranged in about three rows. The cirri have 40-45 segments, few or none of which are longer than broad and the distal are quite short with tolerably well-marked dorsal spines, the opposing spine being sharp and distinct. The radials are barely visible. The elements of the IBr series are united by syzygy. The postradial series divide 4 and sometimes 5 times. Except for the IBr series, all the division series are 4 (3+4). The brachials are smooth, the proximal short and nearly triangular, those after about the sixtieth becoming more oblong. Syzygies 92 BULLETIN 82, UNITED STATES NATIONAL MUSEUM oceur between brachials 3+4, again at about the thirty-first brachial, and distally at intervals of 13-15 muscular articulations. Pp and Pp are large and stout, Pp being rather the longer. They are composed of 40-50 segments of which the lower are large but not specially marked and those following diminish in size but gradually develop a projection of the dorsal edge at their distal end which disappears in the smaller terminal segments. The third and following pinnules decrease rapidly both in length and in stoutness, after which the length slowly increases again. The disk ambulacra are protected by a well-developed calcareous plating, which ceases at the arm bases. The anal tube is also considerably plated, but the other interambulacral areas are unprotected. Sacculi are very abundant on the pinnules. The color in alcohol is brownish white, with the perisome mottled with gray. The specimens from Challenger station 187 were described by Carpenter as a new species under the name Antedon multiradiata. According to Carpenter the centro- dorsal is a thick and slightly convex disk. The cirri are marginal. The cirri are XX- XXYV, 40-50+, rather long. Few or none of the segments are longer than broad, and those in the distal half have a small blunt spine projecting slightly forward. The opposing spine is strong and sharp. The radials are visible. The IBr, are short and free laterally, and are united to the IBr, by syzygy. The postradial series are quite free and may divide four times, the division series being 4(3-+-4). There are about 40 arms, composed of short and smooth triangular brachials, which become blunter and squarer toward the arm ends. Syzygies occur between brachials 3-+-4, the next anywhere from between brachials 17+18 to between brachials 46+-47, and those succeeding at intervals of 8-20 muscular articulations. Pp is of moderate length and is composed of about 25 stout segments. The pinnules following gradually decrease in size to P;, and P, and P, are considerably smaller. The succeeding pinnules increase slowly in size but never become very large. The disk is much incised and is paved with large plates between the ambulacra, which are elevated ridges with plated walls; but the plating scarcely extends beyond the level of the outermost axillary. The disk is about 15 mm. in diameter, and the arms are prob- ably about 60 mm. long. I examined these two specimens in the British Museum and found them to be small examples of the present species. Bell deseribed Antedon microdiscus, which was based upon a single specimen from Port Molle in 22 meters, in the following terms: The centrodorsal is rather large and prominent, bearing marginal cirri in 2 or 3 rows. The cirri are XXX-L, 50-70, nearly 50 mm. long. None of the cirrus segments are markedly longer than broad. As a rule the distal two-thirds bear an inconspicuous dorsal spine, and the opposing spine is hardly more conspicuous. The radials are visible. The radials and the elements of the [Br series are very short and wide, and the IBr, are not in lateral contact. The IIBr series are 4(3+4). ‘Three palmars, the axillary normally a syzygy. The arms may divide again, and of the three joints the axillary may or may not be a syzygy.” There are probably as many as 90 arms in an adult. The arms are stiff and straight and about 150 mm. in length. The earlier brachials have fairly even edges and are well rounded dorsally and flattened laterally. Those suc- ceeding are faintly wedge-shaped, the distal edge of each projecting alternately on each side into a slight protuberance. The arms generally, though slender, are very firm and stiff and are set very close to one another. Syzygies occur between brachials A MONOGRAPH OF THE EXISTING CRINOIDS 93 3+-4, again from between brachials 23-+24 to between brachials 26-27, and from between brachials 42-++-43 to between brachials 43+44, and distally at intervals of from 11 to 13 muscular articulations. The earlier pinnules are exceedingly long in the adult, with very stout and slightly keeled basal segments. 2, which is a good deal longer than P;, has as many as 50 segments and is rather fine at its free end. The more distal segments are provided with a spine or tuft of spines. The disk, which has rounded incisions, is about 12 mm. in diameter. The disk, and the arms as far as their last division, are largely washed with purple. The middle line of the arms is lighter, but patches or spots of purple are to be found at the sides. The ventral surface is a little lighter on the disk than on the arms, where it is almost black. Bell figured this specimen in dorsal view. There are 89 arms, and all the division series beyond the IBr series appear to be 4(3+4). The division series of three ossicles with “‘the axillary not a syzygy’’ mentioned by Professor Bell are undoubtedly 4(3-+4) series in which the syzygy between the two outer elements was overlooked. Bell figured a cirrus with 59 segments of which the longest are at least half again as broad aslong. His figure of the ventral surface, which is to a certain extent schematic, shows a disk in process of regeneration. Bell did not mention the syzygial union of the elements of the IBr series, which he could scarcely have suspected unless one of them had been broken. But his figure shows the extremely short IBr, with a straight distal edge which we have since learned to associate with such a union. I have examined Bell’s type in the British Museum. The disk is in the early stages of regeneration. The enormously long proximal pinnules, which recall those of such species of Himerometra as H. bartschi, and the very long cirri, which are 40-45 mm. in length, are the characteristic features of the species. The specimen recorded by Dr. H. L. Clark from off Port Curtis is small, with only 22 arms. Of the eight IIBr series present, seven are 4(3+-4) and one is 2. There are four IIIBr series, all 4(3+4). The centrodorsal is low and conical. The cirri are XXV-XXX, 40+. P, has 27 segments and is relatively large but less than 20 mm. long. The color, dry, is a uniformly pale brown. Abnormal specimens.—In the specimen from South Passage, Shark Bay, one IIBr, in a IIBr 4(3+-4) series bears, instead of a pinnule, a slightly undersized arm the first division series of which consists of five ossicles, all apparently united by synarthry, and none bearing pinnules. This division series carries two IVBr series, both of which are 4(3+4). The arms have a.knotty and irregular appearance, probably due to parasitization. In an example with about 80 arms from Siboga station 273 one of the division series is 6(3-+4). Remarks.—Zygometra microdiscus and Z. elegans are very much alike, and whether they should be regarded as two species or as merely two incompletely differ- entiated forms of the same species is more or less a matter of personal opinion. The relation between them is of much the same order as the relation between Comatula solaris and C. cratera, between Comatula pectinata and C. purpurea, or between Comanthus timorensis and C. parvicirra. While most specimens may at once be referred either to one or to the other, some are undeniably intermediate and would fit equally well in either. 94 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Excluding obviously young individuals, specimens of Z. microdiscus have 65-100 arms, the average number being 71. The arms are 85-145 mm. long, averaging 118 mm. in length. In Z. elegans the arms are 26-80, though almost always between 35 and 45; the average number is 41. The length is 70-110 mm., averaging 99 mm. The number of the cirri is about the same in both forms. In Z. microdiscus the number of cirrus segments is 30-70, usually varying between 45 and 55; the average number in fully developed cirri is 47. The length of the cirri is 37-55 mm., averaging 42 mm. In Z. elegans the number of cirrus segments is 30-56, usually varying between 35 and 45; the average number in fully developed cirri is 42. The length of the cirri is 25-38 mm., averaging 32 mm. In Z. microdiscus P, is 20-32 mm. in length, averaging 24 mm., and is composed of 25-60 segments, the average number being 44. In Z. elegans Pp is 14-18 mm. long, averaging 16 mm., and is composed of 27-40 segments, averaging 33. There can be no reasonable doubt that Hyponome sarsi is the detached visceral mass of Zygometra microdiscus. It is identical with the visceral mass of this species only, and Zygometra microdiscus is common in the type locality of Hyponome sarsi. Since Antedon microdiscus of Bell, 1882, is the same species as that presented by the fragment described as Hyponome sarsi by Lovén in 1869, the proper course would seem to be to recognize the species as Hyponome sarsi and to reduce Antedon micro- discus to the status of a synonym. But while there can be no reasonable doubt of the identity of Hyponome sarsi and Antedon microdiscus at the same time no wholly conclusive proof of such identity is possible. So it has seemed best to retain the name Zygometra microdiscus for this species, admitting that so far as can be deter- mined it is quite the same as the species represented by the fragment described as Hyponome sarsi, and admitting also that complete and definite proof of such identity is lacking. Localities —Hamburg Southwest Australian Expedition station 16; Shark Bay, Western Australia, northwest of Heirisson Prong; 11-12.5 meters; bottom rocky, with coral; September 13, 1905 [A. H. Clark, 1911] (2, Berl. M., 5966, 6136). South Passage, Shark Bay; 9 meters; Hamburg Southwest Australian Expedi- tion, June 16, 1905 [A. H. Clark, 1912, 1913] (1, H. M.). Lewis Island, Dampier Archipelago, Western Australia [A. H. Clark, 1913] (2, B. M.). Lagrange Bay; 9-15 meters; H. L. Clark, September 1929 [H. L. Clark, 1938]. Broome; H. L. Clark, August and September 1929 and June 1932 [H. L. Clark, 1938]. Mjéberg’s station 13; Broome, Western Australia; on the beach at low tide; July 27, 1911 [Gislén, 1919, 1924]. Baudin Island, northwestern Australia; 15-27 meters [A. H. Clark, 1929] (2, B. M.). Holothuria Bank [A. H. Clark, 1913] (1+, B. M.). Nicol Bay, northwestern Australia; H. M. S. Alert [Bell, 1884; P. H. Carpenter, 1888; A. H. Clark, 1911, 1913] (2, B. M.). aoa Strait; Gazelle [A. H. Clark, 1909, 1911; Hartmeyer, 1916] (1, Berl. M., 5607). A MONOGRAPH OF THE EXISTING CRINOIDS 95 Siboga station 273; anchorage off Pulu Jedan, east coast of the Aru Islands (pearl banks); 13 meters; sand and shells; December 23-26, 1899 [A. H. Clark, 1918] (14, U. S. N. M., E. 448; Amsterdam Mus.). Aru Islands; west of Ngaiguli; 14 meters; coarse yellow sand; Dr. H. Merton, February 18, 1908 [Reichensperger, 1913]. Aru Islands; between Batu Kapal and Meriri; 10 meters; Dr. H. Merton, March 30, 1908 [Reichensperger, 1913]. Aru Islands, near Udjir; 10-14 meters; coral rock and sand; Dr. H. Merton, April 16, 1908 [Reichensperger, 1913). Near Lola, northern Penambulai; 5 meters; Dr. H. Merton [Reichensperger, 1913). Thursday Island; Prof. Richard Semon [Déderlein, 1898]. Albany Island; H. M. S. Alert [A. H. Clark, 1913] (1, B. M.). Cape York, Torres Strait [Lovén, 1869; Liitken, 1869; Billings, 1869; Wyville Thomson, 1871, 1872; P. H. Carpenter, 1879, 1883, 1884, 1888, 1891; Wachsmuth and Springer, 1881; Perrier, 1883; Lockington, 1884; Hartlaub, 1891; Neviani, 1891; A. H. Clark, 1907, 1911, 1912, 1921] (detached disk, H. M.). Cape York [P. H. Carpenter, 1887; refers to Challenger station 187]. Torres Straits; 18 meters; sand [A. H. Clark, 1913] (1, B. M.). Challenger station 186; Prince of Wales channel (lat. 10°30’ S., long. 142°18’ E.); 15 meters; coral mud; September 8, 1874 [von Graff, 1887; P. H. Carpenter, 1888]. Challenger station 187; off Booby Island (lat. 10°36’ S., long. PATS 5 BES) se V1 meters; coral mud; September 9, 1874 [von Graff, 1884; P. H. Carpenter, 1884, 1888] (2, B. M.). Somerset Passage; 9-16 meters [A. H. Clark, 1913] (1, B. M.). Port Molle, Queensland; 22 meters; H. M. S. Alert [Bell, 1884; P. H. Carpenter, 1888; A. H. Clark, 1911, 1913] (1, B. M.). Thirteen miles northeast of North Reef, Capricorn group, off Port Curtis, Queensland; 128-135 meters; Endeavour [H. L. Clark, 1916] (1). Geographical range-—Northern Australia south to Shark Bay on the west and Port Curtis on the east, and the Aru Islands. Bathymetrical range.—Littoral and sublittoral, descending to 128 (?135) meters. Occurrence.—Dr. H. L. Clark says that this truly magnificent comatulid is by no means rare in the Broome region, but it is not so common as either Z. elegans or Z. comata and prefers deeper water. Most of the specimens were dredged in 9-15 meters, but a very few were found on the hard sandy bottom of Roebuck Bay during the extreme low tides of September 1929. History —Few zoological announcements have created more general interest than the description by Prof. Sven Lovén in 1869 of a living cystidean from Cape York, which he called Hyponome sarsi. But the interest in the new cystidean was short lived, for in 1872 Prof. Wyville Thomson wrote that ‘““Hyponome sarsii appears, from Professor Lovén’s description, to be a true crinoid, closely allied to Antedon, and does not seem in any way to resemble the Cystideans.” In 1879 Dr. P. H. Carpenter wrote that the voyage of the Challenger had settled the question of Hyponome sarsii, which was nothing more than the detached disk of a species of comatulid in which the disk is plated. In 1884 Carpenter said that Hyponome sarsii is “nothing more than 96 BULLETIN 82, UNITED STATES NATIONAL MUSEUM the much plated visceral mass of an Antedon common at Cape York” and referred to two figures (pl. 55, figs. 3, 4) showing two visceral masses of Antedon multiradiata |=Zygometra microdiscus|. Some pages further on he said that the original of Hypo- nome sarsii was the disk of a plated Antedon, ‘very probably of this species, Antedon multiradiata.”” In his memoir on the genus Actinometra published in 1879 Carpenter remarked that we are as yet unacquainted with any comatulids in which the elements of the IBr series are united by syzygy and there are more than 10 arms. This remark is followed by a reference to a footnote, and in the footnote he said that in the Chal- lenger collection there are three comatulids that answer to this description. One of these three is the present species. In 1882 Prof. F. Jeffrey Bell published a specific formula for a new species he called Antedon microdiscus. In 1884 he described and figured this new species, which had been obtained by the Alert at Port Molle, and mentioned a specimen in the collection of the British Museum from Nicol Bay. In 1884 Carpenter noted that in some species of Antedon with an incised disk the anambulacral plates are somewhat squamous, with a tendency to overlap one another. He figured two disks illustrating this tend- ency which had been dredged at Cape York as isolated disks. He said that these disks probably belong to Antedon multiradiata, entire individuals of which (unde- scribed) species had been dredged together with them. He went on to say that in these disks— The edges of the interpalmar areas rise rather sharply toward the ambulacra, which are marked by strong ridges with indications of a median groove visible upon their upper surface. The food- groove beneath is really comparatively deep, with its edges plated somewhat regularly and turned in towards one another. This is very marked in the immediate neighborhood of the peristome, which is thus completely closed. It is concealed in the smaller specimen by the large and prominent anal tube which projects forwards over it. In 1884 Prof. Ludwig von Graff described the myzostomes that Carpenter had found upon the specimens from Challenger station 187 and had sent to him, giving Antedon multiradiata as the name of the host. In 1887 Carpenter discussed the distribution of the sacculi in Antedon microdiscus and in his new species A. multiradiata from Cape York, and in the same year von Graff recorded the myzostomes that Carpenter had found upon a specimen of A. microdiscus from Challenger station 186. In 1888 Carpenter described Antedon multiradiata in detail from two imperfect specimens and several isolated disks, which had been dredged by the Challenger at station 187. In the remarks under this species he said that Hyponome sarsii is nothing but one of these Antedon disks covered with a well-developed calcareous plating, both at the sides of the ambulacra and in the interambulacral regions. He said that it is not unlikely to have been the disk of Antedon multiradiata which was dredged in this condition at Challenger station 187, as it has a more extensive plating than the disk of Antedon microdiscus. At the same time Carpenter described a specimen of Antedon microdiscus that had been dredged by the Challenger at station 186. This was the only specimen of this species dredged by the Challenger, and it did not come into Carpenter’s hands with the rest of the Challenger collection, for it had been given by Sir Wyville Thomson to the National Museum at Stockholm where Carpenter found it during his visit to that museum in August 1886. Later Professor Lovén kindly sent it to him in England for A MONOGRAPH OF THE EXISTING CRINOIDS 97 further examination. Carpenter pointed out that in Bel]’s description of Antedon microdiscus he overlooked the syzygy between the elements of the IBr series and also made no mention of any axillaries beyond the IVBr series, although such must be present to bring the number of arms up to 90, the number he describes in the adult while several IVBr series are represented in his figure. Carpenter said that it is the presence of the fourth axillary above the [Br axillary that is one of the characters dis- tinguishing A. microdiscus from A. multiradiata and that he had not seen any specimen without it, though it is much more frequent in the individual from Port Molle than in those from Nicol Bay and Torres Straits. Carpenter said that these last resemble one another in having a smaller number of cirrus segments and a better-developed opposing spine than the type. Carpenter remarked that the segments of the cirrus figured by Bell are much broader than long, whereas in the Challenger specimen this is only the case in the outer part of the cirrus, some of the proximal segments being as long as, or longer than, broad, and in premature cirri the length is distinctly greater than the width, while the opposing spine is especially prominent. Carpenter noted that Bell described the second or palmar pinnule of his type specimen as being a good deal longer than the first or distichal one. But he found that this is not the case in the specimen dredged by the Challenger. Furthermore, the pinnules show no trace of the slightly keeled basal segments described by Bell. But the distal edges of the basal segments are somewhat sharp, and beyond the sixth segment they project slightly over the bases of those succeeding. This feature gradually develops into a blunt slightly spinous process, which is most marked at about the fifteenth segment and dis- appears altogether after the twenty-fifth; but in the palmar pinnule figured by Bell it is not visible until the eighteenth segment and continues until near the end of the pinnule. Carpenter said that it is this feature apparently which led Bell to say that “the more distal joints are provided with a spine or tuft of spines.” In 1898 Prof. Ludwig Déderlein recorded and gave notes upon a specimen from Thursday Island. In 1909 I recorded a specimen that had been dredged by the Gazelle in Mermaid Strait, and in 1911 I recorded two specimens from Shark Bay that had been collected by the Hamburg Southwest Australian Expedition in 1905, and also others from Lewis Island in the Dampier Archipelago, from Holothuria bank, and from north- western Australia, which I had seen in the British Museum. After an examination of the type specimens of Antedon microdiscus and of Antedon multiradiata in the British Museum I, in a memoir on the crinoids of the Hamburg Museum published in 1912, announced that they really represent the same species, I had also examined the type of Hyponome sarsi, which is in the Hamburg Museum. and this I placed, with a query, in the synonymy of Zygometra microdiscus. A speci- men from another locality in Shark Bay was recorded and described. In a paper on the crinoids of the Berlin Museum published in 1912, I listed the specimens of this species in that institution. These are the two from Shark Bay recorded and described in 1911, and the Gazelle specimen from Mermaid Strait. In a memoir on the crinoids of the British Museum published in 1913, I gave notes upon the 10 specimens of this species in its collection and stated that the two specimens that served Carpenter as the types of Antedon multiradiata are simply small examples of the species represented by the type of Bell’s Antedon microdiscus. In a 98 BULLETIN 82, UNITED STATES NATIONAL MUSEUM supplementary paper on the crinoids of the Hamburg Southwest Australian Expedi- tion published in 1913 I again recorded and described the specimen from Shark Bay in the Hamburg Museum. Dr. August Reichensperger in 1913 recorded and described two specimens from the Aru Islands that had been collected by Dr. H. Merton, and at the same time published a detailed description of a new species, which he called Zygometra mertoni. In a paper on the crinoids collected by the Endeavour off the coast of Western Australia published in 1914, I remarked that from an examination of the types in London I believed Antedon multiradiata to be based upon specimens of Antedon microdiscus which had not attained full development, and said further that I would now refer to Zygometra microdiscus the specimens I had recorded from northwestern Australia and from Lewis Island in the Dampier Archipelago. In 1915 Prof. Frank W. Clarke and W. C. Wheeler, of the U.S. Geological Survey, published an analysis of the inorganic constituents of the skeleton of this species, their material consisting of fragments of the specimens collected by the Siboga at the Aru Islands. In 1916 Dr. Hubert Lyman Clark recorded and gave notes upon a specimen dredged by the Endeavour off Port Curtis, and in the same year Dr. Robert Hartmeyer again listed the specimens in the Berlin Museum and gave their catalogue numbers. Inmy report upon the crinoids of the Siboga Expedition published in 1918, [recorded and gave notes upon a series of 14 specimens collected in the Aru Islands. From the characters presented by young individuals in this series it seemed clear that Reichen- sperger’s Zygometra mertoni is merely the young of Z. microdiscus, and the former was therefore placed among the synonyms of the latter. The analysis of the inorganic constituents of the skeleton of this species was republished in an appendix.to the Siboga report. Dr. Torsten Gislén in 1919 recorded and described a specimen from Mjéberg’s station 13 and at the same time suggested that Bell’s Antedon elegans should be re- garded as merely a variety of microdiscus. Dr. Hubert Lyman Clark in 1921 published a general discussion of this species based on published records, and in 1924 Dr. Gislén described in detail the pinnule articulations. In 1929 I recorded a specimen from Baudin Island. In 1938 Dr. H. L. Clark recorded and gave notes on 25 specimens collected by himself at Broome and Lagrange Bay, Western Australia, in 1929 and 1932. ZYGOMETRA ELEGANS (Bell) PLATE 2, Ficure 5; Puate 3, Ficures 7, 8; Prats 4, Ficures 9-12 [See also vol. 1, pt. 2, fig. 710 (disk), p. 346.] Antedon elegans Brut, Proc. Zool. Soc. London, 1882, p. 534 (specific formula).—P. H. CARPENTER, Proc. Zool. Soc. London, 1882 (1883), p. 746 (specific formula) —Bxr.t, Report Zool. Coll. H.M.S. Alert, 1884, p. 155 (specific formula), p. 162 (description; Port Molle; Thursday Island), pl. 13, figs. B, Ba.—P. H. Carpenter, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, pp. 23, 31, 48, 52, 53, 55, 56, 90, 94, 96, 97, 130 (discussion), 264-266 (identity with fluctuans), 276, 362, 366, 367, 375, pl. 8. figs. 1, 2; Journ. Linn. Soe. (Zool.), vol. 21, 1889, p. 305 (discussion, but not the specimens from Mergui or the Philippines, which are Z. comata).—D6pERLEIN, Denkschr. med.-nat. Ges. Jena, vol. 8, pt. 5, 1898, p. 475 (Thursday Island; notes), pl. 36, fig. 1.— A MONOGRAPH OF THE EXISTING CRINOIDS 99 Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1577 (listed). —A. H. Cuarx, Mem. Australian Mus., vol. 4, 1911, p. 714 (of Alert report = Z. elegans), p. 716 (known from Australia), p. 720 (of Déderlein = Z. elegans); Crinoids of the Indian Ocean, 1912, p. 31 (of Bell, 1884 = Z. elegans), p. 33 (of P. H. Carpenter, 1888 = Z. elegans + Z. comata), p. 36 (of P. H. Carpenter, 1889 = Z. comata), p. 39 (of Déderlein, 1898 = Z. elegans). Antedon fluctuans P. H. Carpnntsr, Challenger Reports, Zoology, vol. 11, part 32, 1884, p. 280.— von Grarr, Challenger Reports, Zoology, vol. 10, pt. 27, 1884, pp. 16, 18, 57 (Challenger station 190; myzostomes); vol. 20, pt. 61, 1887, p. 7 (Torres Strait, 10 fathoms, Alert; myzostomes).— P. H. Carpenter, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 94 (description; Chal- lenger station 190; Torres Strait, Alert; notes), pp. 264-266 (identity with elegans), pl. 8, figs. 1, 2.—Bravn, Centralbl. fiir Bakt. und Parasitenk., vol. 3, 1888, p. 186 (myzostomes). Zygometra elegans A. H. Cuark, Smithsonian Misc. Coll., vol. 50, 1907, p. 348 (listed); Zool. Anz., vol. 34, 1909, p. 367 (Mermaid Strait); Amer. Journ. Sci., ser. 4, vol. 32, 1911, p. 130 (signifi- cance of distinctive characters); Die Fauna Siidwest-Australiens, vol. 3, Lief. 13, 1911, p. 441 (tropical Australian species occurring south to Perth and Port Curtis), p. 443 (range on the east coast), p. 444 (range on the west coast), p. 458 (localities), p. 466 (association with other species); Mem. Australian Mus., vol. 4, 1911, p. 717 (known from Australia), p. 722 (occurs south to Port Curtis), p. 723 (Mermaid; Baudin Island), p. 724 (peculiar to north Australia), p. 734 (in key), p. 762 (synonymy; characters; localities); Proc. U. S. Nat. Mus., vol. 43, 1912, pp. 385, 393 (localities); Crinoids of the Indian Ocean, 1912, pp. 31, 33, 39 (identity of records), p. 104 (synonymy; summary of records); Proc. Biol. Soc. Washington, vol. 26, 1913, p. 181 (comparison with Z. comata); Smithsonian Misc. Coll., vol. 61, No. 15, 1913, p. 21 (pub- lished references to specimens in the B. M.; localities; notes)—ALEXANDER, Rec. Western Australian Mus., vol. 1, pt. 3, 1914, p. 108 (between Fremantle and Geraldton).—A. H. Cuark, Rec. Western Australian Mus., vol. 1, pt. 3, 1914, p. 114 (collected by the Endeavour in Western Australia; occurrence near Perth confirmed), p. 121 (between Fremantle and Gerald- ton; characters of the specimens; comparison with related species); Internat. Rev. gesamt. Hydrobiol. und Hydrogr., 1915, pp. 224 ff. (detailed account of distribution in Australia).— H. L. Crarx, Biol. Results Fishing Exper. F. I. 8. Endeavour, 1909-14, vol. 4, part 1, 1916, p. 18 (localities; notes)—Harrmryer, Mitt. zool. Mus. Berlin, vol. 8, Heft 2, 1916, p. 2385 (Western Australia, in Hamburg Mus.; Mermaid Strait, No. 2963).—A. H. Cuarx, Unstalked erinoids of the Siboga-Exped., 1918, p. 59 (in key; range).—Gistin, Kungl. Svenska Vet.-akad. Handl., vol. 59, No. 4, 1919, p. 3 (variety of Z. microdiscus).—H. L. Cuarx, Echinoderm fauna of Torres Strait, 1921, pp. 6, 7 (history), p. 20 (range).—GisLin, Zool. Bidrag Uppsala, vol. 9, 1924, p. 41 (details of arms), p. 44 (reversion); Kung]. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, pp, 37, 40.—A. H. Ciarx, Temminckia, vol. 1, 1936, p. 300 (Amboina; notes).—H. L. Cuarx, Mem. Mus. Comp. Zool., vol. 55, 1938, p. 31 (off Norwest Islet, Capri- corn group, 6 fathoms; Darwin, near Shell Islands, 3-6 fathoms; Broome; notes). Zygometra microdiscus var. elegans GisLHN, Kungl. Svenska Vet.-Akad. Handl., vol. 59, No. 4, 1919, p. 8 (listed), p. 5 (Station 11), p. 19 (detailed discussion and description of specimens), pl. 1, fig. 5. Zygometra microdiscus elegans GisLhN, Zool. Bidrag Uppsala, vol. 9, 1924, p. 77 (syzygies). Diagnostic features.—A large species with usually 35-45 arms 70 to 110 mm. long, and the longest cirri with usually 35-45 segments; the IIIBr and following division series are mostly 2, the IIIBr series being commonly 2 internally and 4 (3+4) exter- nally, as in the species of Himerometra; the proximal pinnules are slender and moderate in length, P, being 14-18 mm. long with 27-40 segments. Description.—The centrodorsal varies from large and flattened hemispherical to small and rather thin discoidal, being usually large and thick discoidal with the dorsal pole slightly concave or almost flat, 2.5-5 (usually 3-5) mm. in diameter. The cirrus sockets are arranged in 2 or 3, sometimes more, crowded and irregular marginal TOws. 100 BULLETIN 82, UNITED STATES NATIONAL MUSEUM The cirri are XV-XXXV (usually XX-XXX), 30-56 (usually 36-47), from 25 to 40 (usually 30-35) mm. long and are typically rather large and stout. The first four or five segments are broader than long, the fifth-eighth or -tenth are from nearly as long as broad to somewhat longer than broad, being usually about as long as broad, and those following gradually diminish in length, soon becoming markedly broader than long. On the ninth-fourteenth (usually on the tenth or eleventh), which is a more or less evident transition segment, sharp distally directed dorsal spines appear, which reach a length of from one-third to one-half the width of the segments that bear them. The dorsal spines decrease somewhat in size distally. The opposing spine is larger and more prominent than the spines on the segments immediately preceding, and its base occupies the entire dorsal surface of the penultimate segment. The terminal claw is two or three times as long as the penultimate segment and is rather strongly curved. The cirri about the rim of the centrodorsal are about 10 mm. longer than those about the dorsal pole and consist of about 10 more segments, which first bear dorsal spines one or two segments later. The distal ends of the radials are just visible beyond the rim of the centrodorsal. The IBr, are short, four times as broad as long or even broader, and are entirely free, or sometimes just in contact, laterally. The IBr, (axillaries) are short, broadly pentagonal, not much longer than the IBr, in the median line. The IIBr series are usually all 4 (3+-4), though occasionally 2, and in rare cases as many as half of them may be 2. The IIIBr series are commonly all 2, but more or fewer of them, rarely as many as half, may be 4 (8+4). When IIIBr 4 (3+4) series are developed, they are usually external in reference to the [Br series, as in the genus Himerometra, and occasionally postradial series are found in which the outer IIIBr series are 4 (3+4) and the inner are 2; rarely most, or even all, of the postradial series are of this character. The IVBr series, when present, are 4 (3+4). They may be developed either on the outermost side of the post-radial series, or on the innermost side of each IIBr series as in Himerometra. The single VBr series reported is 2. The division series are strongly rounded dorsally, and are quite free laterally. The first ossicles immediately following each axillary are interiorly united basally. The elements of the division series and the brachials to about the tenth are markedly smooth and short. The arms are 26-80 (usually 35-40) in number and are 70 to 110 (most commonly about 100) mm. long. They are strongly convex dorsally and laterally compressed and are composed of smooth and short brachials, which sometimes are as short as in the genus Himerometra. The distal ends of the brachials usually do not overlap, but often they are slightly produced and very finely spinous. The first syzygy is between brachials 3+4, the second is anywhere from between brachials 12+13 to between brachials 53+54 (most commonly somewhere near the former position), and the distal intersyzygial interval is 7-21 muscular articulations. The elongate proximal pinnules are slender and are often considerably longer on the outermost arms of each postradial series than on the inner. P, is the longest and largest pinnule, varying from rather slender to moderately stout, long and tapering, 15-18 mm. long with 27-40 segments of which the basal are moderately stout and the terminal are small. P, is similar to P, but smaller; in one case it is 14 mm. long with 31-39 segments. P, is similar to P; but smaller, with 32-36 segments. P, is 9-12 mm. long with 19-28 segments. P, is about 8 mm. long with 18+ segments. P,is 6 mm. A MONOGRAPH OF THE EXISTING CRINOIDS 101 long with 13 segments. The pinnules followed gradually increase in length, the forty-first pinnule being 9 mm. long with 19 segments. The disk is much incised, and the interradial areas are more or less covered with rather large plates. The ambulacra as far as the last axillary are raised and strongly plated ridges, but the ambulacra of the arms and pinnules are unprotected by plates. Sacculi are abundant, especially on the pinnules, and sometimes appear on the outer ends of the plated disk ambulacra. Color in life.—Dr. H. L. Clark says that this species is, on the whole, much more varied than Z. microdiscus, combinations of light grays, buffs, fawn-color, and purple of various shades being usual, while unicolor specimens are rare. A specimen from Darwin was distinctly dull purple and white, the disk whitish or pale cream color, the pinnules tipped with white and also white on the dorsal side. Notes.—The details of the eight specimens dredged by the Endeavour between Fremantle and Geraldton are as follows: (1) 45 arms, 110 mm. in length. Of the 10 IIBr series present 9 are 4 (3+4) and oneis2. Of the 20 IIIBr series present 17 are 2 and 3 are 4 (3+4). The 5 IVBr series are all 4 (3+4); four of them are developed on the outermost side of the post- radial series, the fifth being by the side of one of these. The centrodorsal is large, thick discoidal, with the dorsal pole slightly concave and 5 mm. in diameter. The cirri are composed of 44-47 segments and are 30 to 35 mm. in length. P, is rather slender, about 15 mm. long, and is composed of 27-29 segments. (2) 40 arms, about 100mm. long. Of the 10 I[Br series 7 are 4 (3 +4) and 3 are 2. There are 16 II[Br 4 (3+4) and 2 IIIBr 2 series; two of the I1IBr series are missing. The cirri are composed of 37-42 segments and are 30 to 33 mm. long. P, is 18 mm. long, rather slender, and is composed of 37 segments. (3) 40 arms, about 100 mm. long. The 10 IIBr series are 4 (3+4). Of the 18 IITBr series 12 are 2 and 6 are 4 (3+4). The 2 IVBr series are 4 (3+4). The cirTi are composed of 38-39 segments and are about 30 mm. long. The elongate proximal pinnules are slender. (4) Two of the postradial series are missing. The 3 postradial series present each bear 8 arms, 2 IJBr series and 4 IIIBr series being present in every case. All of the IIBr series are 4 (3-+4). Nine of the IIIBr series are 2 and 3 are 4 (3 +4). (5) 38 arms. One of the I1Br series is missing. The 9 IIBr series are 4 (3+4). Of the IIIBr series 10 are 2 and 6 are 4 (3++4). (6) 36 arms, about 100 mm. in length. The 10 I[Br series are 4 (3+4). Of the IIIBr series 9 are 2 and 7 are 4(3+4). The cirri are composed of 34-42 segments and are 25 to 35 mm. long. (7) 36 arms, about 100 mm. long. Of the 10 IIBr series 5 are 4 (3-++-4) and 5 are 2. Of the IIIBr series 8 are 2 and 8 are 4 (3-+4). The cirri are composed of 36-40 seg- ments and are 25 to 30 mm. long. (8) 34 arms, about 100 mm. in length. The 10 I1Br series are 4 (3+4). The 14 IIIBr series are 2. The cirri are composed of 36-43 segments and are 25-30 mm. in length. The color of these specimens in life was recorded as very variable. The specimen probably from the vicinity of Perth has 38 arms, which are about 110 mm. long. All the IIBr series are 4 (3 +4). Of the IIIBr series 11 are 2 and 7 102 BULLETIN 82, UNITED STATES NATIONAL MUSEUM are 4 (3-+4). The cirri are XXX, 42-44, about 30 mm. long, and are typically large and stout; the dorsal spines commence on the thirteenth or fourteenth segment. The proximal pinnules are considerably larger on the outer arms than on the inner. P,, is moderately stout with 28-30 segments and measures 15-17 mm. in length. The color is purplish white, becoming deep purple on the arms. The proximal part of each segment is crossed by a deep purple band. The cirri are purple, becoming darker in the outer half. In the specimen from Mjéberg’s station 11 as described by Gislén the centro- dorsal is hemispherical. The dorsal pole, which is 3 mm. in diameter, is colored by shallow black margined pits where cirri have fallen off. The cirri are XXX, 33-45, from 25-35 mm. long. The cirrus count includes 10 sockets, from which cirri recently have become detached, and 5 young cirri. The cirri in the ventral row are 32-35 mm. long with 42-45 segments, the dorsal processes beginning on the thirteenth or four- teenth, and those in the dorsal row are 25-28 mm. long, with 33-35 segments and the dorsal processes beginning on the tenth. Intermediate cirri have 36 and 42 segments with the dorsal processes beginning on the twelfth or thirteenth. The cirrus segments are mostly broader than long; the segment preceding the one bearing the first spine is the longest and is as long as, or longer than, broad. The length of the dorsal spines is from one-third to one-half the width of the segments. The opposing spine is a little longer than the preceding spines, and its length equals about the width of the penul- timate segment; its base occupies the entire dorsal surface of that segment. The terminal claw is two or three times as long as the penultimate segment and is rather strongly curved. The dorsal spines and the terminal claw are white. The basal cirrus segments are distally white. The radials are very short and are in contact laterally. The IBr; are four times as broad as long and are laterally separated. The IBr (axillaries) are broadly pentag- onal, a little longer than the IBr, in the median line. The 9 II[Br series are 4 (3+4). Of the 15 IIIBr series 14 are 2 and one is 4 (3+4). All the 3 IVBr series present are 4 (3+4). The first ossicles following each axillary are interiorly united basally. The arms are 37+ (probably about 40) and are 105 mm. in length. The elements of the division series and the brachials to about the tenth are markedly smooth and discoidal. The brachials then become more or less wedge-shaped with produced distal edges. Syzygies occur between brachials 3+4, again from between brachials 37-+-38 to between brachials 40+41, from between brachials 58+-59 to between bra- chials 63-++64, and from between brachials 73-+-74 to between brachials 76+77. On one arm the second syzygy is between brachials 53+-54, and on another the third is between brachials 87+88. On one arm arising from a IIBr 2 series the syzygies are between brachials 3+4, 18+14, 19+20, 25+26, 33+-34 and 424-43. P, is 14 mm. long, with 31-39 segments. P, is 16 mm. long, with 32-36 segments. P, is 9-12 mm. long, with 19-28 segments. P, is about 8 mm. long, with 18+ seg- ments. P; has 13 segments. P, is 6 mm. long, with 13 segments. The pinnules following increase in length to Py, which is 9 mm. long, with 19 segments. P, is 13 mm. long, with about 25 segments. P, is of similar length. P, and Py rapidly de- crease in length. P, is the shortest, the succeeding pinnules increasing again. The disk is lacking. As preserved in alcohol the arms are pale chocolate-brown. The pinnules are dark violet, ribboned with white. The soft parts are dark. A MONOGRAPH OF THE EXISTING CRINOIDS 103 Dr. H. L. Clark, after studying 48 specimens of this species, 44 from Broome and 1 from Darwin collected by himself and 3 from the Capricorn Islands collected by Livingstone and Boardman, wrote that this species is larger but more delicate than Z. comata, with more arms and many more cirrus segments. He said that it is hard to see why Gislén treated Z. elegans as merely a variety of Z. microdiscus, but from his description he suspected that the individual he described as microdiscus was really a large elegans with the maximum number of arms and cirrus segments. He said that the large series of specimens from Broome show well the following differences between the two species. In elegans the IIIBr series are 2, with few series beyond, and those often 2; in microdiscus the IIIBr series are 4 (3+4), and the many subsequent series are almost always 4 (3-+4). The lowest pinnule in elegans is not nearly so flagellate as in microdiscus, and is smaller in every way. The color in elegans is more varied on the whole. The largest specimen studied by him had 51 arms about 130 mm. long, and the cirri were XX, 45-56. A very fine specimen from Darwin has only 36 arms, but they are nearly 150 mm. long; the cirri have about 40 segments. The smallest specimen has 31 arms about 45 mm. long; all the IIIBr series are 2; the cirri have 36 segments. The two specimens from Baudin Island are very small. The specimen from Amboina is a fine example of the species with 28 arms 160 mm. long. There are 8 IIBr 4(3+4) series, and 10 IJIBr 2 series. The cirri are XXI, 35-44, 35-40 mm. long. The four specimens from Mermaid Strait have 35-80 arms. The IIBr series are 4(3+4). The IIIBr and IVBr series are usually 2, more rarely 4(3-+4). In cases where IIIBr 4(3++4) series are developed they are usually external in relation to the IBr series as in Himerometra, and there is also usually one more axillary on the inner side of each IIBr series than on the outer, again as in Himerometra. Further- more, the brachials are exceedingly short and discoidal as in Himerometra, so that at a casual glance these specimens might very well be mistaken for examples of a species of that genus. The form from Challenger station 190 was thus described by Carpenter: The centrodorsal is a thick disk. The cirri are about XXV, 30-35. The fifth-eighth segments are much longer than broad and those following diminish in length and gradually develop a sharp forward-projecting spine, which decreases slightly in the short terminal segments, but increases again on the penultimate as the opposing spine to a strong recurved terminal claw. The radials are visible beyond the rim of the centrodorsal. The IBr, are free laterally. The postradial series are quite free and may divide four times. The IIBr series are 4(3+4). The IIIBr series and 1VBr series (when present) are usually 2. The 26 to nearly 40 arms are 40 to 45 mm. long. They are composed of short, smooth, and obliquely quadrate brachials. Syzygies occur between brachials 3+4, the next anywhere between the thirteenth and sixty-first brachials, and thence at intervals of from 7 to 21 muscular articulations. P, is long and tapering, with about 40 segments, of which the basal are tolerably stout and the terminal are small. The second, and sometimes also the fourth, brachials have similar but smaller pinnules, and those following decrease slowly in size, becom- ing long and slender again toward the ends of the arms. The disk is 8 mm. in diam- eter, much incised, with the interradial regions more or less covered with rather large 104 BULLETIN 82, UNITED STATES NATIONAL MUSEUM plates. The ambulacra as far as the last axillary are raised and strongly plated ridges, but the ambulacra of the arms and pinnules are unprotected. Sacculi are abundant, especially on the pinnules, and sometimes appear on the outer ends of the plated disk ambulacra. Carpenter said that three of the Challenger specimens agree very closely in their general characters, though the frequency of the arm divisions, and therefore the number of arms, varies considerably. All the I[Br series are 4(3+-4), nearly all the I11Br series are 2, and IVBr series are present in each individual, and are in the majority of cases 2. There was secured, however, a fourth specimen that differed from these in many points. The color of the calyx and arms is the same brownish white as in the types, but the cirri have a strong reddish-brown tint (which, according to Carpenter, prob- ably was purple during life) with white bands at the articulations, and the proximal ossicles are relatively shorter, though still longer than broad. The difference between this specimen and the other three is most apparent in the arm divisions. Of the 10 IIBr series, four are 2 and six are 4(3+-4), and of the 12 IIIBr series, six are 4(3+4) and six are 2. There are no IVBr series. The arms are somewhat more massive than in the other specimens, and the brachials, instead of being smooth and obliquely quadrate, are relatively shorter and more wedge-shaped, with a slight tendency to overlap. Bell’s record from Thursday Island is based upon a detached disk. Déderlein recorded a specimen without a disk from Thursday Island. The cirri are about XX, arranged in several closely crowded rows, and are composed of about 36 segments of which the first five are broader than long, those following are nearly as broad as long, and the distal are again broader than long. Dorsal spines are de- veloped from the eleventh segment onward. The IIIBr series are all 2. There are no IVBr series. The color is white, the pinnules spotted and banded with purple; the cirri are for the most part purple, with the basal segments white. Carpenter mentioned the small specimen from Torres Strait collected by Staff- Surgeon R. W. Coppinger of the Alert. He said that, as it was in a somewhat muti- lated condition, it was not described by Professor Bell in his report upon the Alert collections, but was put aside in anticipation of the arrival of better-preserved mate- rial. Carpenter said that it resembles the types of fluctuans from Challenger station 190 in the shape of the brachials, but it has no [VBr series. It resembles the varietal Challenger specimen in having purplish cirri with white bands. The specimen from Prince of Wales Channel is young, with the arms 20 mm. long. On one of the postradial series the original synarthry between the elements of the [Br series has not as yet become transformed into a syzygy. The specimen from the lagoon on Mast Head Island has 40 arms; the single remaining entire cirrus has 56 segments. The example from Mast Head Island has about 60 arms, which are about 70 mm. long. Slightly more division series of 4(3-++4) are present than usual. Of the two specimens from Port Denison, one is a fine example with 60 arms about 100 mm. long, and the cirri XTX, 45-48, from 35 to 40 mm. long. The I[Br series are 4(3+-4); the remaining division series are 2, or rarely, on the exterior of the A MONOGRAPH OF THE EXISTING CRINOIDS 105 postradial series, 4(3-++-4). This individual agrees with the four from Mermaid Strait, with which it was compared directly. The other specimen from Port Denison is considerably smaller, with 29 arms 65 mm. long. The IIBr series are 4 (+4) and the IIIBr series are2. The cirri are XX, 35-39. The type specimen from Port Molle was described by Professor Bell as follows: The centrodorsal is small and flattened, and bears two rows of marginal cirri. The cirri are XXV-XXX, 40, 30 mm. long, with the fifth-tenth segments rather longer than broad, and those succeeding with a short conical spine, which diminishes on the more distal ones but enlarges again somewhat as an opposing spine. The radials are just visible. The IBr, are wide and are barely in contact laterally. The IBr, (axil- laries) are comparatively short. The I[Br series are pretty long, 4(3+4). If the arms divide again the division series are generally 2, but they may be 4(3+-4). There are 30 delicate arms, which are 95 mm. long. The earlier brachials have parallel proximal and distal edges; the succeeding brachials become wedge-shaped, but the distal edges do not overlap. Still further out the brachials become shorter and project a little at the sides. Toward the free end of the arms the dorsal surface of each brachial is sharply convex. Syzygies occur between brachials 3+-4, 12-13, and 24+25 and distally at intervals of 10-14 muscular articulations. The first two pinnules are stiff and long, longer and stouter than the third and fourth. None of the following pinnules are long, but the rather more distal are the longer. The disk is deeply incised, and the margins of the rays are provided with a well-developed and character- istic caleareous plating. Owing to the incisions the disk is only 8 mn. in diameter. The arms are pinkish flesh-color dorsally, much darker ventrally. The cirri are ringed with purplish and white. In a younger specimen there are purplish spots on the arms dorsally. Bell expressly remarked that none of the endocyclic comatulids he described in his report on the Alert collection had the elements of the IBr series united by syzygy, and so when Carpenter first made a cursory examination of most of the comatulids dredged by the Alert he failed to detect the similarity between the three specimens upon which Bell based his Antedon elegans and the four Challenger specimens upon which his own Antedon fluctuans was based. He did, however, recognize his A. fluctuans in a small mutilated specimen from Torres Strait which had been considered by Bell too imperfect for description with the rest of the Alert collection. In connection with the revision of the endocyclic comatulids with a synarthry between the elements of the IBr series and the IIBr series 4(3+4) Carpenter later reexamined the three specimens of Antedon elegans from Port Molle. He found that in all three the elements of the IBr series are united by syzygy. The original description (1882) of Antedon elegans was simply a so-called specific formula in which it was indicated that the IIBr and IIIBr series are of three elements— that is, 4(3-+4). This was later revised by Bell as given above. Carpenter’s reexamination of the type specimens of Bell’s Antedon elegans con- vinced him that it is in reality the same thing as his own Antedon fluctuans. Carpenter remarked that of the three specimens obtained by the Alert at Port Molle one is very considerably different from the other two, both in color and in the 20824440 —8 LO6 BULLETIN 82, UNITED STATES NATIONAL MUSEUM amount of serration of the arms. I examined these specimens at the British Museum. They show that this species has shorter cirri than Z. microdiscus and small and weak proximal pinnules. The specimen from Port Molle in the Australian Museum is small with 18 arms, which are about 50 mm. long, and the cirri XVIII, 22-24. Together with it there are some fragments from a larger individual. Of the three specimens from Port Curtis one has the centrodorsal large, thick discoidal, with the dorsal pole very slightly concave and 5 mm. in diameter. The cirri are XV, 50-53, from 35 to 38 mm. long. There are 35 arms 80 mm. long. The IIIBr series are 4(3-+-4) exteriorly and 2 interiorly, as in the genus Himerometra. Another has the centro-dorsal discoidal with the nearly flat dorsal pole about 4 mm. in diam- eter. The cirri are XXXV, 46-48, from 35 to 40 mm. long. Series of 4(3+-4) are slightly more numerous than usual. The third specimen is small with 20 arms 60 mm. long. All the IIBr series are present. The dorsal pole of the centrodorsal is flat, 2.5 mm. in diameter. The cirri are XXIII, 20-22, 13 mm. long. The specimen from off Noosa Head has 29 arms, which are about 100 mm. long. The I1Br series are all 4(3-++4). All but one of the IIIBr series are 2. The cirri are XXVIII, 45-47. The color (dry) is fawn-color with a faint purplish cast, the articula- tions being distinctly darker than the segments. The cirri are dull purple, the bases of the cirri and the centrodorsal being pale brownish. One of the specimens from off Sandon Bluffs has 38 arms nearly 100 mm. long. The cirri are XXXV, about 54. The second has 36+ arms (several are missing), which are nearly 100 mm. long. The cirri are XXX+, 47-53. In both the IIBr series and the IVBr series, when present, are 4(3-+4). It is common to have the IlIBr series 4(3+-4), although 2 seems to be the usual number. The only VBr series noted is 2. In the first specimen the greater part of the arms is light brown, but basally all the arms are heavily shaded with purple, the articulations in particular being very dark. The cirri basally are dorsally pale brown and ventrally deep purple, but distally the purple becomes very gradually dominant until it includes both sur- faces. In the second the arms are light brown with more or less of a purplish cast; basally three of the postradial series are very deep purple, the other two being lighter. The cirri are light brown basally, becoming purple distally. Localities —Between Fremantle and Geraldton, Western Australia; Endeavour [A. H. Clark, 1914] (8, U.S. N. M., 35174; W. A. M.). ?Vicinity of Perth, Western Australia; Hamburg west Australian Expedition, 1905 [A. H. Clark, 1911] (1, H. M.). Western Australia [= ?Vicinity of Perth] [A. H. Clark, 1912; Hartmeyer, 1916]. Mjéberg’s station 11; 45 miles westsouthwest of Cape Jaubert, Western Aus- tralia; 22 meters; July 17, 1911 [Gislén, 1919]. Broome, Cable Beach; Frances L. S. Clark, September 2, 1929 [H. L. Clark, 1938]. Broome; hard sandy bottom; H. L. Clark, August and September 1929 [H. L. Clark, 1938]. Broome; dredged at various points in 5-15 meters; H. L. Clark, June 1932 [H. L. Clark, 1938]. A MONOGRAPH OF THE EXISTING CRINOIDS 107 Baudin Island, northwestern Australia; 15-27 meters [A. H. Clark, 1911, 1912, 1913] (1, B. M.). Baudin Island [A. H. Clark, 1911, 1912, 1913] (1, B. M.). Amboina; pier; 0-2 meters; Willebrord Snellius, May 6, 1930 [A. H. Clark, 1936] (1, L. M.). Darwin, near Shell Islands; 5-11 meters; on sponge and alcyonarian bottom; H. L. Clark, July 2, 1929 [H. L. Clark, 1938]. Mermaid Strait; Gazelle [A. H. Clark, 1909, 1911, 1912; Hartmeyer, 1916] (4, U.S. N. M., 35157; Berl. M., 2963). Challenger station 190; Arafura Sea (lat. 8°56’S., long. 136°05’E.); 89 meters; green mud; September 12, 1874 [von Graff, 1884; P. H. Carpenter, 1888; A. H. Clark, 1911, 1912, 1913] (8, B. M.). : Thursday Island; H. M. S. Alert [Bell, 1884; A. H. Clark, 1911]. Thursday Island; Prof. Richard Semon [Déderlein, 1898}. Torres Strait; 18 meters;sand; H. M.S. Alert, 1881 [von Graff, 1887; P. H. Car- penter, 1888; A. H. Clark, 1911, 1912, 1913] (1, B. M.). Prince of Wales Channel; 13 meters; sand [A. H. Clark, 1911, 1913] (1, B. M.). Mast Head Island; lagoon [A. H. Clark, 1911] (1, Austr. Mus.). Mast Head Island [A. H. Clark, 1911] (1, U.S. N. M., 35164). Port Denison, Queensland [A. H. Clark, 1911] (2, Austr. Mus.). Port Molle, Queensland; 22-36 meters; rock; H. M.S. Alert [Bell, 1884; A. H. Clark, 1911, 1913] (8, B. M.). Port Molle [A. H. Clark, 1911] (1, Austr. Mus.). Port Curtis, Queensland [A. H. Clark, 1911] (8, U. S. N. M., 35132; 35161; Austr. M.). Norwest Islet, Capricorn group; 11 meters; Livingstone and Boardman, Decem- ber 1930 [H. L. Clark, 1938]. Off Noosa Head, southern Queensland; 29 meters; Endeavour [H. L. Clark, 1916] (1, M. C. Z., 712). Eight miles east of Sandon Bluffs, New South Wales; 64-73 meters; Endeavour [H. L. Clark, 1916] (2, Austr. Mus.). Geographical range-—Northern Australia south to between Fremantle and Geraldton on the west and to Sandon Bluffs, New South Wales, on the east; Amboina. Bathymetrical range —From the shore line down to 89 meters; chiefly sublittoral. Occurrence.—Dr. H. L. Clark says that this is a very common comatulid at Broome, occurring under the same conditions as Z. comata. But while occurring with Z. comata on open sandy bottoms, this species also occurs among rocks and is fre- quently dredged on bottoms covered with algae, sponges, and other marine growths. History —This species was first noticed by Prof. F. Jeffrey Bell who published a specific formula for it in 1882. He described it in detail from a specimen from Port Molle and figured it in his Alert report in 1884. A detached disk from Thursday Island, which had been collected by the Alert, he said probably belonged to this species. In the Challenger report upon the stalked crinoids published in 1884 Dr. P. H. Carpenter mentioned Antedon fluctuans as an example of a comatulid having a syzygy between the elements of the IBr series and in a footnote gave a specific formula for it. 108 BULLETIN 82, UNITED STATES NATIONAL MUSEUM In 1884 Prof. Ludwig von Graff described the myzostomes from Antedon fluctuans, giving as the locality Challenger station 190, and in 1887 he recorded the myzostomes from a specimen of the same species which had been dredged by the Alert in Torres Strait. In 1888 Carpenter described Antedon fluctuans in detail and figured it on the basis of two mutilated individuals and one fragment and one example of a varietal form from Challenger station 190, and a specimen from Torres Strait taken by the Alert in 1881. After the printing off of that portion of the Challenger report that contained the description of Antedon fluctuans. Carpenter had occasion to revise the endocyclic comatulids in which the two elements of the IBr series are united by synarthry and the I/Br series are 4(3+-4). Among these, judged from the original description, was Antedon elegans. Carpenter had made a cursory examination of the greater part of the comatulids dredged by the Alert some time previously, but he had not been able to identify among them any representative of the type that ap- peared in his working list of new Challenger species under the name of Antedon fluc- tuans. Later, however, he had recognized this type in an imperfect specimen from Torres Strait, which was in too mutilated condition for description with the rest of the Alert collection. On going over the Alert collection again in August 1887, he carefully examined the three specimens of Bell’s Antedon elegans for the purpose of definitely making out the relationship between this species and others in which the I Br series are 4(3+-4), and he found that in all three of the specimens the ele- ments of the [Br series are united by syzygy, although Professor Bell had expressly noted that among the species of Antedon described in the Alert report ‘in no case is the radial axillary a syzygy.’’ Carpenter pointed out that Bell’s first formula for Antedon elegans indicated that the I/Br series are 4(8+-4), and sometimes IJIBr 4(3+-4) series are present. But in the Alert report Bell said that the I[1Br series are usually 2, but may be 4(3+4). Bell’s figured specimen has four I1IBr 2 and one I1IBr 4(8+4) series. Carpenter noted that Bell’s formula omitted any reference to IVBr series, which occur in one of his specimens. Carpenter found that Bell’s Antedon elegans is identical with Antedon fluctuans, described earlier in the Chal- lenger report. In his discussion of the species he mentioned a number of speci- mens from the Mergui Archipelago and others from the Philippines, which are in reality Z. comata, and also another species from Mergui with the same extensive plating of the disk and a syzygy between the elements of the IBr series, which is Pontiometra andersoni. In 1889 in his report upon the crinoids collected by Dr. John Anderson in the Mergui Archipelago Carpenter again affirmed the identity of his Antedon fluctuans with the previously described Antedon elegans of Professor Bell. He referred to Antedon elegans five specimens from Mergui, which, however, really represent Z. comata, and mentioned others from Professor Semper’s Philippine collection, which are also Z. comata. In 1898 Prof. Ludwig Déderlein published notes upon and figured a specimen of Antedon elegans from Thursday Island. In 1909 I recorded and published notes upon four specimens from Mermaid Strait that had been collected by the Gazelle. In the discussions I mentioned a very large series at hand from Singapore and the Philippine Islands, which I identified A MONOGRAPH OF THE EXISTING CRINOIDS 109 with Carpenter’s Antedon fluctuans, and on the strength of this identification I stated that I could not agree with Carpenter in considering fluctuans a synonym of elegans. These specimens from Singapore and the Philippines, however, represented in reality Z. comata. Later in the same year I recorded a number of specimens from Singapore as Zygometra fluctuans. I pointed out that these specimens agreed closely with Carpen- ter’s description of fluctuans, except that the brachials are oblong distally, and fur- ther that they all agree among themselves and also with Philippine specimens. I[ said that I had at first followed Carpenter’s lead and had considered them as belong- ing to elegans; but when I examined the collections made by the Gazelle in northwest- ern Australia I found that I was wrong, for in that collection there are specimens of what is undoubtedly true elegans that agree with Bell’s diagnosis and have much longer and more robust cirri with more numerous segments than any of the Singapore or Philippine specimens. I remarked that, judging from the material at hand, elegans and fluctuans are both perfectly good species, and there is not the slightest difficulty in differentiating them. The form I regarded as fluctuans, however, is in reality Z. comata. In 1911 I recorded and gave notes upon a specimen collected by the Hamburg Southwest Australian Expedition presumably in the vicinity of Perth, and in a memoir on the crinoids of Australia published in the same year I recorded specimens in the collection of the Australian Museum from Mast Head Island, Port Denison, Port Molle, and Port Curtis. In a footnote I diagnosed a new species, Zygometra comata (P. H. Carpenter, MS.), to which I referred the specimens from Singapore, the Mergui Archipelago, and the Philippines previously identified by Carpenter as elegans and by myself as fluctuans. Carpenter’s Antedon fluctuans I included in the synonymy of Zygometra elegans. In a memoir on the crinoids of the Berlin Museum published in 1912 I listed one specimen from Western Australia and four from Mermaid Strait, and in my mono- graph on the crinoids of the Indian Ocean published in the same year I gave the sy- nonymy of the species and a list of the known localities. In 1913 I listed and gave notes upon the specimens I had examined in the Brit- ish Museum in 1910. In the same year, in recording specimens of Z. comata from Hongkong, I gave a detailed comparison between this species and Z. elegans. In 19141 recorded and gave notes upon eight specimens dredged by the Australian Fisheries Investigations steamer Endeavour between Fremantle and Geraldton, West- ern Australia, and discussed the relationships between this and the other species of the genus. Dr. Hubert Lyman Clark in 1916 recorded and described three specimens dredged by the Endeavour off Sandon Bluffs, New South Wales, the southernmost locality on the east coast of Australia. In the same year Dr. Robert Hartmeyer noted that the specimen from Western Australia that in 1912 I had listed as in the Berlin Museum was now in the Hambure Museum, the only specimens in the Berlin Museum being those from Mermaid Strait of which he gave the catalogue number. In 1919 Dr. Torsten Gislén described in detail a specimen from Mjéberg’s sta- tion 11. He regarded this form simply as a variety of Z. microdiscus. 110 BULLETIN 82, UNITED STATES NATIONAL MUSEUM In 1921 Dr. Hubert Lyman Clark published an account of the occurrence of this species in northern and eastern Australia. In 1938 Dr. Clark recorded and gave notes on 45 specimens collected by himself and Mrs. Clark at Darwin and Broome in 1929 and 1932, and on three collected in the Capricorn Islands by Livingstone and Boardman in 1930. He compared this form in detail with Z. microdiscus. ZYGOMETRA ANDROMEDA A. H. Clark Zygometra andromeda A. H. Cuark, Ree. Indian Mus., vol. 7, pt. 3, 1912, p. 269 (description; “India’’) ; Rec. Western Australian Mus., vol. 1, pt. 3, 1914, p. 123 (comparison with Z. elegans); Unstalked crinoids of the Siboga-Exped., 1918, p. 59 (in key; range).—GisL£Nn, Kungl. Fysiogr. Salsk. Handl., new ser., vol. 45, No. 11, 1934, p. 27. Diagnostic features.—The cirri are XXI, 27-30 (usually 30), and there are 25 arms 50-55 mm. long. The distal border of the radials and the proximal and distal borders of the elements of the division series and to a lesser extent of the first two brachials are thickened and everted, this eversion being finely scalloped or tubercu- lated so that the edges of the ossicles appear beaded. Description —The centrodorsal is thin discoidal with the bare dorsal pole large, slightly concave and finely granular, 2.5 mm. in diameter. The cirri are XXI, 27-30 (usually the latter), 13 to15mm.inlength. The longest segments are about one-third again as broad as long. Long and sharp dorsal spines are developed from the eighth or ninth segment onward. The arms are about 25 in number, 50 to 55 mm. in length. The division series and arms resemble those of Z. comata, but the distal edges of the radials and the proximal and distal edges of the elements of the division series, and to a lesser extent of the first two brachials, are thickened and everted, this eversion being finely scalloped or tuberculated so that the edges of the ossicles appear beaded. The summit of the eversion may be smooth, but it is usually very finely spinous. The flattened lateral borders of the division series and first two brachials are very finely spinous; the dorsal surface is unmodified. The distal border of the first syzygial pair bears a row of small rounded obscure tubercles, and there is usually a similar, but less evident, row at the syzygial line. Beyond the fourth brachial the arms are smooth, resembling those of Z. comata. P; is 9.5 mm. long and is composed of 24 segments. Locality —‘‘India”’ (probably Ceylon) [A. H. Clark, 1912] (1, I.M ). Remarks.—This species is known only from the type specimen in the Indian Museum, which was described in 1912. Except in the modification of the ends of the radials and ossicles of the division series it does not appear to differ in any way from Z. comata (the following species), of which it is probably the ornate western representative. ZYGOMETRA COMATA A. H. Clark Puate 4, Ficures 13, 14; Puate 5, Figures 15-17 [See also vol. 1, pt. 1, fig. 252 (centrodorsal), p. 253; fig. 333 (cirri), p. 283; fig. 468 (centrodorsal), p. 359; pt. 2, figs. 27, 28 (centrodorsal and radial pentagon), p. 20; figs. 143, 144 (arm base), p. 83; fig. 711 (disk), p. 346; fig. 779 (ambulacral deposits), p. 366.] A MONOGRAPH OF THE EXISTING CRINOIDS 111 Antedon comata von Grarr, Challenger Reports, Zoology, vol. 20, pt. 61, 1887, p. 2 (nomen nudum; Padau Bay, Mergui Archipelago; myzostomes).—A. H. Cuarx, Proc. U.S. Nat. Mus., vol. 34, 1908, p. 465 (status of name). Antedon elegans (part) P. H. Carpenter, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 265 (Philippines; Mergui Archipelago) —Butu, Proc. Zool. Soc. London, 1888, p. 389, footnote 5 (collected by Dr. Anderson at Mergui) —P. H. Carpenter, Journ. Linn. Soc. (Zool.), vol. 21, 1889, p. 305 (detailed account; King Island, Mergui), p. 306 (Philippine variety; characters), p. 307 (compared with [Pontiometra] andersont). Zygometra elegans A. H. CuarK, Smithsonian Mise. Coll., vol. 52, 1908, p. 212 (Albatross stations 5137, 5138). Antedon fluctuans A. H. Cuarx, Zool. Anz., vol. 34, 1909, p. 367 (not as supposed by Carpenter a synonym of elegans). Zygometra fluctuans A. H. Cuarx, Vid. Medd. Naturh. Foren. Kgbenhavn, 1909, p. 151 (Singapore; description of specimens), p. 153 (not the same as elegans), p. 193 (collected at Singapore by Svend Gad). Zygometra comata A. H. Cuarx, Proc. U. 8. Nat. Mus., vol. 39, 1911, p. 537 (Albatross station 5358; number and length of arms, but no further description; also recorded from Singapore) ; Mem. Australian Mus., vol. 4, 1911, p. 734 (in key), p. 762, footnote (description; previous references; localities; differential characters); Rec. Indian Mus., vol. 7, pt. 3, 1912, p. 267 (range; related species farther west), p. 268 (compared with Z. andromeda); Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p. 12 (Hongkong); Crinoids of the Indian Ocean, 1912, p. 33 (identity), p. 36 (= Antedon elegans P. H. Carpenter, 1889), p. 104 (synonymy; localities); Proc. Biol. Soc. Wash- ington, vol. 26, 1913, p. 178 (range in eastern Asia), p. 180 (8 miles outside of Hongkong harbor, 14 fathoms; notes; comparison with Z. elegans); Rec. Western Australian Mus., vol. 1, pt. 3, 1914, p. 123 (comparison with Z. elegans); Journ. Washington Acad. Sci., vol. 5, 1915, p. 214 (Malayan species; range and its significance); Unstalked crinoids of the Siboga-Exped., 1918, p. 59 (in key; range), p. 61 (synonymy; notes; Stations 162, 179, 322), pp. 273, 274, 276 (listed) ; Journ. Linn. Soe. (Zool.), vol. 36, 1929, p. 637, pl. 44, fig. 15 (80 miles northwest of Penang, 40 fathoms; Singapore); Treubia, vol. 14, livr. 2, 1933, p. 211 (Java Sea, notes) Tu. Morten- sEN, Hong Kong Nat., Suppl. 3, 1934, p. 5 (Hongkong).—Gistin, Kungl. Fysiogr. Sask. Handl., new ser., vol. 45, No. 11, pp. 21, 26, 27; Kungl. Fysiogr. Sallsk. Lund Foérh., vol. 7, No. 1, 1936, p. 4 (French Indochina); p. 5 (range), p. 6 (range), p. 8 (Pulo Condor; notes); fig. 1, p. 9H. L. Crarx, Mem. Mus. Comp. Zool., vol. 55, 1938, p. 29 (Broome; notes). Heterometra bengalensis (part) A. H. Cuarx, Mem. Australian Mus., vol. 4, 1911, pp. 722, 723, 734, 768 (specimen from Holothuria Bank, but not specimens from Port Curtis, which are Z. punc- tata); Die Fauna Siidwest-Australiens, vol. 3, Lief. 18, 1911, pp. 440, 443, 444, 446 (record from Holothuria Bank); Crinoids of the Indian Ocean, 1912, p. 130 (record from Holothuria Bank) ; Die Fauna Siidwest-Australiens, vol. 4, Lief. 6, 1913, pp. 310, 313 (Holothuria Bank). Zygometra punctata (part) A. H. Cuark, Smithsonian Misc. Coll., vol. 61, 19138, No. 15, p. 21 (Holo- thuria Bank; 15 fathoms; notes); Die Fauna Siidwest Australiens, vol. 4, Lief. 6, 1913, p. 310 (Holothuria Bank); Rec. Western Australian Mus., vol. 1, pt. 3, 1914, p. 114 (Holothuria Bank); Internat. Rev. gesamt. Hydrobiol. und Hydrogr., 1915, pp. 224 ff. (reference to Holo- thuria Bank); Unstalked crinoids of the Siboga-Exped., 1918, p. 63 (Holothuria Bank). Diagnostic features.—The arms are typically 30 in number, there being a I1IBr 2 series on the inner derivative from each IIBr series, so that each ray bears six arms arranged in 1, 2, 2, 1 order. The arms are 40-135 mm. (but usually between 60 and 100) mm. in length. The cirri have typically 30 segments of which the outer have unusually long and slender sharp dorsal spines. The species is, however, rather variable. There may be as few as 14 or 16 or as many as 41 arms, while the longest cirri consist of anywhere between 28 and 45 segments. Zygometra comata is of more delicate build than Z. elegans, and the spines on the outer cirrus segments are much more prominent. It is rather easily mistaken for a species of Heterometra unless the articulation between the elements of the IBr series 112 BULLETIN 82, UNITED STATES NATIONAL MUSEUM is examined, when the perfectly straight and uniform fine dotted line indicating the presence of a syzygy identifies it at once. The long, slender, and sharp dorsal spines on the outer cirrus segments and the dark ventral saddle-shaped patches commonly present on the cirrus segments also serve to distinguish it at a glance from any similar species of Heterometra. Description.—The centrodorsal is discoidal, broad, with a broad, flat, or slightly concave dorsal pole 5-6 mm. in diameter. The cirri are arranged in usually two irregular and closely crowded marginal rows. The cirri are X-X XXII (usually about XX), 25-45 (usually 29-35), 15-35 (usually between 20 and 30) mm. long. The first segment is about twice as broad as long and those following increase gradually in length to the fourth or fifth, which varies from about as long as broad to one-third or even one-half again as long as broad. The three or four segments following are similar, and those succeeding gradually decrease in length so that the terminal seven or eight are nearly twice as broad as long. From the sixth to the eighth there is a more or less marked transition segment that does not decrease in width distally, and is seldom with any indication of a dark band. From the sixth—-thirteenth (usually seventh—ninth) segment onward long sharp dorsal spines are developed. The opposing spine is long and slender, and is about as long as the penultimate segment. The terminal claw is considerably longer than the penultimate segment, and is rather slender and moderately curved. The longer proximal segments have a slight central constriction. The distal ends of the radials are visible beyond the rim of the centrodorsal. They are usually plain, but in western specimens may be obscurely beaded. The postradial series resemble in general those of Z. elegans but are relatively somewhat longer. The sides of the division series, which are usually in close lateral contact, may be sharply flattened with a distinct dorsolateral line, or they may be barely in contact. The sides of the division series are usually smooth, but occasionally they bear a few irregular spinous processes. The IIBr series are usually all 4(3+4), but one or more may be 2, and in rare cases half, or even the majority, may be 2. The IIIBr series are most frequently all 2 and are usually developed only internally so that commonly the arms on each postradial series are six in number and are arranged in 1, 2, 2, 1 order; but IITBr 4(8+-4) series are not infrequent, and in rare cases half, or even the majority, of the IIIBr series may be 4(3+4). IVBr series are rarely devel- oped, but occasionally one is present, and in one case two were present in a single individual. They are 4(3+-4); but in the individual with two, one was 4(3+4) and the other was 2. The arms are 14—41 (averaging about 25) in number and are 40 to 135 (averaging about 85) mm. long. They are usually of moderate length and moderately tapering, but they may be slender and taper very slowly. The brachials are rather short, in the proximal portion of the arm wedge-shaped and distally oblong; they are some- times very short, as in the genus Himerometra. The brachials after the fourth or fifth have strongly produced and very finely spinous distal edges, making the arms very rough to the touch. The pinnules resemble those of Z. elegans but have fewer segments. The ambulacra of the disk are very strongly plated and the interambulacral areas, especially the anal area, are also more or less strongly plated. A MONOGRAPH OF THE EXISTING CRINOIDS 113 Color in life-—Dr. H. L. Clark says that in life the colors are yellow and purple, but the relative amounts of the two, their distribution, and the shades show infinite variety. For the yellow the shades range from almost white to deep buff, and for the darker color from light grayish purple to a very deep, rich, reddish purple. The color arrangements are too diversified to warrant description, but it is common to have the arms banded either broadly or narrowly; often only the pinnules are banded and the dorsal side of the arms is prettily mottled, or simply unicolor. Notes.—In Dr. John Anderson’s collection from King Island, Mergui Archipelago, there were five specimens of this species. Of these, two were labeled “sublittoral,” two were found on corals, and the youngest had the cirri coiled around a gorgonian (Plexaura). As described by Carpenter, the arms of the largest reached 100 mm. in length. Compared with the specimens of Z. elegans collected by the Challenger in the Arafura Sea and Torres Strait the earlier brachials are relatively shorter with a more wedge-shaped outline and a greater tendency to overlap. The number of arms rarely exceeds 30. There are no IVBr series, and the full complement of IIIBr series is rarely found on any ray. Carpenter noted that as a rule only the 2 inner IJ1Br series are present, so that there are 6 arms to a ray arranged in 1, 2, 2, 1 order. The IIBr series are sometimes absent. There is little tendency to any variation from the IIIBr 2 type. The rays of the largest specimen from King Island are in contact, and the elements of the IIBr series sometimes exhibit a tendency to the straight edged and wall-sided form. Two of the specimens are almost black, another is a dark reddish brown mottled with lighter patches, and one has a more uniform lighter shade of the same color. One of the two specimens from 80 miles northwest of Penang has 30 arms 90 mm. long. Of the 10 IBr series present 9 are 4(3-++4) and one is 2. Two of the I1IBr series are 4(3-+4); these are situated side by side on the IIBr 4(3 +4) series which is paired with the IIBr 2 series. There is a single [VBr 4(3+-4) series on the innermost (nearest the midradial line) branch from a III[Br 2 series. ‘The division series are in close lateral contact. The longest cirri have 29-35 segments. The transition seg- ment, which is not strongly marked, is usually the seventh or eighth. The color is brownish yellow with narrow and widely spaced deep purple cross bands on the arms. The other specimen has 14 arms 110 mm. long. All 4 of the IIBr series are 4(3+4). The cirri are XXI, 33-40, 35 mm. in length. The color is light grayish, the arms with narrow and widely spaced deep purple cross bands. Of the specimens collected by Svend Gad at Singapore one dated April 20, 1906, has 19 arms 70 mm. long. There are 8 IIBr series, all 4(3-++4), and a single IIIBr 2 series, developed internally. The brachials after the fourth or fifth have strongly produced and very finely spinous distal edges, making the arms very rough to the touch. ‘The brachials are very short, in the proximal fourth of the arm wedge-shaped and in the distal portion oblong as in Himerometra. The cirri are XX, 25-28. The first cirrus segment is about twice as broad as long and those following gradually increase in length to the fourth or fifth, which is about as long as broad. The follow- ing three or four segments are similar, and those succeeding gradually decrease in length so that the terminal seven or eight are nearly twice as broad as long. The sixth is a transition segment, its proximal half having a dull and its distal half a highly polished surface. The transition segment does not decrease in width distally, nor is it 114 BULLETIN 82, UNITED STATES NATIONAL MUSEUM encircled by a dark band. The sixth and following segments bear long sharp dorsal spines. The opposing spine is long and slender, about as long as the width of the penultimate segment. The terminal claw is considerably longer than the penulti- mate segment, rather slender, and moderately curved. The longer cirrus segments have a slight central constriction. Another specimen bearing the same date is similar but has 20 arms due to the presence of all the IIBr series, eight of which are 4(3-+-4) and two 2. Dated April 16, 1907, are 15 specimens with 17-25 arms 40 to 50 mm. long. The cirri are XIJI-XX. All the IIBr series are 4(83+4). ILIBr series are present in all the specimens, in all cases 2, and developed internally. Of two specimens dated July 23, 1907, one resembles in general the first mentioned from Singapore. The 20 arms are 75 mm. long. Of the 10 IIBr series 7 are 4(3+4), one is 2, and the remaining two consist of a single ossicle each. The cirri are XXIII, 30, 15 mm. long. The other specimen has 12 arms 50 mm. long. There are 2 IIBr 4(3+4) series. The brachials are somewhat wedge-shaped, longer than in the pre- ceding specimen. The cirri are XX, 25-28, 15 mm. long. An example dated July 27, 1907, has 24 arms 60 mm. long. The 10 IIBr series are 4(3+4). Four IIIBr 2 series are present, 3 developed internally and one exter- nally. The cirri are XX, 25-30, 15 mm. long. A specimen with 16 arms 70 mm. long is dated November 27, 1907. Six IIBr 4(3+-4) series are present, but no IIIBr series. The cirri are XXII, 30, the transition segment being usually the eighth. The specimen from Singapore in the British Museum is small, with 20 arms. The specimen from the Danish Expedition to the Kei Islands station 73 is very small, with 12 arms about 30 mm. long. The example from the Danish Expedition to the Kei Islands station 101 is small with 13 arms 50 mm. long and is in process of adolescent autotomy. The specimen from north-northeast of Indramayu Point, Java, has 18 arms. The specimen from Siboga station 322 is small, with 16 arms 60 mm. long. Five of the six IJBr series are 2 instead of 4 (3-+4). The cirri are XVII, 27, 17 mm. long. The color is deep purple. One of the specimens from the Danish Expedition to the Kei Islands station 38 has 22 arms 130 mm. long, slender and very slowly tapering. There are seven IIBr 4 (3+-4) series and five IIIBr 2 series, of which three are internal and two external. The cirri are XVIII, 26-30, about 25 mm. long. The other is similar with 26 arms 125 mm. long. The 10 IIBr series are 4 (3+4); of the six IIIBr series, five are 4 (3+4) and one is 2. Three of the IIIBr series are developed externally and three internally. Of the three specimens from Amboina, one has 24 arms 135 mm. long; the longest cirri have 35 segments. Another has 22 arms with ten IJBr 4 (3+-4) series and two {11Br 2 series, the latter both internal. The cirri have 29 segments. The third has 22 arms, with eight I1Br 4 (3+-4) series and four III Br 2 series, the latter all internal. The specimens from Siboga station 162 are both young. One with 14 arms is deep purple with a broad mediodorsal line of white on the arms. The other, which is much broken, is entirely deep purple. A MONOGRAPH OF THE EXISTING CRINOIDS 115 The specimen from Holothuria Bank has 18 arms, and the cirri are X, 29-31, 20 mm. long. On the cirri long dorsal spines are developed from the eleventh seg- ment onward. Dr. H. L. Clark says that the number of arms in specimens from Broome ranges from 11 in a specimen 70 mm. across to 25 in a specimen 115 mm. from tip to tip. Of the 39 specimens he collected and examined the largest are 175-200 mm. across and have 16-24 arms. He remarks that it is hard to see where the line is to be drawn between this species and punctata. So far as the number of segments in the cirri is concerned, the series of specimens from Broome shows that there is a considerable range of variation, but only a few large specimens show as many as 25-27; most of them have 20-23. None of the specimens from Broome have fewer than 20 cirrus segments. The example from Siboga station 179 has 20 arms 90 mm. long. There are six IiBr 4 (8+4) series and four IIIBr 2 series present, the latter all internal. The distal borders of the radials are obscurely beaded. ‘The sides of the division series and rays bear a few irregular spinous processes. The cirri are XVIII, 28-33 (usually 30), 25 mm. long. Long dorsal spines are developed from the eleventh to the thir- teenth segment onward. The specimen collected by Dr. Th. Mortensen off Jolo has 41 arms about 100 mm. long. Of the 10 IIBr series five are 4 (3-++4), three are 2, and two are 3. Of the 16 IIIBr series eight are 2 and eight are 4 (8+4). The centrodorsal is discoidal, broad, with a broad flat circular dorsal pole 6 mm. in diameter. The cirri are XXXII, 36-45, 30 mm. long. The specimen from Albatross station 5358 has 28 arms 110 mm. long. It resem- bles others from Singapore with which it was compared. Of the five specimens from Bantayan Reef four have about 40 arms, and the fifth has about 20 arms. Carpenter noted (1888) that, while in the specimens from the Mergui Archi- pelago the ambulacra of the disk are very strongly plated, and also the interpalmar areas at their sides, this feature is less marked in Philippine examples. He also remarked (1889) that the straight-edged and ‘‘wall-sided” character of the elements of the IIBr series, a tendency to which is shown by the largest specimens from Mergui, is much more distinct in the specimens from the Philippines. The specimen from Pulo Condor, as described by Gislén, has 20 arms; of the 10 IIBr series six are 2 and four are 4 (8-+4). The cirri are XXII, 25-26. The color is mottled violet and white. In the better of the two specimens from Hongkong there are about 38 arms, which are about 90 mm. long. There are nine IIBr series, five of which are 4 (3+-4), the other four being 2. Of the 15 IIIBr series, 11 are 4 (8+-4) and four are 2. The single IVBr series is 4 (3-+4). The radials, division series, and arms are similar to those of specimens from Singapore. The centrodorsal is discoidal, very slightly concave—almost flat—on the broad dorsal pole, which is 5 mm. in diameter. The cirri are arranged in two irregular closely crowded marginal rows, and are XXIX, 34-36, from 20 to 27 mm. long. The longest proximal cirrus segments are usually from one-third to one-half again as broad as long, and the outer cirrus segments are two and one-half or three times as broad as long. The eighth or ninth and following 116 BULLETIN 82, UNITED STATES NATIONAL MUSEUM bear prominent rather slender dorsal spines which in the outer part of the cirri become more or less broadened laterally, forming short high transverse ridges or ending in a more or less completely bifurcated tip. Toward the extremity of the cirri the dorsal spines again resume their normal character. The other specimen is similar. There are about 40 arms. The 10 IIBr series are 4 (3+-4). Three of the IIJBr series are 2, the remainder being 4 (3+4). Two IVBr series are present, one 2 and one 4 (3+4). The cirri are XXVIII, 33-35, about 25 mm. long. Both of these specimens are dull pinkish with purple cirri, the earlier segments of the latter with white borders and a white dorsal surface. Except for the greater number of arms these specimens agree perfectly with a series at hand from Singapore and the Philippines. The specimen from Hongkong collected by A. Wieler in 1888 has 20 arms and seems not to differ from others from Singapore. Abnormal specimens.—In one of the specimens from Singapore two of the IIBr series consist of a single ossicle each, and in one from Jolo two of the IIBr series consist of three ossicles. In a specimen from Hongkong the dorsal spines in the outer, but not terminal, portion of the cirri are more or less broadened, forming a short high transverse ridge or ending in a more or less completely bifurcated tip. Local variation.—The smallest specimens, with 12-25 arms 40 to 70 mm. long and the cirri with 25-30 segments, are from Singapore. But this does not necessarily mean that this species does not reach a larger size there. It is quite possible that all the specimens were collected in a restricted area where they never grow to full size, or where at the time they happened to be all young. The specimens from the Mergui Archipelago are essentially like those from Singapore. They rarely have more than 30 arms, which are up to 100 mm. in length. The greatest arm length occurs in specimens from the Kei Islands and the Moluceas, although here the number of arms is small. The arms are 22—26 in number, reaching a length of 125 to 135 mm. The longest cirri, while they usually have 30 segments, may have as many as 35. As is commonly the case among the multibrachiate comatulids, the largest number of arms is found in specimens from the Philippine Islands. Here the arms vary from 20 to 41 in number, with a maximum length of 110 mm. The cirri may have as many as 45 segments. In Philippine specimens the disk is less heavily plated than in specimens from the Mergui Archipelago, and the sharp lateral flattening, or “wall-sided” character, of the I[Br series is much more marked. Specimens from Hongkong agree with those from the Philippines in having 38-40 arms and a maximum of 36 cirrus segments. Localities.—King Island, Mergui Archipelago; Dr. John Anderson [P. H. Carpenter, 1888, 1889; Bell, 1888; A. H. Clark, 1911). Eighty miles northwest of Penang; 73 meters; Eastern and Associated Telegraph Co.’s cable-ship Patrol; taken in May 1923 from a cable laid 2 years previously [A. H. Clark, 1929] (2, B. M.). Singapore; Svend Gad [A. H. Clark, 1909, 1911] (32, U. S. N. M., 36262, 36263, E. 1076; C. M.). Singapore; 13 meters; December 12, 1898 [A. H. Clark, 1929] (1, B. M.). A MONOGRAPH OF THE EXISTING CRINOIDS il7 Danish Expedition to the Kei Islands station 73; Sunda Straits (lat. 5°57’S., long. 105°57’ E.); 30 meters; hard bottom; July 28, 1922 (1). Danish Expedition to the Kei Islands station 101; Java Sea; 49 meters; sand, stones, and sponges; August 5, 1922 (1). Java Sea, northnortheast of Indramayu Point, Java (lat. 4°56’ S., long. 108°56’ E.); October 23, 1907; Gier 3, Exp. 20 [A. H. Clark, 1933] (1, Buitenzorg Mus.). Siboga station 322; 1.5 miles south of Tandjong Lajar, south coast of Bawean Island; 22 meters; coral; February 24, 1900 [A. H. Clark, 1918] (1, U.S. N. M., E. 474). Danish Expedition to the Kei Islands station 38; about 35 meters; sand; April 24, 1922 (2). Danish Expedition to the Kei Islands; Amboina; breakwater; about 1 meter; February 28, 1922 (3). Siboga station 162; between Loslos and Broken Islands, western coast of Sala- watti; 18 meters; coarse and fine sand, with clay and shells; August 18, 1899 [A. H. Clark, 1918] (2, Amsterdam Mus.). Holothuria Bank, northwestern Australia; 27 meters [A. H. Clark, 1911, 1913] (1, B. M.). Roebuck Bay (Broome), Western Australia; littoral; H. L. Clark, 1929 and 1932 [H. L. Clark, 1938]. Lagrange Bay, south of Broome, Western Australia; littoral; H. L. Clark, 1929 [H. L. Clark, 1938]. Siboga station 179; Kawa Bay, western coast of Ceram ; 36 meters; stony bottom; September 2-3, 1899 [A. H. Clark, 1918] (1, Amsterdam Mus.). Albatross station 5137; in the vicinity of Jolo (Sulu), Philippines; Jolo light bearing S. 61° E., 1.3 miles distant (lat. 6°04’ 25’’ N., long. 120°58’30’ E.); 36 meters; sand and shells; February 14, 1908 |A. H. Clark, 1908, 1909, 1911] (2, U.S. N. M., 35142, 35151). Albatross station 5138; in the vicinity of Jolo; Jolo light bearing S. 19° E., 2.5 miles distant (lat. 6°06/00’’ N., long. 120°58’50’’ E.); 35 meters; sand and coral; February 14, 1908 [A. H. Clark, 1908, 1909, 1911] @, U.S. N. M., 35130, 35137). Off Jolo; about 36 meters; lithothamnion; Th. Mortensen, March 17, 1914 (1). Albatross station 5358; Jolo (Sulu) Sea; Sandakan light bearing S. 34° W., 19.7 miles distant (lat. 6°06’40’’ N., long. 118°18/15’’ E.); 53 meters; mud; January 7, 1909 [A. H. Clark, 1911] (1, U. S. N. M., 35149). Bantayan Reef, Cebu, Philippine Islands; Lawrence E. Griffin (5, M. C. Z., 388, 397; original numbers 69, 70). Philippines; Prof. C. Semper [P. H. Carpenter, 1888, 1889; A. H. Clark, 1911]. Pulo Condor, Cochinchina; 5 meters; Dr. C. Dawydoft |Gislén, 1936]. Hight miles outside of Hongkong Harbor (lat. 22°12’ N., long. 114915’ E.); 25 meters; Captain Suenson, November 16, 1911 [A. H. Clark, 1913] (2, UsiSsNiM., BH. 1078; C. M.). Hongkong; A. Wieler, 1888 |A. H. Clark, 1912; Mortensen, 1934} (1, H. M.). Geographical range.—From the Mergui Archipelago to the Lesser Sunda Islands, northwestern Australia, the Moluccas, Salawatti, the Phillippine Islands, and Hongkong. 118 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Bathymetrical range.—From the shore line down to 73 meters. Occurrence.—Dr. H. L. Clark says that this is a very common species at Broome, and like all the Zygometras of that region it delights in the clean shallow water and hard sandy bottom of Roebuck Bay. At extreme low water in the greatest spring tides, as in September 1929, a large part of the water ebbs out of Roebuck Bay and vast areas of a clean firm gray sand are exposed. On this bottom are countless patches or isolated plants of coralline and other algae, besides partly buried rock fragments or shells scattered abundantly about, well separated from one another, and each forming the nucleus of a little animal community seeking shelter from the pitiless exposure to which the unusual ebb of the tide subjects them. On almost every alga or other projection a comatulid will be found, usually but one, and the vast majority of these are Zygometras, chiefly comata. It is a hardy species, enduring this exposure to the sun, transferring to a bucket, and transportation to the laboratory without damaging itself in any way. Left by the retreating tide, the comatulids fall relaxed on their sides, the arms closed on the disk and pointing to the water that has abandoned them. But when the tide returns, enduring for a time the washing back and forth of the now coming, now going, wavelets, they soon find the water deep enough to enable them to sit erect, expand their arms, and renew their normal life. Dr. Clark says that seeing this flowerlike expanding of the wilted comatulids is one of the most interesting pleasures of watching the incoming tide. History.—In sending some myzostomes taken from a comatulid collected by Dr. John Anderson at King Island in the Mergui Archipelago to Prof. Ludwig von Graff for identification, Dr. P. H. Carpenter gave the name of the host as Antedon comata. This name was published as a nomen nudum by Professor von Graff in 1887, together with the locality. It was never mentioned again either by von Graff or by Car- penter. Von Graff gave as the locality ‘“Padau Bay in the Mergui Archipelago.” Carpenter said (later) that all the comatulids collected by Anderson at Mergui were obtained at one locality, King Island. Therefore Padau Bay, which I can not locate in any of the available books of reference, must be on the shore of King Island, and is possibly the same as King Island Bay, the main anchorage at King Island. In discussing Antedon elegans in the Challenger report in 1888 Carpenter said that Professor Semper’s Philippine collection contained representatives of the type and that he had recently found a most valuable series of varying forms of this species among the comatulids dredged by Dr. Anderson in the Mergui Archipelago. These Philippine and Mergui specimens represent not Zygometra elegans but Z. comata. In 1888 Prof. F. Jeffrey Bell mentioned that Dr. Anderson had collected Antedon elegans [= Zygometra comata] in the Mergui Archipelago. Under the name Antedon elegans Carpenter in 1889 recorded and gave notes upon the five specimens collected by Dr. Anderson at King Island in the Mergui Archipelago. He compared these with the specimens obtained by the Challenger and by the Alert (representing true elegans) and noted the differences. He mentioned a feature exhibited by the Mergui specimens in common with, and more distinct in, others from the Philippines in Semper’s collection. Carpenter said that there were in the collection of comatulids from the Mergui Archipelago three species of “Antedon” infested with myzostomes, and that these myzostomes had been placed in the hands of Professor von Graff for identification A MONOGRAPH OF THE EXISTING CRINOIDS 119 and description. In his account of the several species, however, he mentioned the presence of myzostomes on only two, Antedon elegans and Antedon milberti. In von Grafi’s description of the myzostomes he likewise listed only two comatulids from the Mergui collection among the hosts, Antedon comata and Antedon milberti. It is thus evident that Carpenter at first believed that the zygometrids collected at Mergui represented a new species, to which he tentatively applied the manuscript name comata, but later decided that they should be referred to Antedon elegans. The conclusion that the form referred to as Antedon elegans in Carpenter’s report upon the collection from the Mergui Archipelago is really the one he had originally called in manuscript Antedon comata is further borne out by the detailed comparisons he gives between the Australian specimens (representing true elegans) and those from elsewhere. In 1908 I recorded specimens of this species, as Zygometra elegans, from Albatross stations 5137 and 5138 in the Philippines. In discussing Zygometra elegans in my paper on the comatulids collected by the Gazelle, which was published on June 1, 1909, I remarked that I could not agree with Carpenter in regarding Antedon fluctuans as a synonym of Antedon elegans. I said that “judging from a very large series from Singapore and the Philippine Islands, fluctuans has uniformly about thirty arms with wedge-shaped brachials, and com- paratively short cirri with not more than thirty-five joints; it is a smaller and more delicate species, with the IIIBr series always 2, developed internally, and bears a very close resemblance to the species of Heterometra.” In a memoir on the comatulids in the Zoological Museum of Copenhagen, pub- lished on November 16, 1909, under the name Zygometra fluctuans, I recorded and gave notes upon 23 specimens from Singapore collected by the Danish Consul, Svend Gad. I said that “these specimens agree closely with Carpenter’s description of fluctuans, except that the brachials are oblong distally ; moreover, they all agree among themselves, and with Philippine specimens.”” I remarked that I had at first followed Carpenter’s lead and considered them, as Carpenter had his own fluctuans, to be representatives of Bell’s previously described elegans; but when I examined the col- lections made by the Gazelle in northwestern Australia I found that I was wrong, for in that collection there are specimens of what is undoubtedly true elegans which agree with Bell’s diagnosis and have much longer and more robust cirri, composed of more numerous segments, than any of these specimens. I concluded that, judging from the material at hand, elegans and fluctuans are both perfectly good species, and there is not the slightest difficulty in differentiating them. In a paper on some crinoids dredged by the Albatross in the Philippines published on February 15, 1911, I recorded ““Zygometra comata (A. H. Clark)” from station 5358, giving the number and length of the arms of the single specimen and remarking that it resembled others from Singapore in the collection of the University of Copenhagen. In my memoir on the recent crinoids of Australia published later in the same year, in a footnote appended to the differential characters of Zygometra elegans I proposed the name Zygometra comata, “reinstating a nomen nudum long ago applied to it in MS. by Carpenter,” for the species represented by the specimens from the Mergui Archipelago, Singapore, and the Philippine Islands, which I had previously considered as representing Carpenter’s Antedon (Zygometra) fluctuans. I said that I 120 BULLETIN 82, UNITED STATES NATIONAL MUSEUM had found, in working over the magnificent collection of comatulids made at Singapore by Svend Gad, that the common Zygometra occurring there and in the Mergui Archi- pelago is not the same as that called elegans by Bell, though it had been united with that species by Carpenter. I gave a few of the characters of the new species and said that a satisfactory description could be found in my report upon the comatulids of the Copenhagen Museum. In 1912 I recorded and gave notes upon a specimen of Zygometra comata from Hongkong that I had examined in the Hamburg Museum in 1910. In my monograph on the crinoids of the Indian Ocean published later in the same year I gave the synonymy of Z. comata and its geographical and bathymetrical ranges. In 1913 I recorded and published notes upon two specimens that had been col- lected by Captain Suensson, commanding one of the Danish cable repair ships, off Hongkong. The specimen in the British Museum recorded by me from Holothuria Bank in 1913 under the name of Zygometra punctata is probably an example of this species. This is inferred from the notes given on the cirri. The specimen has not been reexamined. In a memoir on the comatulids collected by the Australian Fisheries Investiga- tions steamer Endeavour between Fremantle and Geraldton, Western Australia, published in 1914, I discussed the relationships between Z. comata and the other species of the genus. In the Siboga report upon the unstalked crinoids published in 1918, I recorded and described four specimens from three Siboga stations, and in 1929 I recorded two specimens from the collections made by the Eastern and Associated Telegraph Co.’s cable repair ship Patrol, and another from Singapore. In 1934 Dr. Th. Mortensen listed Z. comata as one of the crinoids occurring at Hongkong. In 1936 Dr. Torsten Gislén recorded and gave notes upon a specimen collected by Dr. C. Dawydoff at Pulo Condor, Cochinchina. In 1938 Dr. H. L. Clark recorded and gave extensive notes on 37 specimens from Roebuck Bay and vicinity and 2 from the vicinity of Lagrange Bay collected by him- self in 1929 and 1932. ZYGOMETRA PUNCTATA A. H. Clark Puate 6, Figures 18-21 Heterometra bengalensis A. H. Crarx, Mem. Australian Mus., vol. 4, 1911, p. 722 (occurs south to Port Curtis), p. 723 (occurs on Holothuria Bank [= Z. comata]), p. 734 (in key), p. 768 (synonymy (=bengalensis}; characters, in part; Port Curtis [specimen from Holothuria Bank, 15 fathoms = Z. comata); range, in part); Die Fauna Siidwest-Australiens, vol. 3, Lief. 13, 1911, pp. 440, 443, 444, 446 (Australian records [Holothuria Bank refers to Z. comata]); Crinoids of the Indian Ocean, 1912, p. 130 (Queensland [Holothuria Bank refers to Z. comata]); Die Fauna Siidwest- Australiens, vol. 4, Lief. 6, 1913, pp. 310, 313 (all Australian records refer to Z. punctata (Holo- thuria Bank refers to Z. comata)). Zygometra punctata A. H. Cuark, Proc. Biol. Soc. Washington, vol. 25, 1912, p. 23 (description; Siboga station 273); Die Fauna Siidwest-Australiens, vol. 4, Lief. 6, 1913, p. 310 (Australian records [Holothuria Bank refers to Z. comata])—Rr1cHENSPERGER, Abh. Senck. naturf. Ges., vol. 35, Heft 1, 1913, p. 82 (Aru Islands), p. 89 (Aru Islands; Dobo Strait, 50 m.; characters) .— A. H. Crarx, Rec. Western Australian Mus., vol. 1, pt. 3, 1914, p. 114 (records [Holothuria Bank refers to Z. comata]), p. 123 (comparison with Z. elegans); Internat. Rev. gesamt. Hydro- A MONOGRAPH OF THE EXISTING CRINOIDS 121 biol. und Hydrogr., 1915, pp. 224 ff. (detailed account of the distribution (Holothuria Bank refers to Z. comata]); Unstalked crinoids of the Siboga-Exped., 1918, p. 60 (in key; range), p. 61 (synonymy; detailed description; station 273 [specimen from Holothuria Bank = Z. comata)), p. 275 (listed), pl. 19, figs. 40, 41—Guisiin, Kungl. Fysiogr. Sallsk. Lund Handl., new ser., vol. 45, No. 11, 1934, p. 21. Diagnostic features —The cirri are short and rather stout, strongly recurved in the distal half, with not more than 21 segments of which the longest are about as long as broad. There are 11-21 arms 25-35 mm. inlength. The division series are broad, in close lateral contact, with straight lateral edges which are more or less flattened against those of their neighbors. The color—white with numerous circular purple spots on the division series and arm bases and in bandlike areas on the outer portions of the arms—and the conspicuously banded white and purple cirri seem to be quite distinctive. But it is always unsafe to rely too much on color when dealing with comatulids. This species is much more likely to be confused with the young of certain species of Heterometra than with the other species of Zygometra. It is easily distinguished from these, however, by the syzygy between the elements of the [Br series. Description.—The centrodorsal is discoidal, thin, with a broad, flat, circular dorsal pole 2.7 mm. in diameter. The cirri are XV, 18-19 (usually 18), 9 mm. long. The first segment is very short, the second and third are about three times as broad as long, the fourth is about twice as broad as long, and the fifth or sixth is about as long as broad. The next two or three segments are similar and those following decrease very gradually in length so that the outermost 10 or 11 are slightly broader than long. The sixth and following segments bear prominent sharp dorsal spines. The radials are entirely concealed by the centrodorsal. The IBr, are very short and bandlike, six or more times as broad as long, united to the IBr, by syzygy. The IBr, (axillaries) are low triangular with the lateral angles slightly truncated, three times as broad as long. There are 9 I1Br series present, 8 of which are 4 (3+4) and one of which is 2. A single IJIBr series of two elements is developed internally on one of the IIBr 4 (3-+4) series. The I]Br series are broad like the IBr series with straight lateral edges which are more or less flattened and almost or quite in apposition. The lateral outer portions of the elements of the division series are roughened or very finely papillose as in certain species of Mariametra. The 21 arms are 35 mm. long and are comparatively/short and"rather stout. The first brachials are slightly wedge-shaped, three times as broad as the median length, and are entirely united internally. The second brachials are much more obliquely wedge-shaped, twice as broad as the external length. The first syzygial pair (com- posed of brachials 3+-4) is slightly longer internally than externally, and twice as broad as long internally. The next three brachials are oblong, three to four times as broad as long, and those following are wedge-shaped, twice as broad as long, becom- ing as long as broad terminally. The brachials have rather strongly produced distal edges. Py resembles P; and is composed of 20 segments. P, is about 5 mm. long, rather stout basally but in the distal half tapering to a slender tip, and is composed of 20 208244409 122 BULLETIN 82, UNITED STATES NATIONAL MUSEUM segments of which the first is about three times as broad as long, and those following gradually increase in length becoming as long as broad on the ninth and slightly longer than broad terminally. The longer proximal segments are rather strongly carinate, the carination having a straight profile which is parallel with the long axis of the segments. The short outer segments have slightly prominent distal ends. P, is similar, but very slightly smaller and shorter. P; is 2.5 mm. long, composed of 12 segments; except for its smaller size it is similar to the preceding pinnules. P, is 2 mm. long, with 12 segments, of which the first three are much broader than long, the fourth is about as long as broad, and the distal are twice as long as broad. P, is 2.5 mm, long, with 12 segments, and resembles P,, but it is very slightly stouter basally and the distal segments are longer. The following pinnules resemble P;. The distal pinnules are 4 mm. long, with 15 segments, of which the distal are nearly or quite three times as long as broad. The color in alcohol is white with numerous regular purple spots on the division series and arm bases and in bandlike areas on the distal portions of the arms. The cirri are white with a band of purple on each segment. Notes.—A specimen from Siboga station 273, which is similar to the type, has 17 arms, with 5 I[Br 4 (3+4) series and 2 IIBr 2 series; the cirri are XVII, 18, 9 mm. long. The remaining two specimens from this station are small. Reichensperger referred to this species a 10-armed young individual from the Aru Islands. The 10 graceful, slender arms are 27 mm. long. The centrodorsal is thin, with the cirri arranged in a single irregular row. The cirri are XV, 13-15, up to 5 mm. in length. The first segment is short and laterally compressed, the second is notice- ably longer than broad, and the third, fourth, and fifth are the longest, about twice as long as broad. ‘The length of the segments then decreases slowly so that the last is broader than long. The opposing spine is very long and straight. The terminal claw is about twice as long as the penultimate segment, long and strongly curved. From the fourth or fifth segment onward strong dorsal spines are developed. P, is 4 mm. long, with 20-22 segments, of which the first is broader than long, the next 10 or so are about as long as broad, and the remainder slightly elongated, the pinnule becoming flagellate terminally. P, is 3.5 mm. long, with about 18 segments, and resembles P,. P; is scarcely 2 mm. long and is composed of 11 or 12 segments. FP, is the shortest, about 1.5 mm. long, moderately stout, with 8 or 9 segments. The following pinnules increase in length and in the number of their component segments. P; has 12 seg- ments. The distal pinnules are about 3mm. long. The IBr series are rather short, and laterally free. The brachials are relatively long, with the ends only slightly oblique. The lowest brachials bear slight ventrolateral processes. The disk is lacking. The color in life was light red. Of the three specimens from Port Curtis one has 12 and two have 11 arms, 25 mm, to 33mm.long. The cirri of the largest specimen are XV, 14-18 (usually 15). Localities.—Siboga station 273; anchorage off Pulu Jedan, eastern coast of the Aru Islands (pearl banks); 13 meters; sand and shells; December 23-26, 1899 [A. H. Clark, 1912, 1913, 1918] (4, U.S. N. M., E. 442; Amsterdam Mus.). Aru Islands; Dr. H. Merton’s station 4; Dobo Strait; 50 meters; calcareous rock; March 20, 1908 [Reichensperger, 1913]. A MONOGRAPH OF THE EXISTING CRINOIDS 123 Port Curtis, Queensland [A. H. Clark, 1911, 1913] (3, U.S. N. M., 35150; Austra- lian Mus.). Geographical range.—Aru Islands and the northern coasts of Australia south to Port Curtis, Queensland. Bathymetrical range.—Littoral and sublittoral; from the shoreline down to 50 meters. History —This species was first mentioned in my monograph of the recent crinoids of Australia, published in 1911, under the name of Heterometra bengalensis. The three small specimens from Port Curtis referred to Heterometra bengalensis represent this species, but the one from Holothuria Bank in the British Museum is in reality an example of Zygometra comata. In 1912 Zygometra punctata was described as a new species on the basis of a specimen from Siboga station 273. In 1913 in a supplementary paper on the crinoids of southwestern Australia all of the Australian records of Heterometra bengalensis were referred to this species. In a paper on the crinoids in the British Museum published in 1913, Zygometra punctata was recorded from Holothuria Bank. The specimen from Holothuria Bank is in reality, however, Z. comata. It was this specimen that served as the basis for all the records from Holothuria Bank in both earlier and later papers. In 1913 Dr. August Reichensperger described a young 10-armed specimen from the Aru Islands. In the Siboga report published in 1918 this'species was redescribed and figured, and another specimen from station 273 was mentioned. Holothuria Bank was again erroneously included in the range. Dr. Hubert Lyman Clark’s Heterometra delicata, based upon a single specimen that he had found on the sand flat on the southern end of Friday Island, Torres Strait, was included in the synonymy of this species. This is an error, as the type specimen of Heterometra delicata is a young specimen of some species of Heterometra. ZYGOMETRA PRISTINA A. H. Clark PLATE 7, FiauREs 22, 23 Zygometra pristina A. H. Cuark, Proc. U. 8. Nat. Mus., vol. 39, 1911, p. 537 (description; Albatross station 5276); Crinoids of the Indian Ocean, 1912, p. 106 (synonymy; habitat) —RrIcHEN- SPERGER, Abh. Senck. naturf. Ges., vol. 35, Heft 1, 1913, pp. 91, 92 (comparison with Z. mertoni [=young Z. microdiscus]).—A. H. Cuarx, Unstalked crinoids of the Siboga-Exped., 1918, p. 60 (in key; range)—GusLin, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 17, 18. Diagnostic features —The possession of only 10 arms and the partially developed syzygy between the elements of the IBr series distinguish this species at once from all the others in the genus. The arms are 50 mm. long. The cirri are XII, 20-21, 12 mm. long, with the longest segments about one-third again as long as broad. Description.—The centrodorsal is low hemispherical with the bare polar area papillose, 1 mm. in diameter. The cirrus sockets are arranged in a single irregular not especially crowded row. The cirri are XII, 20-21, 12 mm. long, and moderately slender. The first segment is short, the following gradually increasing in length and becoming about one-third again as long as broad on the fourth. The next two segments are similar, and those 124 BULLETIN 82, UNITED STATES NATIONAL MUSEUM succeeding slowly decrease in length, those in the outer part of the cirri being about as long as broad. The sixth or seventh and following segments bear long sharp dorsal spines. The opposing spine is nearly as long as the width of the penultimate segment, and is slender, sharp, and erect. The terminal claw is twice as long as the penultimate segment and is strongly curved proximally, becoming straighter distally. The cirrus segments are somewhat constricted centrally with slightly expanded and overlapping distal ends. The disk is covered with rounded isolated flat plates and is thickly plated along the ambulacral grooves and on the anal tube. The radials are short, about four times as broad as long, sometimes with a faintly marked row of small tubercles along the distal border. The IBr, are about as long as the radials, oblong, four times as broad as long. The IBr, (axillaries) are pentagonal, twice as broad as long. The elements of the IBr series are united by a curious syzygy (pseudosyzygy) in which the outer portion of the joint face for about one-half the distance from the periphery to the rim of the central canal is marked with radiating ridges, the space within this border being smooth and flat except for the low and narrow synarthrial longitudinal ridge. The 10 arms are 50 mm. long. The first two brachials are subequal, slightly wedge-shaped, twice as broad as the exterior length. The first syzygial pair (com- posed of brachials 3+4) is slightly longer interiorly than exteriorly, and twice as broad as the exterior length. The next four brachials are oblong, slightly over twice as broad as long, and those following become obliquely wedge-shaped, about as long as broad, after the proximal third of the arm less obliquely wedge-shaped and about as long as broad, and terminally slightly longer than broad. The brachials after the tenth have somewhat produced distal ends. Syzygies occur between brachials 34-4, again most commonly between brachials 13-+-14, and distally at intervals of 6-10 (usually 8 or 9) muscular articulations. P; is 8 mm. long, moderately stout basally but tapering rapidly in the proximal half and becoming very slender and thread-like distally, with about 25 segments of which the first is about twice as broad as long and those following gradually increase in length becoming about as long as broad on the fifth and following and slightly longer than broad terminally. The first three to five segments are slightly carinate and rounded prismatic. P, is about 10 mm. long, stouter than P, but otherwise similar to it, with about 30 segments. PP; is 3.5 mm. long, small, weak, and slender, and becoming exceedingly slender in its distal half. P, is similar, 3 mm. long. P, is 2 mm. long, not tapering so rapidly as the preceding pinnules and therefore appearing somewhat stouter, with 10 segments. The following pinnules are similar, soon slowly increasing in length and slenderness. The distal pinnules are very slender, 6 mm. long, with 15 or 16 segments. Locality.— Albatross station 5276; China Sea, in the vicinity of southern Luzon; Malavatuan Island (N. W.) bearing N. 61° 30’ E., 6.5 miles distant (lat. 13°49/15’’ N., long. 120°14’45’’ E.); 33 meters; shells, pebbles and sand; July 17, 1908 [A. H. Clark, 1911, 1912, 1918; Reichensperger, 1913] (1, U.S.N.M., 27489). Remarks.—This species is as yet known only from the single specimen dredged by the Albatross in 1908. A MONOGRAPH OF THE EXISTING CRINOIDS 125 Genus CATOPTOMETRA A. H. Clark Antedon (part) A. H. Cuarx, Proc. U. 8. Nat. Mus., vol. 33, 1907, pp. 72, 150. Zygometra (part) A. H. Cuarx, Smithsonian Misc. Coll., vol. 50, 1907, pp. 339, 345, 347. Catoptometra A. H. Cuarx, Proc. U. S. Nat. Mus., vol. 34, 1908, p. 317 (listed, with 3 species men- tioned, hartlaubi, koehleri, and rubroflava; in family Zygometridae; no diagnosis and no type designated), p. 505 (type designated as Antedon hartlaubi A. H. Clark, 1907); vol. 35, 1908, fig. 16, p. 119 (arm structure); Smithsonian Misc. Coll., vol. 52, 1908, p. 217 (characteristic rough feeling of arms); Proc. Biol. Soc. Washington, vol. 22, 1909, p. 2 (cirri and Jower pinnules compared with those of Charitometridae); Vid. Medd. Naturh. Foren. Kgbenhavn, 1909, p. 152 (cirri and brachials compared with those of Zygometra fluctuans [comata]), p. 193 (probably occurs at Singapore though not yet discovered there) ; Amer. Journ. Sci., ser. 4, vol. 32, 1911, p. 129 (characteristic of the Japanese fauna; significance); Mem. Australian Mus., vol. 4, 1911, p. 725 (absent from Australia); Crinoids of the Indian Ocean, 1912, p. 9 (absent from Austra- lia), p. 10 (great development of local species in Japan), p. 11 (absent from west coast of Malay Peninsula and from farther west), p. 17 (significance of conditions in this genus in southern Japan), p. 21 (range; reaches to 153 fathoms), p. 56 (in key), p. 106 (original reference; type); Die Crinoiden der Antarktis, 1915, p. 112 (synonymy; range); Amer. Journ. Sci., vol. 40, 1915, p. 62 (list of recent species), p. 63 (number of recent and fossil species; range), p. 65 (con- fined in recent seas to the western Pacific; fossil in Patagonia and Europe).—F. W. CLARKE and Wuere.er, U. 8S. Geol. Surv. Prof. Pap. 90-L, 1915, p. 195 (inorganic constituents of the skeleton); Professional Paper 102, 1917, pp. 23 and following (same).—A. H. Cuarx, Un- stalked crinoids of the Siboga-Exped., 1918, p. 63 (key to the included species); Univ. Iowa Stud. Nat. Hist., vol. 9, No. 5, 1921, p. 14 (includes both recent and fossil species; not known from West Indies); Smithsonian Misc. Coll., vol. 72, No. 7, 1921, p. 17.—F. W. Cuarxe and WHEELER, U.S. Geol. Surv. Prof. Pap. 124, 1922, p. 20 (inorganic constituents of the skeleton). — Gistin, Nova Acta Reg. Soc. Sci. Upsaliensis, ser. 4, vol. 5, No. 6, 1922, p. 64; Zool. Bidrag Uppsala, vol. 9, 1924, pp, 35, 84, 100, 107, 117, 159; Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, p. 18. Diagnosis.—A genus of Zygometridae in which the outer cirrus segments are but little, if at all, shorter than the proximal and do not bear dorsal processes. Geographical range.—Southern Japan from the Korean Straits to Sagami Bay, the Bonin and Philippine Islands, Hong Kong, and southward to the Lesser Sunda Islands. Bathymetrical range.—From 25 to 914 meters; most abundantly represented be- tween 81 and 196 meters. Remarks.—The genus Catoptometra is one of the most easily recognized of all the genera of multibrachiate comatulids. The stout strongly recurved smooth cirri with more or less swollen articulations and consequently a knobby appearance, combined with the presence of a syzygy between the elements of the [Br series, separate it at once from all other genera including multibrachiate forms. The syzygy in the IBr series is very brittle, and almost invariably during capture one or more of the post- radial series—not infrequently all—are broken across at this point. Superficially the species of Catoptometra resemble to a certain extent some of the species in certain genera of Charitometridae, but these last are easily recognized by the sharply carinate pinnules and the occurrence of conspicuous side plates and covering plates on the ventral surface of the pinnules. Of the four species included in the genus one, C. hartlaubi, seems to occupy a rather isolated position. The other three are more closely allied, but one of these, C. magnifica, differs markedly from the others in having all the division series of two ossicles only and in having a greater number of arms. The two remaining species, 126 BULLETIN 82, UNITED STATES NATIONAL MUSEUM C. rubroflava and C. ophiura, seem to be very closely related, though on the basis of our present knowledge they are quite distinct. One species, C. magnifica, has a very extensive range, occurring from southwestern Japan and the Bonin Islands to the Lesser Sunda Islands; two species, C. hartlaubi and C. rubroflava, occur only in southern Japan, the latter ranging south to Hongkong; and the fourth species, C. ophiura, is found only from the Lesser Sunda Islands to the Philippines. Though occurring in rather shallow water, the species of Catoptometra live at a greater depth than the species of Zygometra, the upper limits of the former slightly overlapping the lower limits of the latter. History —The generic name Catoptometra was first used in a list of the crinoids of Japan published July 15, 1908. No diagnosis was given. ‘Three species were mentioned, Catoptometra hartlaubi, C. koehteri, and C. rubroflava. In another paper published in the same volume on August 25, 1908, the genus Catoptometra was listed and the type species was given as Antedon hartlaubi A. H. Clark, 1907. No diagnosis was published until 1912, when it was inserted in a key to the genera of the family Zygometridae and the characters separating it from Zygometra were given. The question whether the endocyclic comatulids with a syzygy in the IBr series and smooth cirri should retain the name Catoptometra, or assume the unfamiliar name Zygometra, has already been discussed (see page 81). KEY TO THE SPECIES IN THE GENUS CATOPTOMETRA a'!, Arms 30-80 in number; all division series 2 (southwestern Japan, Bonin Islands, and Philippines to southern Annam and Lesser Sunda Islands; 36-914 meters) -__----_-- magnifica (p.126) a?. Less than 30 arms; I/Pr series 4 (3+-4). b'. Cirri short and stout, few or none of segments longer than broad; 11-14 arms; in life broadly banded with bright red and bright yellow (southern Japan from Sagami Bay to Korean Straits and southward to Hongkong; 25-183 meters) ________-_-_-__-- rubroflava (p.133) b*. Cirri longer and less stout, longest segments being twice as long as median width; 20 or more arms. c!, Arms 20-23 in number; cirrus segments 14-17 (usually 14 or 15); elements of division series and first two brachials with prominently everted and produced spinous distal edges; dorsal pole of centrodorsal papillose or finely spinous; color in life uniform, brown or greenish (southern Japan from Sagami Bay to Korean Straits; 137 [7115] — 278 meters) hartlaubi (p. 139) @. Arms 20-27 in number; cirrus segments 15-19 (usually 17-19); elements of division series and arm bases smooth; dorsal pole of centrodorsal smooth; color in life usually dark green, rarely red and yellow (from Lesser Sunda Islands to Philippines; 69-106 meters)— ophiura (p. 136) CATOPTOMETRA MAGNIFICA A. H. Clark _PLATE 7, Ficures 24, 25; Piatre 8, Figure 26; Piate 9, Ficure 31 Catoptometra magnifica A. H. Crarx, Smithsonian Mise. Coll., vol. 52, 1908, p. 208 (description; Albatross station 5137); Crinoids of the Indian Ocean, 1912, p. 106 (synonymy; habitat); Amer. Journ. Sci., vol. 40, 1915, p. 62 (listed); Unstalked crinoids of the Siboga-Exped., 1918, p. 63 (in key; range).—Grsi6n, Nova Acta Reg. Soc. Sci. Upsaliensis, ser. 4, vol. 5, No. 6, 1922, p. 67 (compared with C. m. minor) —A. H. Crarx, Journ. Linn. Soc. (Zool.), vol. 36, 1929, p 637 (off Cape Padaran, 80 fathoms; Rotti Strait, 100 fathams.).—Gist4n, Kungl. Fysiogr. Siillsk. Handl., new ser., vol. 45, No. 11, 1934, p. 33; Kungl. Fysiogr. Sallsk. Lund Férh., vol. 7, No. 1, 1936, p. 2 (southern Annam), p. 4 (French Indochina), p. 6 (range). A MONOGRAPH OF THE EXISTING CRINOIDS 127 Catoptometra magnifica var. minor Gistin, Nova Acta Reg. Soc. Sci. Upsaliensis, ser. 4, vol. 5, No. 6, 1922, p. 4 (127-182 m.), p. 65 (description; Bock’s stations 46, 47, 59, 61), p. 67 (also in Mortensen’s collection), figs. 46, 47, p. 70; Vid. Medd. Dansk Naturh. Foren., vol. 83, 1927, p. 2 (stations 10, 13, 24, 26; 109-900 m.), p. 20 (stations 10, 13, 24, 26; notes), pp. 68, 69 (listed). Catoptometra magnifica minor Gistin, Nova Acta Reg. Soc. Sci. Upsaliensis, ser. 4, vol. 5, No. 6, 1922, p. 6 (Bonin Islands), p. 8, p. 27 (tabulation of characters), pp. 182, 183 (listed), figs. 46, 47, p. 70; Zool. Bidrag Uppsala, vol. 9, 1924, p. 71 (nonmuscular articulations), p. 74 (syzygies), p. 285 (Kiu Shiu, Japan, 900 m.; details), p. 286, fig. 59, p. 75 (syzygial face); Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, p. 24. Diagnostic features—All the division series are of two ossicles only, and the arms are more than 30 in number. The cirri are long, stout, recurved, and knobby, com- posed of 18-25 segments and reaching 30-35 mm. in length. The arms are 130-140 mm. long. Specimens from the Philippines were apparently in life bright yellow with regular bands of bright red on the arms. They were readily distinguished by this coloration from the species with which they were associated. But Catoptometra rubroflava always and C. ophiura sometimes is of the same color, while it is not certain that this species is constant in its coloration. Description—The centrodorsal is discoidal, large, with a broad polar area 5 to 10 mm. in diameter, which is moderately concave with a deep rounded pit in the center. The cirrus sockets are arranged in two closely crowded and irregular more or less alternating rows. The cirri are XX X-XL, 18-25 (usually 20-24), 30-35 mm. in length, stout and knobby in appearance. The first segment is about twice as broad as long, the second is not quite so long as broad, the third is about as long as broad or very slightly longer than broad, the fourth is slightly longer, and the fifth is slightly longer still, about half again as long as its median width. The next three segments are similar to the fifth and those following gradually decrease in length so that the terminal six or seven are about as long as broad. The opposing spine, though prominent, is small, terminal- ly situated, rarely reaching to more than one-third the width of the penultimate segment in length. The terminal claw is large, longer than the penultimate segment, usually half again as long and sometimes even longer, stout and moderately curved. The lateral and dorsal profiles of the cirrus segments are deeply concave, though the ventral profile is nearly straight. This causes the articulations to stand out promi- nently and gives the cirri a characteristic appearance, like those of C. rubroflava. This feature becomes less and less marked as the segments decrease in length distally. The disk is more or less plated along the ambulacral grooves. The radials, and usually also the IBr,, are concealed by the centrodorsal. When visible the IBr, are very short and are united in their proximal half but widely sep- arated distally. The IBr, (axillaries) are short, triangular, in the smaller specimens about three times as broad as long, in the larger four or five times as broad as long. The I[Br, I1[Br, and IVBr series are 2. The first ossicles following each axillary are interiorly united for their proximal half, but their distal halves diverge from the point of union almost in a straight line so that the division series and the arms are well separated. 128 BULLETIN 82, UNITED STATES NATIONAL MUSEUM The arms are 20-80 (ussally 30-40) in number, 140 mm. long. The first brachials are usually rather large, sometimes nearly as long exteriorly as broad, interiorly united in the proximal half, diverging in almost a straight line in the distal. There is a considerable diversity in size, some first brachials being very short, while most of them are about twice as broad as long exteriorly. The second brachials are nearly oblong, about twice as broad as long. The first syzygial pair (composed of brachials 3-+-4) is oblong, somewhat less than twice as broad as long. The next six or seven brachials are oblong, about twice as broad as long, those following becoming wedge- shaped and then almost triangular, about twice as broad as long, and slowly less and less obliquely wedge-shaped and very gradually increasing in length so that the termi- nal brachials are wedge-shaped, about as long as broad or rather longer, with rather prominent articulations. The second and following brachials have projecting and finely spinous distal edges. Syzygies occur between brachials 3+-4, again from between brachials 13-14 to between brachials 17-++-18 (with rarely an additional one between brachials 9+10), and distally at intervals of from 3 to 13 (usually 6-8) muscular articulations. The lower pinnules are very slender and flagellate. P, is 15 to 20 mm. in length, very delicate, with about 60 segments of which the first five are broad and are pro- vided with a dorsal carinate process and the remainder are about as long as broad. P, is longer, about 22 mm. in length, and slightly stouter, with the same number of segments of which the first five are modified as in P, and the remainder are about as long as broad. P, is similar to P. and is of about the same length. The following pinnules decrease gradually in length, in basal stoutness, and in the number of their component segments, P, being 12 mm. long with 30 segments of which those in the proximal third are similar to those of the second and third pinnules, and those in the distal portion are longer than broad, becoming about twice as long as broad terminally, and P, being 8 mm. long, with about 20 segments, of which the first two are not quite so long as broad, the third and fourth are about as long as broad, and the following become progressively elongated and about twice as long as broad distally. The distal pinnules are very slender, 9 mm. long with about 30 segments of which the first is wedge-shaped, not so long as broad, the second is trapezoidal, about as long as its greater width, the third is longer than broad, and the remainder are about twice as long as broad. The pinnule segments have slightly projecting and very finely spinous distal ends. Notes.—The larger of the two specimens from Rotti Strait is rather small and slender, with about 30 arms, which are about 130 mm. long; the arms are arranged in 2,1,1,2 order. The other specimen from Rotti Strait is very small and much broken; it is apparently undergoing adolescent autotomy. The specimen from the Danish Expedition to the Kei Islands station 27 is young with 14 very slender arms 70 mm. long. One ray bears one IIBr series, which carries 2 ILIBr series, and another ray has one IIBr series. The specimen from off southern Annam is a slender example with about 20 arms, about 130 mm. long. The longest cirri have 22 or 23 segments. The two youngest cirri, at the periphery of the centrodorsal, differ from the others in lacking the strongly produced and everted distal ends of the segments, which gives them quite a distinctive aspect; they are also not so strongly curved as are the other cirri. A MONOGRAPH OF THE EXISTING CRINOIDS 129 Of the five specimens from south of the Goto Islands one has 38 arms 180 mm. long; there are several slender cirri with no expansion of the ends of the segments. Another has 34 arms 135 mm. long; the lateral borders of the division series have rather numerous fine tubercles. The cirri are XLV, 19-20 (usually 20), about 30 mm. long. Among the cirri there is a single one 18 mm. long with 18 segments that is slenderer than the others, scarcely at all curved, in which the component segments do not have expanded ends. The centrodorsal is 6 mm. in diameter, with the dorsal pole slightly concave. A third specimen has 33 arms 135 mm. long. All the division series are 2. The second and third segments of the first seven or eight pinnules bear a very high carinate process, and the fourth and sometimes the fifth carry a much smaller carinate process. The lateral edges of the elements of the division series have a few small tubercles. The centrodorsal is discoidal with the broad dorsal pole slightly concave, 5 mm. in diameter. The cirri are XXXIII, 19-23 (usually 22). Another example has about 30 arms 140 mm. long. The outer portion of the proximal and distal edges of the elements of the division series are swollen and more or less beaded, and the lateral edges bear small tubercles. The fifth specimen is small, with 25 arms. In the specimen from Mortensen’s station 10 the cirri are XXX, 19-28, from 17 to 25 mm. long. The IIBr and IIIBr series are 2, the latter being externally developed. The division series have well-developed ventrolateral tubercles, 1 to 3 on each segment. The 29 arms are about 100 mm. long. There are 14 brachials to each 10 mm., or 13 if the syzygial pairs are counted as units. The distal intersyzygial interval is 14-19 muscular articulations. P, is 7 mm. long, with about 30 segments. P,is 13 mm. long, with 37 segments. P; is composed of 33 segments. P, is about 10 mm. long. The distal pinnules are 11 mm. long, with about 25 segments. The disk is lacking. In one of the specimens from Mortensen’s station 13 the cirri are XXXIII, 22-25, from 20 to 30 mm. long. The centrodorsal is 8 mm. in diameter, with the concave dorsal pole 7 mm. across. The 35 arms are 80+- mm. long. There are two IVBr 2 series. In another specimen from Mortensen’s station 13 the cirri have 19-25 seg- ments. There are 26+ (probably 30) arms 80 mm. long. The disk is incised, 8 mm. in diameter with the anal tube 3 mm. high. Ina third specimen two postradial series bear six and four arms, the three other postradial series being broken away. In the fourth specimen from station 13 the cirri have 19-21 segments. There are 22 arms 90+ mm. long. In the specimen from Mortensen’s station 24 the cirri are XXX, 22-25, from 21 to 27 mm. long. The 26 arms are 110 mm. long. The distal intersyzygial interval is 4-11 (usually 8) muscular articulations. P, is 8 mm. long, with 25 segments. P; is 11 mm. long, with 33 segments. P; is 10 mm. long, with 24 segments. The distal pinnules are 9 mm. long, with 22 segments. The first 10 pinnules have the usual characteristic processes on the proximal segments. The disk, which is detached, is incised and measures 12 mm. in its longest and 8 mm. in its shortest diameter. The perisome contains calcareous granules. The species is represented from Mortensen’s station 26 by detached arms and groups of arms. The arm length is 125 mm. The specimen from Bock’s station 46 is young, with 12 arms 35-40 mm. long. The arms are slender, and the brachials are smooth, rather long, and slightly constricted centrally. The arm bases are well separated, leaving broad strips of the perisome 130 BULLETIN 82, UNITED STATES NATIONAL MUSEUM visible. The radials are rounded, smooth, and distally well separated laterally. The [Br, are three times as broad as long. The IBr, (axillaries) are twice as broad as long. The I[Br series are 2, both being young regenerates. The centrodorsal is almost completely concealed by the cirri. The cirri are XXIII, 10-14, from 3 to 7 mm. in length. The first segment is short, the second is about as long as broad and is centrally constricted, the third is half again as long as broad, and the fourth is twice as long as broad and is strongly constricted centrally. The succeeding segments are shorter, and the antepenultimate is one-quarter again as long as broad. The height of the opposing spine is equal to one-third the width of the segment bearing it. The terminal claw is a little shorter than the penultimate segment. P, is 4 mm. long, with 22 segments. P, is 4.2-5.5 mm. in length, with 20 segments. P3is 4.3 mm. long, with 19 segments. P, is 3.8 mm. long, with 15 segments. P, is 3mm. long. The length of the distal segments is twice the breadth. The basal segments have weak prom- inences. The distal pinnules are 5 mm. long, with 18 segments which are very long, being four to five times as long as broad. The disk is incised and bears calcareous granules along the ambulacral furrows. The anal cone is long and narrow, 1.5 mm. high. The three specimens from Bock’s station 47 are all young. In one the [Br, are twice as broad as long. The distal intersyzygial interval is three or four muscular articulations. In another the IBr, are one-quarter again as broad as long. The distal intersyzygial interval is 4-8 muscular articulations. In the third the IBr, are twice as broad as long. The distal intersyzygial interval is six muscular articulations. P, is 3.5 mm. long with 14 segments. The disk is “lean’’ and incised, 2-3 mm. in diameter. The anal funnel is 1.5 mm. high. In one of the specimens from Bock’s station 59 the centrodorsal is flat with the free dorsal pole granular, 3.5 mm. in diameter. The cirri are XXV, 20-21, from 17 to 22 mm. long, arranged in a partially double marginal row. The first and second segments are short, the third is about as long as broad, the fourth is a little longer, and the fifth and sixth are about one-third again as long as the distal width. The length of the following segments slowly decreases. The proximal segments are constricted centrally with their distal ends especially broadened. The antepenultimate segment is one-quarter again as long as broad. The opposing spine is inconspicuous, its height being equal to one-quarter the width of the segment which bears it. The terminal claw is pointed, and is about as long as the penultimate segment. The radials are partly concealed by the centrodorsal and are visible only as very narrow bands. The IBr, are four times as broad as long and are free laterally. The IBr, (axillaries) are low pentagonal, three times as broad as long. The I[Br and IIIBr series are 2. The 24 arms are all broken. The first 10 brachials are discoidal, and those following have oblique ends that are somewhat overlapping and are armed with small spines. There is a weak synarthrial tubercle on the articulation between the first and second brachials. Syzygies occur between brachials 3+4, 8+-9, and 14+-15 and distally at intervals of 4 or 5muscular articulations. P, is 5.5mm. long, with 22 segments. P, is about 10 mm. long, with 36 segments. P, is 13 mm. long, with 40 segments. Ps is about 9 mm. long, with 29 segments. Pj is 6 mm. long, with 15 segments. P, to about P, have the first-fifth or -sixth sezments provided with large prominences, which in height are equal to from one-third to one-fourth the width of the segments. The A MONOGRAPH OF THE EXISTING CRINOIDS 131 distal pinnules are 6 mm. long, with about 20 segments. The proximal segments are distally somewhat serrate and broadened. The disk is incised and measures 13 mm. in its longest and 8 mm. in its shortest diameter. It bears fine calcareous granules. The mouth is central. A second specimen from Bock’s station 59 has the centrodorsal as in the preceding, 4 mm. in diameter, with the dorsal pole studded with small papillae and 3 mm. across. The cirri are XXIV, 15-19, from 9 to 11 mm. long. The height of the opposing spine equals one-third that of the segment which bears it. The [Br and IIIBr series are 2. The arms, which are partly broken, are 21 (with perhaps 2 more) in number and 60 mm. long. The IBr, are three times as broad as long. The lateral borders of the IBr and IIBr series are slightly spiny. The intersyzygial interval is four or five muscu- lar articulations. P, is 3.5 to 4 mm. long, with 22-24 segments. P2, on a regenerating arm, is 5mm. long, with 25 segments. P,is 6 mm. long, with 22 segments. Only the first three pinnules are carinate. The distal pinnules are 4.5 mm. long, their segments being three times as long as broad. The disk is 6 mm. in diameter, and the anal cone is 1.7 mm. high. In the third specimen from Bock’s station 59 the centrodorsal is 5 mm. in diam- eter, with the radially furrowed dorsal pole 3.5 mm. in diameter. The cirri are XXII, 19-20, from 16 to 18 mm. long, and are arranged in a double row. The IBr, are eight times as broad as long and are free laterally. The IBr. (axillaries) are three times as broad as long. One of the IIBr series is 3, without a pinnule; the other IIBr series and the IIIBr series are all 2. The 23 arms are all broken. The proximal ossicles of the division series have three or four small lateral tubercles, which on the IBr,; orna- ment the whole proximal margin. A similar, though less distinct, ornamentation occurs on the IIBr; and on the IJIBr,. The ossicles immediately following the axillaries are interiorly united as far as the distal third. P, is about 5 mm. long, with 25 segments. The disk is incised and measures 12 mm. in its longest and 6 mm. in its shortest diameter. Gislén notes that this specimen differs from the two others from Bock’s station 59 in its lighter color and in the ornamentation of the proximal ossicles. The specimen from Bock’s station 61 is young. The JBr, are twice as broad as long. The intersyzygial interval is 5-7 muscular articulations. Gislén noted that in this, as in the specimen from station 46 and the three from station 47, the promi- nences on the proximal pinnules are indiscernible. Localities —Rotti Strait, between Rotti and Timor; 183 meters; from the Banju- wangi-Darwin No. 2 cable; cable repair ship Cable, Eastern and Associated Telegraph Co. [A. H. Clark, 1929] (2, B. M.). Danish Expedition to the Kei Islands station 27; 60-70 meters; fine sand; April 17, 1922 (1). Off Cape Padaran, southern Annam (lat. 11°38/N., long. 109°41’E.) ; 146 meters; from the Cape St. James-Hongkong cable; R. H. Ellis, cable repair ship Patrol, Eastern and Associated Telegraph Co., June 4, 1927 [A. H. Clark, 1929] (1, B.M.). Albatross station 5137; in the vicinity of Jolo (Sulu), Philippines; Jolo light bearing S. 61° E., 1.3 miles distant (lat. 6°04’25’’ N., long. 120°58’30’’ E.) ; 36 meters; sand and shells; February 14, 1908 [A. H. Clark, 1908, 1912, 1915, 1918; Gislén, 1922] (1, U.S.N.M., 25436). 132 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Albatross station 5139; in the vicinity of Jolo; Jolo light bearing S. 51° W., 3.6 miles distant (lat. 6°06’ N., long. 121°02’30’’ E.); 36 meters; coral sand; February 14, 1908 [A. H. Clark, 1908, 1912, 1915, 1918; Gislén, 1922] (1, U.S.N.M., 35172). Korean Straits, south of the Goto Islands (lat. 32°15’ N., long. 128°20’ E.); 183 meters; Capt. H. Christiansen, April 17, 1926 (5). Dr. Th. Mortensen’s Expedition to Japan station 10; off the Goto Islands (lat. 33°41’ N., long. 128°50’ E.); 137 meters; sand; May 17, 1914 [Gislén, 1927]. Dr. Th. Mortensen’s Expedition to Japan station 13; Korean Straits, east of Tsushima (lat. 34°20’ N., long. 130°10’ E.); 110 meters; sand and shells; May 18, 1914 [Gislén, 1927]. Dr. Th. Mortensen’s Expedition to Japan station 24; Sagami Bay; 914 meters; June 26, 1914 [Gislén, 1927]. Dr. Th. Mortensen’s Expedition to Japan station 26; Sagami Bay, off Okinose; 366 meters; hard bottom; July 1, 1914 [Gislén, 1914]. Dr. Sixten Bock’s Expedition to Japan station 46; Bonin Islands, east of the Channel; 128 meters; August 1, 1914 [Gislén, 1922, 1924]. Dr. Sixten Bock’s Expedition to Japan station 47; Bonin Islands, east of the Channel; 146 meters; August 1, 1914 [Gislén, 1922, 1924]. Dr. Sixten Bock’s Expedition to Japan station 59; eastnortheast of Anojima, Bonin Islands; 183 meters in published record, 150 meters on label of specimen examined; August 15, 1914 [Gislén, 1922, 1924] (1, U.S.N.M., E. 1114). Dr. Sixten Bock’s Expedition to Japan station 61; Bonin Islands, eastsoutheast of the Channel; 152 meters; August 16, 1914 [Gislén, 1922, 1924]. Geographical range.—From the Lesser Sunda Islands and southern Annam to the Philippines, southern Japan, and the Bonin Islands. Bathymetrical range-—From 36 to 914 meters; the average of 14 records is 188 meters. History —This species was originally described in 1908 from specimens collected by the Albatross in the Philippines earlier in the same year. In 1922 Dr. Torsten Gislén recorded and gave notes upon eight specimens dredged by Dr. Sixten Bock in the Bonin Islands and southwestern Japan and mentioned incidentally that there were other specimens in Dr. Th. Mortensen’s Japanese collec- tion. He found that the specimens from the Bonin Islands and southern Japan included in Bock’s and Mortensen’s collections differed from C. magnifica as originally described in having fewer arms, shorter cirri, and very much shorter proximal pinnules. He therefore proposed that they be recognized as representing a new variety, which he called minor. As in many species of crinoids individuals from certain localities in the Philippines and from some other localities in the East Indies are considerably larger with more numerous arms than those from elsewhere, it does not seem advisable, at least for the present, to recognize this new variety. In 1927 Gislén recorded and gave notes upon the six specimens in Mortensen’s collection from southern Japan, referring them to his new variety minor. A MONOGRAPH OF THE EXISTING CRINOIDS 133 CATOPTOMETRA RUBROFLAVA (A. H. Clark) Piars 8, Figure 27; Piare 9, Ficures 30, 33 Antedon rubroflava A. H. Cuarx, Proce. U. S. Nat. Mus., vol. 33, 1907, p. 127 (color in life), p. 150 (description; Albatross station 4880), p. 151 (comparison with A. [Catoptometra] hartlaubi); vol. 34, 1908, p. 493 (listed). Zygometra rubroflava A. H. Crarx, Smithsonian Mise. Coll., vol. 50, 1907, p. 348 (listed); Proce. U.S. Nat. Mus., vol. 34, 1908, p. 271 (relation to the species of EHudiocrinus). Catoptometra rubroflava A. H. Cuarx, Proce. U. S. Nat. Mus., vol. 34, 1908, p. 317 (listed); Smith- sonian Mise. Coll., vol. 52, 1908, p. 210 (compared with C. magnifica); Vid. Medd. Naturh. Foren. Kgbenhavn, 1909, p. 153 (off the Goto Island, 36 and 100 fathoms; description of specimens); Crinoids of the Indian Ocean, 1912, p. 106 (synonymy; habitat); Proc. Biol. Soe. Washington, vol. 26, 1913, p. 178 (range in eastern Asia), p. 181 (22°12’ N., 114°15/E., 14 fathoms; description of specimens) ; Journ. Washington Acad. Sci., vol. 5, 1915, p. 214 (southern Japanese species; range and its significance); Amer. Journ. Sci., vol. 40, 1915, p. 62 (listed) ; Unstalked crinoids of the Siboga-Exped., 1918, p. 63 (in key; range).—GisLin, Zool. Bidrag Uppsala, vol. 9, 1924, pp. 74, 77 (syzygies), fig. 67, p. 75 (syzygial face), fig. 97, p. 87 (synarthrial face); Vid. Medd. Dansk Naturh. Foren., vol. 83, 1927, p. 2 (stations 2, 10; 0 [?]}-135 m.), p. 21 (Mortensen’s stations 2, 10; notes), p. 68 (listed), figs. 9, 10, p. 15; Kungl. Fysiogr. Sallsk . HandL., new ser., vol. 45, No. 11, 1934, p. 21. Diagnostic features —The arms are 11-14 in number, the IIBr series being 4 (3-4), and are 100-180 mm. long; the cirri are short and stout, strongly and evenly recurved, with 15-20 segments, few or none of which are longer than broad. So far as is known the color in life is always bright yellow broadly and regularly banded with bright red. Except for the very different C. magnifica there is no other comatulid in the area where this occurs that is at all similar in color. Description —The centrodorsal is broad, discoidal, with the bare polar area concave. The cirrus sockets are arranged in two closely crowded marginal rows. The cirri are XXXV-XL, 15-20, stout, about 17 mm. in length. The first segment is about twice as broad as long, the second is about as long as broad, the third- eighth or third-tenth are slightly longer than broad with a strong central constriction and prominent ends, and those following become about as long as broad, the expan- sion of the ends gradually dying away. The penultimate segment bears a short and small, but prominent, opposing spine, which is subterminal in position but sometimes, owing to the occasionally rather strongly trapezoidal shape of the penultimate seg- ment, appears to be terminal. The opposing spine is directed diagonally outward. The terminal claw is considerably longer than the penultimate segment, moderately stout and moderately curved. Like the opposing spine it is semitransparent and glassy in strong contrast to the dull and brownish cirrus segments. The cirri are laterally compressed in the outer two-thirds. This disk is lacking. Large and closely crowded sacculi are found on the pinnules. The radials are very short in the midradial line but are visible as low triangles in the interradial angles of the calyx. The IBr, are very short, oblong, strongly rounded dorsally, and well separated laterally. The IBr, are broadly triangular, two and one-half or three times as broad as long, and about three times as long as the IBr,. The IIBr series are 4(3+4), in the type specimen only present on one ray. The 11 arms (in the type specimen) are 180 mm. in length, very long and slender. The first 9 or 10 brachials are oblong or slightly wedge-shaped, over twice as broad as 134 BULLETIN 82, UNITED STATES NATIONAL MUSEUM long, those following becoming triangular, not quite so long as broad, in the outer portion of the arm wedge-shaped and about as long as broad, and in the terminal portion elongated. All the brachials are evenly rounded dorsally with projecting distal edges which are armed with fine spines. Syzygies occur between brachials 3-++-4, again between brachials 9+ 10 (morerarely 10-+-11), and distally at intervals of 4-8 muscular articulations, the former being the usual number in the distal portion of the arms. In the two arms arising from the single I1Br series there are in both syzygies between brachials 3+-4 and 15+16, at which point one of the arms is broken off. In the other the following syzygies are distributed first at intervals of from 5 to 8 muscular articulations, soon becoming pretty uniformly 4 as in the arms arising from the [Br axillaries. Py is similar to P,, 8 mm. long, slender and flagellate, composed of about 30 seg- ments all of which are about as long as broad with the corners cut away; the second- fourth are furnished with thin rounded dorsal keels. The whole pinnule has an appearance recalling that of the oral pinnules in Heliometra. Pz, is 11-15 mm. long, flagellate, considerably stouter than P; but resembling it in general proportions; it is composed of 35-40 segments, which are about as long as broad, the first two or three with thin rounded dorsal keels. P3; and P, are similar, stouter than P,, 12-16 mm. in length, composed of short segments like the preceding pinnules, but with the carination of the basal segments less marked. The succeeding pinnules gradu- ally become shorter and less stout, reaching a minimum about the tenth or twelfth, after which they become still more slender and slowly increase in length. The distal pinnules are 12 mm. long, with about 35 segments, of which the first is trapezoidal and not so long as broad, the second is similar but rather longer, and the remainder are about twice as long as broad or rather longer with somewhat abruptly expanded articulations. The segments of all but the distal pinnules have prominent distal ends which are fringed with fine spines. The color in alcohol is uniform yellowish brown. Notes.—In the specimen from Mortensen’s station 2 as described by Gislén the centrodorsal is discoidal, 5.2 mm. in diameter, with the bare dorsal pole 4.8 mm. in diameter. The cirri are XXXVI, 15-19, from 11 to 16 mm. long. In the longest cirrus segments the ventral length is about equal to the width. The opposingspine is indistinct. The 14 arms are 125 mm. long. The IIBr series are 4(3+4). The IBr, are partly visible. There is a synarthrial tubercle on the articulation between the two outer elements of the I1Br series, and another on the articulation between the first two brachials. The distal intersyzygial interval is five or six muscular articula- tions. P, is 9.5 mm. long, with 32 segments. P, is 7 mm. long, with 28 segments. P; is 7 mm. long, with 23 segments. The distal pinnules are 8.5 mm. long, with about 22 segments. P, is 10 mm long, with 25 segments; and P, is 11 mm. long, with 27 segments. The basal segments of the proximal pinnules have inconspicuous flanges. Gislén said that this specimen differs from the type as described only in the slightly shorter P;, which has fewer segments. In one of the specimens from Mortensen’s station 10 the centrodorsal is 6 mm. in diameter. The cirri are XXXVIII, 15-19, from 9 to 15 mm. in length. The 11 arms are about 100 mm. long. The single IIBr series, which is developed on the right side of a ray, is 4(3+4). P, is 6 mm. long, with 25 segments. P, is 8.5 mm. A MONOGRAPH OF THE EXISTING CRINOIDS 135 long, with about 30 segments. P, is 10 mm. long, with 29 segments. The first six pinnules have a sharp keel on the first-fourth or -fifth segments, this feature becoming less pronounced on the more distal pinnules. Ps is 5.3 mm. long with about 18 segments. There are three or four sacculi to each pinnule segment. In the second specimen from Mortensen’s station 10 the centrodorsal is 3.8 mm. in diameter, with the bare dorsal pole 2 mm. in diameter. The cirri are XXIV, 16-19, from 9 to 16 mm. long. The radials are 10 times as broad as long, and the IBr, are eight times as broad as long. The 10 arms are all broken. Slight syn- arthrial tubercles are present. The specimen from off the Goto Islands in 80 meters has 11 arms about 40 mm. long. The single IIBr series is 2. Of the two specimens from off the Goto Islands in 66 meters, one is just passing from the 10-armed young stage to the adult condition. There are 13 arms 75 mm. long, three IIBr 4(3+4) series being present. Of the three IIBr series, two are of nearly full size while the third is very small, showing that adolescent autotomy had taken place not long before capture. In this specimen there is the same overlapping of the brachials and roughness of the dorsal surface of the arms as in the type, and the relative proportions of the proximal pinnules and of their component segments is similar, though the segments of the pinnules in the proximal third of the arm have rather more prominent distal ends and the lower segments of the first three or four pinnules are rather more strongly carinate. The cirri are XIX, and measure 13 mm. in length. As in the type they are arranged in one and a partial second crowded and irregular marginal row on the discoidal centrodorsal. The cirri show the char- acteristic somewhat abrupt expansion at the articulations, and consequently the same knobby appearance, as the cirri of the type, but the component segments are slightly shorter, none of them being quite so long as broad. The opposing spine, well marked as in the type, is characteristically small. The other specimen is small with 12 arms about 45 mm. long. The cirri are XX, and measure 12 mm. in length. They are arranged in a single marginal row on the centrodorsal. The centrodorsal is thin and discoidal, the slightly concave broad bare polar area 4 mm. in diameter. The best of the five specimens from 8 miles off Hongkong has 13 arms 100 mm. long, the 3 ITBr series being 4 (3 +4). The cirri are XXVII, 16-17, about 15 mm. long; the longest proximal cirrus segments are nearly or quite as long as the median width as viewed laterally. A second specimen has 12 arms about 110 mm. long, the 2 ILBr series being 4 (+4). The cirri are XXVIII, 15, about 13 mm. long. A third example also has 12 arms with the two IIBr series 4 (3+4). There are 14—16 cirrus segments. A small specimen has 10 arms 60 mm. long. The fifth specimen is much broken. ‘The color in life appears to have been bright yellow broadly banded with bright red, as in the type. These specimens agree with the type and with other from southern Japan with which they were compared. Localities—Dr. Th. Mortensen’s Expedition to Japan station 2; Misaki, shore (but probably taken in dredging) ; April, 1914 [Gislén, 1927]. Dr. Th. Mortensen’s Expedition to Japan station 10; off Kiu Shiu (lat. 33°41’ N., long. 128°50’ E.); 137 meters; sand; May 17, 1914 [Gislén, 1927]. 136 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Albatross station 4875; in the eastern channel of the Korean Straits, in the vicinity of the Oki Islands; Oki Shima bearing S. 24° W., 4 miles distant (lat. 34°19’ N., long. 130°09’ E.); 108 meters; fine gray sand and broken shells; August 2, 1906 (frag- ments, U.S. N. M., 36283). Albatross station 4880; in the eastern channel of the Korean Straits, in the vicinity of the Oki Islands; Oki Shima bearing S. 79° W., 7.5 miles distant (lat. 34°16’ 00’ N., long. 130°16’ E.); 108 meters; fine gray sand and broken shells; August 2, 1906 [A. H. Clark, 1907, 1908, 1909, 1912, 1913, 1915, 1918; Gislén, 1924] (1, U.S.N.M., 22639). Korean Straits, off the Goto Islands (lat. 33°10’ N., long. 129°18’ E.); 80 meters; Schgnau, November 15, 1895 (1, C. M.). Korean Straits, off the Goto Islands (lat. 33°09’ N., long. 129°18’ E.); 183 meters; Schgnau, May 23, 1898 [A. H. Clark, 1909, 1912, 1913, 1915, 1918; Gislén, 1924] (1, C. M.). Korean Straits, off the Goto Islands (lat. 33°08’ N., long. 129°20’ E.); 66 meters; Schgnau, March 14, 1891 [A. H. Clark, 1909, 1912, 1913, 1915, 1918; Gislén, 1924] (2, C. M.). Korean Straits, off the Goto Islands (lat. 32°48’ N., long. 129°37’ E.); 73 meters; bottom temperature 11.67° C.; Capt. H. Christiansen, cable repair ship Nordiske, April 23, 1912 (2, C. M.). Northeast of Swatow, China (lat. 23°15’ N., long. 117°40’ E.); Capt. H. Chris- tiansen, July 26, 1912 (2, C. M.). Eight miles outside of Hongkong Harbor (lat. 22°12’ N., long. 114°15’ E.); 25 meters; Captain Suensson, November 16, 1911 [A. H. Clark, 1913, 1915, 1918; Gislén, 1924] (5, U.S. N. M., EB. 1071; C. M.). Geographical range.-—From Sagami Bay to the Korean Straits and southward to Hongkong. Bathymetrical range.—From 25 to 183 meters; the average of eight records is 98 meters. Thermal range.—One record, 11.67° C. History.—This species was first described in 1907 from a single specimen dredged by the Albatross in the Korean Straits in 1906. In 1909 the present author recorded three specimens from two localities in the Korean Straits, where they had been dredged by Danish cable repair ships under the command of Captain Suensson, and in 1913 he recorded five more from near Hongkong, received from the same source. In 1927 Dr. Torsten Gislén recorded and gave notes upon three specimens which had been collected by Dr. Th. Mortensen in southern and southwestern Japan. CATOPTOMETRA OPHIURA A. H. Clark Puate 9, Ficures 29, 32 Catoptometra ophiura A. H. Cuark, Proc. U. 8S. Nat. Mus., vol. 39, 1911, p. 589 (description; Alba- tross stations 5356 and 5414; comparison with C. hartlaubi) ; Crinoids of the Indian Ocean, 1912, p. 106 (synonymy; habitat)—IF. W. Ciarke and Wuee.er, U. S. Geol. Surv. Prof. Pap. 90-D, 1914, pp. 33 ff. (inorganic constituents of skeleton).—A. H. Cuarx. Amer. Journ. Sci., vol. 40, 1915, p. 62 (listed).—F. W. CLarks and Wuerter, U. S. Geol. Surv. Prof. Pap. 102, 1917, pp. 20 ff. (Albatross station 5036; inorganic constituents of the skeleton).—A. H. Cuarkx, A MONOGRAPH OF THE EXISTING CRINOIDS 137 Unstalked crinoids of the Siboga Exped., 1918, p. 63 (in key; range), p. 64 (synonymy; notes; stations 49a, 294), pp. 271, 275 (listed)—F. W. Cuarkre and Wuee.emr, U. S. Geol. Surv. Prof. Pap. 124, 1922, p. 17 (inorganic constituents of the skeleton)—Gistin Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 21, 27. Diagnostic features —The arms are 20-27 in number, the IIBr series being 4(3-+4), 180 mm. long in fully grown individuals, and the cirri are composed of 15-19 (usually 17-19) segments of which the longest are twice as long as the median width and the outermost are half again as long as broad. In its general appearance this species most nearly resembles C. rubroflava, but the greater number of arms and the longer cirri, which are composed of longer segments, easily distinguish it from that form. In color it appears to be usually olive-green or brown, but occasionally red and yellow like C. rubroflava. Description.—The centrodorsal is thin discoidal with a broad moderately concave polar area 5-8 mm. in diameter marked in the center by a shallow rounded pit. The cirrus sockets are arranged in two closely crowded irregular marginal rows. The cirri are XXVII, 15-18 (usually 17), 22 mm. long. The first segment is about twice as broad as long, the second is not quite so long as broad, the third is half again as long as the median width, the fourth and fifth are twice as long as the median width, and those following gradually become very slightly shorter so that the terminal segments are half again as long as broad. The dorsal and ventral profiles of the longer proximal segments, especially the former, are rather strongly concave, this feature gradually dying away distally. The lateral profiles of these longer proximal segments are very strongly concave, this character gradually becoming less marked but persist- ing to the end of the cirri. The penultimate segment is about one-third again as long as broad. The opposing spine is terminal, small and inconspicuous. The terminal claw is half again as long as the penultimate segment, slender and slightly curved. The radials are only just visible in the interradial angles of the calyx. The IBr, are oblong, very short and bandlike. The IBr, (axillaries) are triangular, nearly three times as broad as long. The [Br series are well separated laterally. The IIBr series are 4(3++-4). The IIBr, are interiorly united for the proximal half or two-thirds, their interior sides diverging from that point so as to make a right angle with each other. The IIIBr series, when present, are 2, always developed interiorly in 1, 2, 2, 1 order. The synarthrial tubercles are prominent and sharp. The arms are 20-27 in number, 180 mm. in length. The first brachials are wedge-shaped, about twice as broad as long exteriorly, interiorly united for about the proximal two-thirds. The second brachials are of about the same size but more ob- liquely wedge-shaped. The first syzygial pair (composed of brachials 3 +-4) is slightly longer interiorly than exteriorly, about twice as broad as the median length. The next five brachials are oblong or slightly wedge-shaped, two to three times as broad as long, and those following become triangular and as long as broad, distally obliquely wedge-shaped, and terminally longer than broad. The distal borders of the tenth and following brachials are moderately overlapping and finely spinous. Syzygies occur between brachials 3-++-4, again from between brachials 19+ 20 to between brachials 36+37 (usually about brachials 27+-28), and distally at intervals of 8-16 (usually 10 or 11) muscular articulations. 208244—40——10 138 BULLETIN 82, UNITED STATES NATIONAL MUSEUM P, is small and slender, 6 mm. long, with 25-30 segments, most of which are about as long as broad; the second-fifth bear strong dorsal projections. P, is 7-8 mm. long, slender, the second-fourth segments with dorsal projections which are not so large as those on the basal segments of P,. P», resembles P, but is 15 mm. long, with 40-42 segments, which are about as long as broad, the first-fourth with slight dorsal projec- tions. P,issimilar to P,,12mm.long. P,is similar to P;, 11mm. long. Ps is similar to Py, 11 mm. long. The following pinnules gradually decrease in length to 7 mm., then slowly increase to 10 mm. distally. The carination of the proximal segments of the lower pinnules becomes progressively less and less, disappearing entirely after the tenth or twelfth. Notes.—One of the specimens from Siboga station 49a has the centrodorsal broad and flat, discoidal, with the dorsal pole 6.5 mm. in diameter, marked with faint radiating lines and pitted in the center. The cirri are XXX, 17-19, 23 or 24 mm. long. The cirri, and in fact the whole animal, are less robust than is the case in the specimens in the type series. The other specimen from Siboga station 49q is slightly smaller, though exactly similar. The dorsal pole of the centrodorsal is 6 mm. in diameter. The cirri are XXVII, 17, 20 mm. long. In these specimens the arm bases, centro- dorsal and cirri are light flesh color, the latter becoming pinkish distally; the arms beyond the second syzygy are bright yellow orange regularly banded with bright red as in C. rubroflava and in C. magnifica. The specimen from Siboga station 294 has 13 arms. . The dorsal pole of the centro- dorsal is 3 mm. in diameter. The cirri are XX, 16, from 20 to 22 mm. long. The three I1Br series are 4(3+-4). This example resembles the two preceding, but the centrodorsal is only half as broad, consequently the cirri appear shorter and more of the division series is visible. The four specimens recorded in 1911 from Albatross station 5356 had 20, 26 (2), and 27 arms. The specimen from Albatross station 5414 is small, with 11 arms about 90 mm. long. Localities—Siboga station 49a; Sapeh Strait, between Sumbava and Komodo (lat. 8°23'30’" S., long. 119°04’36’’ E.); 69 meters; coral and shells; April 14, 1899 [A. H. Clark, 1918] (2, Amsterdam Mus.). Siboga station 294; off southwestern Timor (lat. 10°12’12’’ S., long. 124°27/18” E.); 73 meters; soft mud with very fine sand; January 23, 1900 [A. H. Clark, 1918] (1, Amsterdam Mus.). Albatross station 5356; North Balabac Strait; Balabac light bearing S. 64° W., 15.5 miles distant (lat. 8°06’40’’ N., long. 117°18’45’’ E.); 106 meters; sand and shells; January 5, 1909 [A. H. Clark, 1911, 1912, 1918; F. W. Clarke and Wheeler, 1914, 1917] (6+, U.S. N. M., 27490, 35131, 35168, 35171, 35175, 35239). Albatross station 5414; between Cebu and Bohol; Lauis Point light bearing N. 67° W., 9.5 miles distant (lat. 10°10’40’” N., 124°02’45’’ E.); 77 meters; coral and sand; March 24, 1909 [A. H. Clark, 1911, 1912, 1918] (1, U. S. N. M., 35170). Geographical range.—From the Lesser Sunda to the Philippine Islands. Bathymetrical range-—From 69 to 106 meters; the average of four records is 81 meters. A MONOGRAPH OF THE EXISTING CRINOIDS 139 History—This species was first described in 1911 from five specimens collected by the Albatross in the Philippines in 1909, four of them at station 5356 and a small individual at station 5414. In 1918 three additional specimens were recorded by the author from two Siboga stations, CATOPTOMETRA HARTLAUBI (A. H. Clark) Puate 8, Figure 28; Pyare 9, Figure 34; Puate 11, Ficures 43, 44 {See also vol. 1, pt. 1, fig. 251 (centrodorsal), p. 253; fig. 334 (cirrus), p. 283; fig. 467 (centrodorsal), p. 359; pt. 2, figs. 32, 33 (radial pentagon), p. 20.] Antedon hartlaubi A. H. Cuarx, Proc. U. S. Nat. Mus., vol. 33, 1907, p. 72 (description; Albatross station 4934); vol. 34, 1908, p. 475 (listed). Zygometra koehlert A. H. Cuarx, Smithsonian Mise. Coll., vol. 50, 1907, p. 339 (description; Albatross station 3717), p. 348 (listed); Proc. U. S. Nat. Mus., vol. 34, 1908, p. 271 (relation to species of Eudiocrinus), p. 542 (listed); Unstalked erinoids of the Siboga-Exped., 1918, p. 63, footnote (synonym of C. hartlaubi). Zygometra hartlaubi A. H. Cuark, Smithsonian Mise. Coll., vol. 50, pt. 3, 1907, p. 348 (listed) ; Proc. U. S. Nat. Mus., vol. 34, 1908, p. 271 (relation to species of Eudiocrinus), p. 276, figs. 6, 7 (centrodorsal and radial articular faces). Catoptometra hartlaubi A. H. Cuarx, Proc. U. S. Nat. Mus., vol. 34, 1908, p. 317 (listed); vol. 39, 1911, p. 540 (compared with C. ophiura); Crinoids of the Indian Ocean, 1912, p. 21 (attains greatest depth of any species of the family), p. 106 (synonymy; habitat); Journ. Washington Acad. Sci., vol. 5, 1915, p. 214 (southern Japanese species; range and its significance); Amer. Journ. Sci., vol. 40, 1915, p. 62 (listed); Unstalked crinoids of the Siboga-Exped., 1918, p. 63 (in key; range).—GisL&in, Nova Acta Reg. Soc. Sci. Upsaliensis, ser. 4, vol. 5, No. 6, 1922, p. 4 (245 m.), p. 6 (Kiu Shiu and Goto Islands), p. 64 (Bock’s station 5; detailed notes), p. 180 (listed), figs. 48-51, p. 70.—Gistin, Vid. Medd. Dansk Naturh. Foren., vol. 83, 1927, p. 2 (Mortensen’s station 5; 243 m.), p. 20 (station 5; notes), p. 22 (compared with C. rubroflava), p. 68 (listed); Kungl. Fysiogr. Sillsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 21, 29. Antedon hartlanbi S. and R. M. Pacer, Zool. Rec., 1907, V. Echinoderma, Dec. 1908, p. 38 Ech. Catoptometra koehleri A. H. Cuark, Proc. U.S. Nat. Mus., vol. 34, 1908, p. 317 (listed) ; Crinoids of the Indian Ocean, 1912, p. 107 (habitat). Diagnostic features —The arms are 20-23 (usually 20) in number and 70 to 120 mm. in length; the I[Br series are 4(3-+-4); the cirri are composed of 14-17 (usually 14-15) segments of which the fourth and following are all about twice as long as broad. This species is easily distinguished from C. rubroflava, which occurs in the same region, by the greater number of arms and by the longer and more slender cirri, which are recurved only in the distal half and are composed of much longer segments. From C. ophiura it is distinguished by the more slender and less regularly recurved cirri, which have longer distal segments, by the prominently everted spinous distal ends of the elements of the division series and first two brachials, and by the papillose or finely spinous dorsal pole of the centrodorsal. Description.—The centrodorsal is a thick disk, with the large bare polar area flat or very slightly concave and beset with well separated small spines. The cirrus sockets are arranged in 2 crowded rows of about 15 each. The cirri are XXX, 15,19 mm.long. The first segment is twice as broad as long, the second is somewhat longer, the third is rather longer than broad, and the fourth and following are all approximately twice as long as broad. The ends of the segments in the proximal half of the cirri are slightly enlarged, this feature gradually dying 140 BULLETIN 82, UNITED STATES NATIONAL MUSEUM away distally. The opposing spine is represented by a minute tubercle, which may be obsolete. The terminal claw is as long as the penultimate segment, or rather less, and is moderately stout and moderately curved; it is semitransparent and glassy, in contrast to the cirrus segments which are dull and opaque. The disk is lacking. The sacculi on the pinnules are small but numerous. The radials are concealed by the centrodorsal. The IBr, are also concealed by the centrodorsal. The IBr, (axillaries) are very broadly triangular. The IIBr series are 4 (3-+4), with the distal ends of the component segments everted and finely serrate. The 20 arms are 120 mm. in length. The first 9 or 10 brachials are oblong or slightly wedge-shaped, rather more than twice as broad as long, those following be- coming triangular, nearly as long as broad, in the distal portion of the arm wedge- shaped again, and in the terminal portion elongate. All the brachials have slightly overlapping and finely spinous distal ends. Syzygies occur between brachials 3+-4, again about brachials 16+ 17, and distally at intervals of five to eight muscular articulations. P, is 4 mm. long, small, weak, and slender, becoming very slender and flagellate in the distal half. It is composed of about 20 segments of which the first four or five are broader than long and the remainder are about as long as broad; the second-fourth or -fifth have rounded triangular dorsal keels, which in shape almost suggest over- lapping spines. The basal segments have prominent and finely spinous distal ends, and the dorsal carinate processes are covered with fine spines. P, is longer and stouter than P,, though not quite so long as P,. P, is the longest pinnule, exactly like P, and P, butslightly stouter basally, 12 mm. long, with about 40 segments, of which the basal five or six are short and the remainder are about twice as long as broad. The second and third segments in P;, P,, and P; have an indicated dorsal process similar to that on the basal segments of P,. The following pinnules gradually decrease in length to about 7 mm. though remaining similar to the lower pinnules except that the tip is not so slender, then gradually increase in length distally. The distal pinnules are 11 mm. long with the first segment short and approximately oblong, about twice as broad as long, the second about as long as broad, and those following becoming very rapidly elongated and very long and exceedingly slender distally. The elongated segments have somewhat abruptly expanded articulations, and all the segments, especially the lower, have the ends more or less spinous and often scattered spines on the dorsal surface. Notes.—The eight specimens from off the Goto Islands are all much broken. The centrodorsal is discoidal, moderately thick, that of the largest being 6.2 mm. in basal diameter. The dorsal pole is flat or very slightly concave, in the largest being 4 mm. in diameter. The cirrus sockets are arranged in one and a more or less complete sec- ond irregular closely crowded marginal rows. Interradially the centrodorsals are 1.5-2 mm. high. The lateral borders slope inward only very slightly. The longest cirri are 19 mm. long with 16 segments, which are subequal, one-third again to twice as long as broad. All the specimens appear to have had 20, or about 20, arms. Those of the largest are about 140 mm. long. As in the type specimen, the distal edges of the ossicles of the division series are everted and spinous. A small specimen has 10 arms between 55 and 60 mm. long and the cirri 8 mm. long with 10-12 segments. A MONOGRAPH OF THE EXISTING CRINOIDS 141 The specimen from Bock’s station 5 was thus described by Gislén: The centro- dorsal is large, flattened, slightly concave in the center, with calcareous granules which toward the margin are arranged in about 10 rays. It is 5 mm. in diameter and 0.8 mm. high, with the bare dorsal pole 3.5 mm. in diameter. The cirri are XXVIII, 14-17, from 11 to 17 mm. long. The shorter dorsal cirri have 14 or 15 and the longer ventral have 16 or 17 segments. The cirri are arranged in two closely crowded alter- nating rows. The second segment is about as long as broad, and the third is a little longer and is slightly constricted centrally, as are most of the segments following. The fifth-seventh segments are twice as long as broad, and those succeeding are from one-half to one-quarter again as long as broad, and are without dorsal spines. The penultimate segment is one-fifth again as long as broad with a median very weak opposing spine the height of which is equal to about one-fifth the width of the segment. The terminal claw is slender, somewhat curved, and is longer than the penultimate segment. The radials and the IBr; are concealed by the centrodorsal except for the distal edge of the latter. The IBr, are united laterally. The IBr, (axillaries) are tri- angular, four times as broad as long. The IIBr series are 4(3+4). The IIBr, are twice as long exteriorly as interiorly and are united interiorly. There is a synarthrial tubercle on the articulation between the first two elements of the I[Br series, and another inconspicuous one on the articulation between the first two brachials. The 20 arms (or perhaps 19, as one of the IIBr series is broken off at the syzygy between the two outer elements) are 70 mm. long. The first-eighth brachials are discoidal. The proximal brachials are short, and the outer are about as long as broad. All the distal brachials are somewhat compressed in the middle, with the distal ends everted and armed with small spines. There are 17 brachials in each 10 mm., or 15 if the syzygial pairs are counted as units. On a typical arm syzygies occur between brachials 3+4, 11+12, 21+22, 30+31, and distally at intervals of about seven muscular articulations. The pinnule on the IIBr series is 7 mm. long with 29 segments, of which the first seven are stout and those following more slender. P, is 13 mm. long with 33 segments, which are a little longer than broad, the second-fourth with small prom- inences resembling those on the proximal pinnules of, for instance, Comatula pectinata, but very much less developed. P, is of about the same length. P; is about 10 mm. long. P, is 5.5 mm. long, with about 20 segments. The genital pinnules are 12 mm. long. The distal pinnules are 8 mm. long, with 20 segments. The basal pinnule segments, especially on the proximal pinnules, have small spiny prominences. There are three pairs of small sacculi to each pinnule segment. The disk is lacking. The color in formalin is white. Dr. Gislén remarked that this specimen comes closest to Catoptometra hartlaubi from which it differs in the occurrence of an opposing spine and in the pinnules which, compared with the arm length, are somewhat longer. In the specimen from Mortensen’s station 5, as described by Gislén, the cirri are XXXII, 14-16, from 11 to 15 mm. long. The penultimate segment bears an indistinct opposing spine. The 23 arms are 80+ mm. long. Of the 10 IIBr series eight are 4(3-+4) and two are 2(1+2). Of the five IIIBr series three are 4(3-+-4) and two are 3, both of the last on a I[Br 2 series. With 10 I1Br and five IIIBr series present the number of arms must be 25 instead of 23, as given by Gislén. The distal border of the brachials is rather strongly everted and serrate. The distal intersyzygial interval is five or six, or eight, muscular articulations. The pinnule on the IIBr series is 142 BULLETIN 82, UNITED STATES NATIONAL MUSEUM 6.5 mm. long, with about 23 segments. P, is 9.5 mm. long, with about 30 segments. Some of the proximal cirrus segments, except the first, have an insignificant flange. The disk is detached. It is 9 mm. in diameter. The mouth is central, slitlike, extending from the anterior interradius toward the posterior radius. The perisome is naked except for insignificant granules along the ambulacral furrows. The specimen from Albatross station 3717 originally described under the name of Zygometra koehleri presents the following characters: The centrodorsal is a flat disk. The marginal cirri are about XX, 10-15, short and rather stout. The first two segments are short, and those following are longer than broad and somewhat constricted centrally. The opposing spine is very small, and the terminal claw is short and curved. The radials are just visible beyond the edge of the centrodorsal, and are free distally. The LBr, are nearly three times as broad as long and are united to the low triangular IBr, (axillaries) by syzygy. The IBr series are dorsally rounded and are widely free laterally. The 10 arms are 45 mm. long and are composed of 65-80 brachials, of which the first seven are oblong or slightly wedge-shaped, broader than long, those following are wedge-shaped and as long as or longer than broad, and the distal become elongate. Syzygies occur between brachials 3+-4, 9+10 and 14+15, and distally at intervals of four or five muscular articulations. P, is long and slender, with more than 20 segments, of which the basal three or four are broader than long and the remainder are elongate. P, is distinctly, and often much, shorter, with fewer segments. P, is the longest, with about 25 elongated segments. The following pinnules are not much shorter but are very much more slender. The anal tube and the posterior interradius are heavily plated, but the rest of the disk is almost naked. The color in life was bright yellow. Localities.—Korean Straits, off the Goto Islands (lat. 32°20’ N., long. 128°15’ E.); 201 meters; bottom temperature 11.67° C.; Captain Suensson, May 17, 1911 (8, U.S.N.M., E. 1088; C.M.). Albatross station 4933; Eastern Sea, off Kagoshima Gulf; Sata Misaki light bearing N. 84° E., 8.5 miles distant (lat. 30°59’00” N., long. 130°29’50” E.); 278 meters; rocky bottom; August 16, 1906 (1, U.S.N.M., 35169, 36232). Albatross station 4934; Eastern Sea, off Kagoshima Gulf; Sata Misaki light bearing N. 77.5° E., 7 miles distant (lat. 30°58’30” N., long. 130°32’00” E.); 188-278 meters; rocky bottom; August 16,1906 [A. H. Clark, 1907, 1912, 1918] (1, U.S.N.M., 22606). Dr. Sixten Bock’s Expedition to Japan station 5; off Kagoshima; 247 meters; May 13, 1914 [Gislén, 1922]. Dr. Th. Mortensen’s Expedition to Japan station 5; off Kiu Shiu, 9 miles W. to N., % N., of Bonomisaki; 247 meters; sand and shells; May 13, 1914; presumably identical with the locality just preceding [Gislén, 1927]. Albatross station 3704; Sagami Bay; Ose Zaki bearing S. 53° W., 2.25 miles distant; 115-137 meters; volcanic sand and gravel; May 8, 1900 (1, U.S.N.M., 35134). Albatross station 3717; Sagami Bay; Ose Zaki bearing S. 34° E., 0.8 mile distant; 137-183 meters; volcanic sand, shells and rock; May 11, 1900 [A. H. Clark, 1907, 1912, 1918] (1, U.S.N.M., 22660). Geographical range.—Southern Japan, from the Korean Straits to Sagami Bay. Bathymetrical range—From 137 (?115) to 278 meters; the average of six records is 196 meters. A MONOGRAPH OF THE EXISTING CRINOIDS 143 Thermal range.—One record, 11.67° C. History.—This species was described originally under the name of Antedon hartlaubi in 1907 from a specimen dredged by the Albatross in the Korean Straits in 1906. Later in the same year another young individual from Albatross station 3717 in Sagami Bay was described as Zygometra koehleri. In 1918 in the report upon the Siboga collection koehleri was placed in the synonymy of hartlaubi. In 1922 Dr. Torsten Gislén described in detail a specimen which had been collected by Dr. Sixten Bock off Kagoshima, and in 1927 he described another from Dr. Th. Mortensen’s collection, which apparently was taken at the same time and place. Family EUDIOCRINIDAE A. H. Clark Eudiocrinidae A. H. Cuarx, Smithsonian Mise. Coll., vol. 50, 1907, p. 344 (in part; new family to include Eudiocrinus and Decametrocrinus); Proc. Biol. Soc. Washington, vol. 21, 1908, p. 135 (name replaced by Pentametrocrinidae, the genus Eudiocrinus being transferred to the Zygo- metridae); Proc. U. S. Nat. Mus., vol. 34, 1908, p. 210 (in part; in key; includes Eudiocrinus and Decametrocrinus; range), p. 211 (represented in Hawaiian Islands), p. 212 (range of com- ponent genera); Rec. Indian Mus., vol. 34, pt. 4, 1932, p. 554 (including Eudiocrinus only); Temminckia, vol. 1, 1936, p. 300; John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1936, pp. 99, 103. Zygometridae (part) A. H. Cuark, Proc. Biol. Soc. Washington, vol. 21, 1908, p. 135 (includes Zygometra and Eudiocrinus). [For subsequent references see page 76.) Diagnosis —A family of the superfamily Mariametrida in which the elements of the IBr series are united by syzygy and the IBr, is not axillary, but bears a pinnule, the arms being therefore only five in number. The perisome of the disk is completely covered by a pavement of irregular plates (see Part 2, p. 226, fourth paragraph), and the lateral perisome of the pinnules contains an inner and an outer row of spicules (see Part 2, p. 238, and figs. 787 and 788, p. 366). Geographical range-—From the Maldive Islands, Cape Negrais, Burma, and the Andaman Islands to the Lesser Sunda and Kei Islands, the Moluccas, the Philippine and Bonin Islands, and southern Japan. Bathymetrical range—From 36 to 183 meters. Remarks.—The possession of only five arms easily distinguishes the species of the family Eudiocrinidae from all other comatulids except the species of Pentametrocrinus in the family Pentametrocrinidae and the species of Atopocrinus in the family Atele- crinidae of the suborder Macrophreata (see Part 5). In the Eudiocrinidae the first two postradial ossicles are the equivalent of the [Br series in comatulids with 10 or more arms, but the IBr is not axillary, simply bearing a pinnule. The third postradial ossicle is the equivalent of the first brachial of the free undivided arms in other comatulids, and consequently does not bear a pinnule. In the Pentametrocrinidae the IBr series are absent, the first ossicle following the radials representing the first brachial of the free undivided arms in other comatulids. It is interesting that in the Eudiocrinidae the sequence of the pinnules is that seen on the free undivided arms in other comatulids—that is, P, and P, are similar, forming the first pair; P, and P, are similar, forming the second pair; and so on—whereas it might be expected that the pinnule of the IBr series (Pc) would be an odd pinnule and that the grouping of the pinnules in pairs would commence with P;, as is the case in other comatulids with pinnules on the division series. 144 BULLETIN 82, UNITED STATES NATIONAL MUSEUM In the occurrence of a syzygy between the elements of the IBr series and in the plating of the disk the Eudiocrinidae agree with the Zygometridae. These two families also agree in the structure of the articular faces of the radials, the agreement between Eudiocrinus (see Part 2, p. 22, figs. 29, 30, p. 20) and Catoptometra (see Part 2, p. 21, figs. 32, 33, p. 20) being especially close. In those species of Zygometridae in which adambulacral deposits are found in the lateral perisome of the pinnules these deposits are of the type characteristic of the Eudiocrinidae, although they are more or less reduced. In the Budiocrinidae the disk is more heavily plated than it is in the Zygometridae, the adambulacral rods are always large, and there are large filmy plates over the gonads. It may be remarked that in the genus Mariametra of the family Mari- ametridae the disk is always well plated, although there are no other features to suggest affinity with the Zygometridae or Eudiocrinidae. The Zygometridae and the Eudiocrinidae, though quite distinct, possess in com- mon a number of features that clearly indicate a rather close relationship and at the same time set them off rather sharply from the 3 other families in the Mariametrida— the Himerometridae, Mariametridae, and Colobometridae—which similarly are closely allied. History —The family name Eudiocrinidae was first proposed by me in 1907 to include the genera Fudiocrinus and Decametrocrinus. The genus Hudiocrinus was accepted in the sense in which it was used by Carpenter—that is, as including the species now assigned to it and also the species now referred to Pentametrocrinus. At that time I had at hand many specimens of Carpenter’s Hudiocrinus (Penta- metrocrinus) varians and E. (P.) japonicus, which I had collected in southern Japan in the previous year, but I had never seen a specimen of Hudiocrinus indivisus or of any closely related species. My concept of the genus Eudiocrinus was therefore based entirely on species of Pentametrocrinus, and the family Eudiocrinidae as proposed was in intent the equivalent of the family Pentametrocrinidae as now understood. Shortly after the publication of this paper I acquired a specimen of a species related to Hudiocrinus indivisus—the type specimen of EF. variegatus—and a study of this individual showed that the group of species represented by E. indivisus is entirely different from the group represented by £. varians and E. japonicus. In a paper published on April 11, 1908, I established the new genus Pentametrocrinus for E. varians and its allies and created the new family Pentametrocrinidae for the reception of Pentametrocrinus and Decametrocrinus. I showed that Hudiocrinus, as represented by E. indivisus, is characterized by the possession of [Br series of which the IBr, is not axillary, which are wholly unrepresented in £. varians and its allies (the genus Pentametrocrinus), and that it is most closely related to the genus Zygometra. I therefore created the new family Zygometridae to include the genera Zygometra (which at that time included also Catoptometra) and Eudiocrinus. In a paper published later in the same year (May 14), but written before the receipt of the type specimen of Hudiocrinus variegatus, the family Eudiocrinidae appears as including Eudiocrinus and Decametrocrinus, and the family Zygometridae as including Zygometra only. Until the present time the genus Hudiocrinus has remained in the family Zygo- metridae to which it was assigned in 1908. A MONOGRAPH OF THE EXISTING CRINOIDS 145 Genus EUDIOCRINUS P. H. Carpenter Ophiocrinus (not of Salter, 1856, nor of Angelin, 1878) Sumpxr, Archiv fiir Naturg., 1868, vol. 1, p. 68 (diagnosis; genotype O. indivisus).—P. H. Carpenter, Nature, vol. 15, 1877, p. 198.— ScuLiter, Zeitschr. deutsch. geol. Ges., Jahrg. 1878, p. 40.—P. H. Carpenter, Nature, vol. 19, 1879, p. 450; Trans. Linn. Soc. (Zool.), ser. 2, vol. 2, 1879, p. 17; Proc. Roy. Soc., No. 194, 1879, p. 385.—pE Lorrov, Monogr. des crinoides fossiles de la Suisse, 1879, p. 227.—Zi11TEL, Handbuch der Palaeontologie, vol. 1, 1876-1880, p. 346 (referred to the Comatulidae), p. 397 (a recent type).—P. H. Carpenter, Quart. Journ. Geol. Soc., vol. 36, 1880, p. 41; Bull. Mus. Comp. Zool., vol. 9, 1881, p. 161 (p. 11 of separate) (sacculi).—PERRIER, Compt. Rend. Acad. Sci., vol. 96, 1883, p. 725.—A. H. Cuarx, Proc. U. S. Nat. Mus., vol. 34, 1908, p. 271 (history and status). Comatula (Ophiocrinus) pn Loriou, Denkschr. allgem. schweiz. Ges. ges. Naturw., vol. 23, 1869, p. 57. Eudiocrinus P. H. Carpenter, Journ. Linn. Soc. (Zool.), vol. 16, 1882, p. 493 (new name for Ophio- crinus Semper, 1868, preoccupied); Proce. Roy. Soc., vol. 35, 1883, p. 138.—Perrriur, Compt. Rend. Acad. Sci., vol. 96, 1883, p. 725; Ann. Mag. Nat. Hist., ser. 5, vol. 11, 1883, p. 294.— P. H. Carrenrer, Phil. Trans. Roy. Soc., vol. 174, pt. 3, 1883 (1884), p. 926; Challenger Re- ports, Zoology, vol. 11, pt. 32, 1884, pp. 36, 47, 58, 68, 137, 138, 140, 143, 291, 378; Narrative, vol. 1, pt. 1, 1885, p. 310.—Prrrier, Explorations sous-marines, 1886, p. 275.—P. H. Car- PENTER, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 73.—DE Lorion, Paléontologie francaise, Terrain jurassique, vol. 11, pt. 2, 1889, p. 484.—Prrrimr, Traité de zoologie, 1893, p. 784.—BatueEr, Proc. Zool. Soc. London, 1895, p. 995.—Lane, A text book of comparative anatomy, vol. 2, 1896, p. 313.—BarTuer, in Lankester, A treatise on zoology, pt. 3, Echino- derma, 1900, p. 195.—Drtace and H&rovarp, Traité de zoologie concréte, vol. 3, 1903, p. 394.—Mincxert, Archiv fir Naturg., Jahrg. 71, 1905, vol. 1, p. 166.—A. H. Cuarx, Proc. U. 8. Nat. Mus., vol. 32, 1907, p. 569 (found abundantly off southern Japan by the Albatross; refers to Pentametrocrinus), p. 573 (list of known species with original references and type localities; all but one [indivisus] are species of Pentametrocrinus).—HAMANN, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1575 (listed) —A. H. Cuarx, Smith- sonian Misc. Coll., vol. 50, 1907, p. 343 (included by Carpenter in the family Comatulidae), p. 344 (most nearly related to Decametrocrinus; here united with it in the family Eudiocrinidae) ; Proc. Biol. Soc. Washington, vol. 21, 1908, p. 134 (restricted by the removal of 5 species to Pentametrocrinus), p. 135 (assigned to the family Zygometridae); Proc. U. 8S. Nat. Mus., vol. 34, 1908, p. 210 (referred to the Eudiocrinidae), p. 212 (occurs in West Indies and Japan [refers to Pentametrocrinus]), p. 269 (arm structure), p. 271 (history; structure; included species dis- tributed among 2 genera, Eudiocrinus and Pentametrocrinus, the former assigned to the Zygo- metridae, the latter to the Pentametrocrinidae), p. 276 (revised diagnosis; list of species), pp. 307, 436, 519; vol. 35, 1908, p. 113; fig. 4, p. 118; p. 125 (arm structure); Proc. Biol. Soc. Washington, vol. 22, 1909, p. 2 (cirri and lower pinnules compared with those of Charito- metridae); Vid. Medd. Naturh. Foren. Kgbenhavn, 1909, p. 152 (cirri and brachials compared with those of Zygometra fluctuans [comata]), p. 193 (probably occurs at Singapore though not yet discovered there); Mem. Australian Mus., vol. 4, 1911, p. 725 (absent from Australia) ; Crinoids of the Indian Ocean, 1912, p. 9 (absent from Australia), p. 11 (occurs in the Anda- mans), p. 21 (range); p. 56 (in key), p. 99 (synonymy; type).—REICHENSPERGER, Zeitschr. fiir wiss. Zool., vol. 101, Heft 1-2, 1912, p. 32 (sacculi); A. H. Cuarx, Die Crinoiden der Antarktis, 1915, p. 121 (belongs to the Zygometridae, though with only 5 arms); Amer. Journ. Sci., vol. 40, 1915, p. 62 (list of recent species), p. 63 (number of recent and fossil species; range), p. 65 (confined to western Pacific); Smithsonian Mise. Coll., vol. 65, No. 10, 1915, pp. 32 ff. (phylo- genetic study); Unstalked crinoids of the Siboga-Exped., 1918, p. VIII (discovery of 3 remark- able species by the Siboga); p. 64 (key to the included species); Univ. Iowa Stud. Nat. Hist., vol. 9, No. 5, 1921, p. 14 (includes both recent and fossil species; not known from West Indies) ; Smithsonian Mise. Coll., vol. 72, No. 7, 1921, p. 17.—Gistin, Nova Acta Reg. Soc. Sci. Upsali- ensis, ser. 4, vol. 5, No. 6, 1922, pp. 68, 73; Zool. Bidrag Uppsala, vol. 9, 1924, pp. 25, 30 (has P, [Pc] to the left), p. 35 (brachial homologies), p. 44 (reversion), p. 49, p. 84 (distal intersyzygial interval), pp. 106, 107, 115, 117; Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, p. 17. 146 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Endiocrinus Baruer, Rept. Brit. Assoc. for 1898, 1899, p. 923.—-Cuun, Aus den Tiefen des Welt- meeres, 1900, p. 488.—Hvurron, Index faunae Novae Zelandiae, 1904, p. 291. Eudiocrinis A. H. Cuarx, Smithsonian Misc. Coll., vol. 60, No. 10, 1912, p. 12. Eudiocriuus, GisLGN, Nova Acta Reg. Soc. Sci. Upsaliensis, ser. 4, vol. 5, No. 6, 1922, p. 182. Diagnosis.—A genus of comatulids with five arms only; [Br series are present in which the two elements are united by syzygy and the distal bears a pinnule instead of being axillary. This is the only genus in the family Eudiocrinidae; for further information re- garding its characters and relationships see page 143. Geographical range—From the Maldive Islands, Cape Negrais, Burma, and the Andaman Islands to the Lesser Sunda and Kei Islands, the Moluccas, the Philippine and Bonin Islands, and southern Japan. Bathymetrical range.—F rom 36 to 183 meters. Remarks.—While the genus Hudiocrinus is one of the most clearly delimited and most definitely characterized of all comatulid genera the interrelationships of the in- cluded species are as yet by no means entirely clear. The curious elongated cirri of E. junceus seem to be quite distinctive, differentiat- ing that species sharply from all the others. But as it has only been taken once we know nothing of its geographical or other variations. Similarly the short and strongly recurved cirri composed of short segments charac- teristic of LZ. variegatus, E.. loveni, and FE. pinnatus seem to indicate that these species form a natural group. But these species are very insufficiently known, while the cirri in certain individuals of other species sometimes approach the short-segmented type rather closely. The small and slender Z. venustulus appears to be a well-marked form, although it is sometimes rather easily confused with the young of the larger and stouter species. Though on the basis of our present information FE. gracilis, E. pulchellus, and E. eoa appear to be quite distinctive, as yet we know very little about them. Whether £. ornatus, E. philenor, and E. serripinna are to be regarded as species or as forms of E. indivisus is more or less a matter of personal opinion. The inter- relationships of serripinna, ornatus, and indivisus are extraordinarily similar to the interrelationships between Oligometra serripinna and its corresponding varieties. History.—In 1868 Prof. Carl Semper established the genus Ophiocrinus for the reception of a curious little comatulid that he called Ophiocrinus indivisus from the Philippine Islands. In the following year a second species was described by Prof. Percival de Loriol under the name of Comatula (Ophiocrinus) hyselyi from the Neo- comian of Switzerland. The generic value of the character separating Ophiocrinus from Antedon or Comatula, as those genera were understood at that time, was doubted by Schliiter in 1878, and P. H. Carpenter had some hesitation in regarding it as equivalent to Antedon, Actinometra, and Promachocrinus. The Challenger had dredged three species of 5-armed comatulids in the western Pacific, and in 1882 Carpenter described these, at the same time redescribing Semper’s Ophiocrinus indivisus, the type and only specimen of which had been acquired by his father. Carpenter pointed out that the generic name Ophiocrinus had been used by Salter in 1853 and was not therefore available for Semper’s genus, for which he pro- A MONOGRAPH OF THE EXISTING CRINOIDS 147 posed the name Eudiocrinus. Carpenter noted that “in its central mouth and in the structure of the calyx Eudiocrinus is essentially an Antedon. But the sacculi which are usually so abundant at the sides of the ambulacra of this genus are not so constant in Eudiocrinus. E. indivisus has numbers of them, while they are scanty in £. varians, and altogether absent in the two remaining species [semperi and japonicus], which I have never found to be the case in Antedon, though I have examined over one hundred species of this genus.’’ He noted that in E. indivisus the sacculi are ‘‘tolerably close on the arms, but larger and more closely set at the sides of the pinnule-ambulacra, which have only the very slightest trace of any superficial limestone deposits.” In 1884 in his report upon the stalked crinoids collected by the Challenger, Car- penter discussed Eudiocrinus at considerable length. He remarked that in EL. in- divisus the two ossicles following the radials are united by syzygy, and the epizygal of this syzygial pair bears a pinnule “which clearly shows that they must be considered as arm-joints and not as belonging to the calyx, although they undoubtedly represent the so-called second and third radials [the ossicles of the IBr series] of a ten-armed Cri- noid.” He said further that in the other species of Hudiocrinus — that is, in the species now assigned to Pentametrocrinus — these primitively separate ossicles are not united by syzygy but are articulated, just as in Thawmatocrinus [= the young of Decametrocrinus]. The second one bears a pinnule both in Thawmatocrinus and in Eudiocrinus [Pentametrocrinus] varians; but in Eudiocrinus [Pentametrocrinus] sempert and in E. [P.] japonicus the first pinnule is on the fourth ossicle after the radial. He said that this would correspond to the second brachial of a 10-armed crinoid, but it is really the fourth brachial in Eudiocrinus. Lastly, he remarked, in Perrier’s Hudio- crinus atlanticus the first pinnule is on the fifth brachial, which corresponds to the third brachial of an Antedon. In the Challenger report on the comatulids published in 1888 Carpenter gave no new information regarding E. indivisus. His account of the genus Eudiocrinus is based almost wholly upon the three species of Pentametrocrinus secured by the Chal- lenger and the one (atlanticus) described by Perrier. His description of the disk and of the articular faces of the radials is entirely based upon the Challenger species of Pentametrocrinus. In 1894 Prof. F. Jeffrey Bell described and figured a second species of Eudiocrinus from the Macclesfield Bank, which was closely related to E. indivisus. In fact Bell’s new species, E. granulatus, subsequently proved to be a synonym of E. indivisus. In 1908, having obtained from southern Japan a single specimen of a new species, which I described as Eudiocrinus variegatus, I showed that whereas in E. indivisus and its close relatives the two ossicles united by syzygy immediately following the radials are the equivalent of the IBr series in comatulids with 10 or more arms (as was suggested by Carpenter in 1884) there are no representatives of the IBr series at all in Eudiocrinus varians, E. japonicus, E. semperi, E. atlanticus, or E. tuberculatus. These last I therefore removed from the genus Hudiocrinus, placing them in the new genus Pentametrocrinus. At the same time I showed that Eudiocrinus, as restricted by the removal of various species to Pentametrocrinus, is most closely related to Zygometra, and I united Eudiocrinus and Zygometra in the new family Zygometridae. In another paper published in 1908 I described Hudvocrinus serripinna, which had been dredged by the Albatross at station 5136 in the Philippines, and in 1909 I 148 BULLETIN 82, UNITED STATES NATIONAL MUSEUM described E. ornatus, which had been dredged by the Royal Indian Marine Surveying steamer Investigator at station 61 north of the Andamans. In 1911 I recorded £. indivisus from Albatross station 5356 in the Philippines. In 1912, Eudiocrinus junceus, E. pinnatus, and E. venustulus were described from the collections of the Dutch steamer Siboga in the East Indies, and in a monograph on the crinoids of the Indian Ocean E. ornatus was redescribed and figured, a new species, E. minor, was also described and figured, and another new species, . gracilis, was described. In the same year I described also a second specimen of the Japanese E. variegatus, which I had studied at the Hamburg Museum, and in 1913 I published notes upon the type specimen of Bell’s 2. granulatus, which I had examined in the British Museum in 1910. In a report upon the crinoids collected by the Siboga published in 1918, Hudto- crinus junceus, E. punnatus, and E. venustulus were redescribed and figured, and £. ornatus, E. serrupynna, and E. wndwsus were recorded from various Svboga stations. In 1922 Dr. Torsten Gislén recorded EF. indivisus from the Bonin Islands, and from the same region described E. gracias var. pulchellus and E. lovena, the specimens having been secured by Dr. Sixten Bock during his expedition to Japan in 1914. In 1924 Dr. Gislén published extensive notes on the structure of the genus based upon the material from the Bonin Islands. KEY TO THE SPECIES IN THE GENUS EUDIOCRINUS a!. Cirri composed of much elongated segments of which the longest are 3% to 4 times as long as median width or longer, and the distal are twice as long as broad. b!. Cirri greatly elongated with 22 elongated segments, tapering from base to tip, 23 mm. long; enlarged proximal pinnules stout basally, very long, tapering gradually to a delicate, almost flagellate, tip; arms 90 mm. long (Moluccas; 95 meters) -_--------------- junceus (p. 149) 2. Cirri less elongated with 13-15 segments, not tapering distally, 10 mm. long; enlarged proximal pinnules not especially long, tapering evenly to tip; arms 40 mm. long (Mergui Archipelago; ERO bers) ee ee ee ee eee philenor (p. 151) a?. Cirri shorter, with the longest segments never more than 2% times as long as broad. 1. Cirri of moderate length, not especially stout, with the earlier segments markedly longer than broad, about twice as long as median width or longer. c!. First segment of Pc with a very large spatulate or fan-shaped process as high as width of seg- ment, and following segments with high carinate processes, which rapidly diminish in height distally; ossicles of IBr series and earlier brachials with edges swollen and everted and more or less scalloped or beaded; a slender and delicate species (Philippines to Kei and Lesser Sunda Islands; 73-112 meters) -_----_----.-------_-------- venustulus (p. 160) @. Slight and uniform, if any, processes on basal segments of proximal pinnules; not markedly slender and delicate. d'. Cirri very slender, strongly recurved, and composed of 11 or 12 segments of which the longest are about 24 times as long as median width, and antepenultimate is half again as long as broad (Philippines; about 46 meters)_...---.---------.-_---- eoa (p. 162) @. Cirri not especially slender with shorter segments of which distal are not, or at least are not appreciably, longer than broad. e!, Ossicles of IBr series and earlier brachials with everted and more or less produced edges, dorsal profile of earlier portion of arms being strongly serrate. jf’. Proximal pinnules very sharply triangular in section; on P; and P, distal edges of segments are very strongly produced at prismatic angles dorsally so that pinnules have a strongly serrate dorsal profile (Philippines to Kei and Lesser Sunda Islands; 20-100 MGvers) Soon ee ae ae eee eee serripinna (p. 169) A MONOGRAPH OF THE EXISTING CRINOIDS 149 f?. P, and P, rounded triangular, distal edges of segments everted and spinous all around dorsal side, this being slightly, if at all, accentuated at prismatic angles (Andaman to Lesser Sunda and Kei Islands; 73-90 meters) ---------------- ornatus (p. 172) e. Edges of ossicles of IBr series and earlier brachials only slightly, if at all, everted, so that dorsal profile of earlier portion of arms is smooth. f'. Third segment of P, and P, not longer than broad (Bonin and Philippine Islands and the Macclesfield Bank to Moluccas and Kei and Lesser Sunda Islands; 36-183 Meters) = So es indivisus (p. 163) j?. Third segment of P; and P, half again as long as broad. gi. P, and P, scarcely longer or slenderer than other proximal pinnules; arms up to 40 mm. in length and cirri up to 7 mm. in length (Bonin Islands; ?128—183 meters) pulchellus (p. 177) g?. P, and P, markedly longer than succeeding pinnules; arms 55 mm. and cirri 9 mm. long (Maldive Islands to Burma and Kei Islands; 73-89 meters) gracilis (p. 175.) l?. Cirri stout and strongly curved, none of component segments appreciably longer than broad. c!. P, and P:, though elongated and stiffened, are slender like the succeeding pinnules and not enlarged; third and fourth segments are half again as long as broad and outer are 3 or 4 times as long as broad. d'. P, and P2 are one-third or one-fourth again as long as succeeding pinnules and are composed of 16 or 17 segments (Lesser Sunda to Kei and Philippine Islands; 46-85 meters) pinnatus (p. 153.) @. P, and P, are not longer than succeeding pinnules and are composed of 11 segments (Bonin Islands; 7146 meters) ~-----=--==4_~_2-----------____--_-_- loveni (p. 156.) c2. P, and P, are enlarged as well as stiffened; they are composed of 11 or 12 segments of which first 4 are broader than long and outer are about twice as long as broad (southern Japan; LlOjmeters) 19) 222 tes senna tS ee ee eet a variegatus (p. 157 ) EUDIOCRINUS JUNCEUS A. H. Clark Puate 10, Ficures 35, 36 [See also vol. 1, pt. 1, fig. 83 (side view), p. 136; pt. 2, fig. 788 (ambulacral deposits), p. 366.] Eudiocrinus junceus A. H. Cuarx, Proc. Biol. Soc. Washington, vol. 25, 1912, p. 25 (description ; Siboga station 167); Amer. Journ. Sci., vol. 40, 1915, p. 62 (listed); Unstalked crinoids of the Siboga-Exped., 1918, p. 64 (in key; range), p. 65 (references; detailed description; Sta. 167), fig. 3, p. 66, p. 273 (listed), pl. 17, figs. 31, 32; Smithsonian Mise. Coll., vol. 72, No. 7, 1921, pl. 2, fig. 20 (ambulacral deposits); Rec. Indian Mus., vol. 34, pt. 4, 1932, p. 556 (compared with E. philenor). Diagnostic features —The long tapering cirri with elongated segments and the greatly enlarged and slowly tapering P, and P, easily distinguish this species from all the others in the genus. Description.—The centrodorsal is discoidal, moderately thick, the sides converging rather strongly; the flat dorsal pole is 1.0-1.5 mm. in diameter. The cirrus sockets are arranged in two closely crowded and irregular rows. The cirri are XXV, 22, 23 mm. long, greatly elongated with elongate segments, very slender, tapering from the base to the tip, rather more strongly in the first three or four segments than subsequently. The first segment is short, the second is not so long as broad, the third is nearly or quite twice as long as the proximal diameter, the fourth is about four times as long as the median diameter, and the sixth, seventh, and eighth are very slightly longer. From this point onward the length almost imper- ceptibly decreases so that the seventeenth and following segments are slightly over twice as long as broad. The penultimate segment is half again as long as broad, 150 BULLETIN 82, UNITED STATES NATIONAL MUSEUM tapering somewhat distally. The opposing spine is represented by a small rounded subterminal tubercle. The terminal claw is slightly longer than the penultimate segment, very slender and sharp, and only slightly curved. The second and third segments are rather strongly constricted centrally; the sixth and following have mod- erately expanded and slightly overlapping distal ends; both of these characters gradu- ally die away distally. From the fifth segment onward the cirri are rather strongly compressed laterally. The radials are just visible beyond the edge of the centrodorsal; their distal border is swollen and turned outward, smooth or evenly tuberculated. The ossicles of the Br series (united in a syzygial pair) taken together form an element that is oblong, not quite twice as broad as long; both the proximal and the distal borders are turned outward, the former slightly, but the latter standing up at right angles to the general surface of the segment, with a smooth and somewhat thickened edge; the proximal edge may be more or less scalloped, and bears just within the border a prominent rounded tubercle; the produced distal edge is thickest and most prominent in the middorsal half, this portion being distally evenly concave; the remainder of the distal edge may be broadly scalloped. The five arms are 90 mm. long. The first brachial is oblong, two and one-half to three times as broad as long, with the proximal and distal edges slightly thickened and everted. The second brachial is similar, but the distal edge is prominently everted, especially in the central third where the eversion is thickened and distally concave, standing up vertically from the dorsal surface of the segment. The first syzygial pair (composed of brachials 3+4, or the fifth and sixth postradial ossicles) is slightly longer on one side than on the other, about twice as broad as the lesser length; this syzygial pair resembles the preceding syzygial pair (composed of the two elements of the 1Br series), but the tubercle just within the proximal border is only very slightly indicated. The following three brachials are slightly wedge-shaped, about twice as broad as the median length, with the distal borders everted as described for the second brachial, but progressively Jess and less so. The following brachials are triangular, about as long as broad, with slightly produced and overlapping distal edges. From the fourth to the ninth brachials there is a low median carination which after the ninth becomes the low rounded zigzag keel characteristic of the arms of all the species of this genus, which is traceable throughout the entire length of the arms. Pe is 6.5 mm. long, with 15 segments, rather stout basally, but tapering rapidly to a very delicate tip, and strongly prismatic. The first segment is much broader than long, the following gradually increasing in length and becoming about as long as broad on the fourth or fifth, and terminally twice as long as broad. The second- sixth segments have a narrow sharp carination the crest of which is straight and parallel to the longitudinal axis of the pinnule; the outer edge of the prism formed by the pinnule is sharp; the outer surface of the pinnulars between the prismatic angles is flat or very slightly concave. P, is similar, 6.5 mm. long, with 13 or 14 segments. P, is 11 mm. long, with 15 segments, much larger and stouter than those preceding, tapering evenly from the base and becoming very delicate distally. The first segment is much broader than long, the following gradually increasing in length and becoming about as long as broad on the fourth and three times as long as broad terminally. The second, third, and fourth segments have a low even carination. The second and A MONOGRAPH OF THE EXISTING CRINOIDS 151 following have the distal edges all around produced and finely spinous. Py is 11 mm. long, with 15 segments, exactly resembling P,. P, is 8-10 mm. long, with 19 segments, of which the basal are as small as the basal segments of P,. The first segment is short, more or less crescentic, the second is about twice as broad as the median length, the third is not quite so long as the distal breadth, the fourth is from one third to one half again as long as broad, and the following gradually increase in length, after the eighth being three or four times as long as broad; the fifth and following have slightly produced and spinous distal edges. P,; is 10 or 11 mm. long, similar to P,. P, is 8 mm. long, with 17 segments, slenderer than P, and with more elongated segments. P, is similar to P,. Pa is 7 mm. long, with 18 segments, resembling P, but even more delicate and with longer segments. P, is similar to Py. P, is 6.5 mm. long, with 16 segments, and resembles P,. P, is similar to P,. The distal pinnules are excessively slender, 8.5 mm. long, composed of 20 segments. Each ambulacral lappet contains usually two similar smooth spicules, an outer and an inner, the adjoining ends of which overlap more or less; each is bent at an obtuse angle in the middle; the outer may be branched or forked at the outer end, or may be replaced by two or three small straight spicules irregularly arranged. Each tentacle contains at the base a double (distal and proximal) continuous row of small spicules which soon coalesce, forming a single row which, becoming gradually narrower, is continued almost to the tip; when the tentacle is contracted this row of spicules may be more or less zigzag. These tentacular spicules are much the most conspicuous of all the perisomic deposits in the pinnules. Locality — Siboga station 167; near Misool (lat. 2°35’30’’S., long. 131°26/12’E.); 95 meters; hard coarse sand; August 22, 1899 [A. H. Clark, 1912, 1918] (16, US.N.M., E.426; Amsterdam Mus.). Remarks.—This very distinct species is as yet known only from the original specimens dredged by the Siboga. EUDIOCRINUS PHILENOR A. H. Clark Puate 10, Fieurss 37, 38 Eudiocrinus philenor A. H. Cuarx, Rec. Indian Mus., vol. 34, pt. 4, 1932, p. 551 (listed), p. 554 (Investigator station 534; description), pl. 19, figs. 1-4. Diagnostic features —The cirri are long and slender but do not taper distally; they are composed of 13-15 segments of which the longest are three and one-half to four times as long as broad, and the distal are about twice as long as broad; the en- larged proximal pinnules are not especially long; the arms are about 40 mm. long, and the cirri are 10 mm. long. Description —The centrodorsal is discoidal, thin, with a rather broad, flat, and more or less rugose polar area. In small specimens the dorsal pole is more or less strongly convex. The cirrus sockets are arranged in a crowded zigzag row, alternating higher and lower; rarely there are two sockets one above the other. The cirri are XX-XXV, 13-15 (usually 14), 10 mm, long, slender and delicate and only moderately curved. ‘The first segment is very short, the second is longer, about twice as broad as long or somewhat shorter, the third is half again as long as the median width, the fourth is about three and one-half times as long as the median 152 BULLETIN 82, UNITED STATES NATIONAL MUSEUM width, and the three following are about four times as long as the median width; the segments succeeding slowly decrease in length so that the antepenultimate is about twice as long as broad and the penultimate is about as long as broad or slightly longer than broad. The third segment is rather strongly constricted centrally. The next few segments are moderately constricted centrally, and those succeeding have a slightly concave ventral profile, but soon a less strongly concave dorsal profile. The antepenultimate segment is practically oblong in lateral view. The opposing spine is prominent and sharp, terminally situated, and is equal in height to about half the distal width of the penultimate segment. The terminal claw is about as long as the penultimate segment or slightly longer, and is slender and moderately and evenly curved. The gently concave distal edge of the radials extends slightly beyond the rim of the centrodorsal. The syzygial pair formed by the two elements of the IBr series is oblong and half again as broad as long. The IBr, has a low and broadly rounded elevation occupying the median line. The IBr, has a corresponding but somewhat less strongly marked midradial elevation, and the distal edge is slightly everted. The five arms are about 40 mm. long. The first brachial is short, oblong, about four times as broad as long, of about the same size as the IBr, and with a similar midradial elevation. ‘The distal border is slightly everted. The second brachial is slightly larger than the first, and is trapezoidal in shape; it bears a slight median elevation in the proximal half or two-thirds, and the distal edge is slightly everted. The syzygial pair formed by brachials 3+-4 is slightly longer on the side bearing the pinnule than on the opposite side, and is not quite twice as broad as long. The hypo- zygal is oblong, nearly four times as broad as long, and the epizygal is longer on the side toward the pinnule base than on the opposite side, in the midradial line being of the same length as the hypozygal. The next two brachials are very bluntly wedge- shaped, about two and one-half times as broad as the midradial length. After the second syzygy the brachials become triangular, as long as or somewhat longer than broad, but soon wedge-shaped again with less oblique ends, later becoming longer with slightly concave sides. The brachials as far as the second syzygy have a low and broadly rounded tubercle in the proximal portion. The succeeding brachials have the middle of the dorsal line almost imperceptibly elevated. Syzygies occur between the elements of the IBr series, between brachials 3-+-4, 8+-9, and 13+ 14, and distally at intervals of 3 muscular articulations. Pc is 4 mm. long, with 11 segments, of which the first is short and trapezoidal, half again as broad as the greater (inner) length, the second is almost oblong, twice as broad as the median length, the third is about as long as broad, the fourth is half again as long as the median length, and those following slowly increase in length, becoming nearly three times as long as broad terminally. The third and following segments have the distal edge very slightly everted. The pinnule is moderately stout at the base and tapers evenly to the tip. The middorsal line is very obscurely and roundedly elevated, scarcely suggesting the carination found in related species. P, is4 mm. long, with 12 segments, and resembles Pc. P, is § mm. long, with 15 seg- ments, roughly half again as broad at the base as Pe, moderately slender and tapering evenly to a fine tip. It is somewhat flattened laterally and is broadly rounded dor- sally, not being prismatic. The first segment is from half again to twice as broad as A MONOGRAPH OF THE EXISTING CRINOIDS L653 long, the second is slightly broader than long, the third is slightly longer than broad, the fifth is twice as long as the median width, and the eighth or ninth and following are about three times as long as broad. The longer segments are very slightly constricted centrally, and the distal ends are slightly prominent and are armed with exceedingly fine spines. Locatity.— Investigator station 534; Mergui Archipelago, west of Tavoy Island (lat. 12° 59’ N., long. 96° 48’ 30’ E.); 110 meters; coral and sand; April 17, 1913 (14, U. S. N. M.; I. M.). Remarks.—This species is more or less intermediate between E. gunceus, in which the cirri are greatly elongated, slender, composed of usually 22 greatly elongated segments, and taper gradually from the base to the tip, and the other species of the genus in which the cirri are much shorter and are composed of fewer segments of which the longest are never more than twice as long as broad and the distal are about as long as broad. It seems to be most closely related to HL. ornatus from which it differs in having the cirri more slender with more elongated segments, in the longer P, and P., in having the ossicles of the IBr series and the earlier brachials distinctly, though roundedly, carinate, and in having the distal edge of the earlier brachials much less produced. EUDIOCRINUS PINNATUS A. H. Clark Puate 11, Fiagure 39 (See also vol. 1, pt. 1, fig. 84 (side view), p. 137; pt. 2, fig. 787 (ambulacral deposits), p. 366.] Eudiocrinus pinnatus A. H. Cuark, Proc. Biol. Soc. Washington, vol. 25, 1912, p. 27 (description; Siboga Station 310); Amer. Journ. Sci., vol. 40, 1915, p. 62 (listed); Unstalked crinoids of the Siboga-Exped., 1918, p. 1x (relationship with ZH. variegatus), p. 64 (in key; range), p. 67 (refer- ences; detailed description; station 310), p. 68, fig. 4, p. 276 (listed), pl. 17, fig. 30.—GisLin, Nova Acta Reg. Soe. Sci. Upsaliensis, ser. 4, vol. 5, No. 6, 1922, p. 73 (compared with EZ. loveni).— A. H. Cuarx, Temmincekia, vol. 1, 1936, p. 295 (listed), p. 300 (Snellius station 123*; notes). Diagnostic features —The cirri are short and stout, with none of the component segments longer than broad; P, and P, are elongated and stiffened, but slender like the succeeding pinnules and not enlarged, with the third and fourth segments half again as long as broad; these pinnules are one-third or one-fourth again as long as the pinnules following, and are composed of 16 or 17 segments. The short, stout, short-segmented cirri distinguish this species from all the others in the genus except E. variegatus and E. loveni. From E. variegatus it is distinguished by the slenderness of P, and P,, which have longer proximal segments, and from E. loveni it is distinguished by the fact that these pinnules are markedly longer than those succeeding and are composed of 16 or 17 instead of only 11 segments. Description —The centrodorsal is discoidal, with the flat dorsal pole 2 mm. in diameter. The cirrus sockets are arranged in two marginal rows. The cirri are XVII, 17-22, from 10 to 11 mm. long; all the segments beyond the third are subequal, none being quite so long as broad; the fourth-eighth or -ninth have slightly prominent distal edges, especially dorsally. The cirri are rather stout and are of the same type as those of F’. variegatus. The five arms are about 75 mm. long; the brachials are unornamented, and their distal edges are not produced. 208244—40——11 154 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Po is 4 mm. long, moderately stout, strongly prismatic, evenly tapering to the tip, the distal border of the segments sharply, though narrowly, carinate, composed of 11 segments. P, similar to Po, 4.5 mm. long, with 12 segments. P, is 11 mm. long, with 17 segments, very slender like the succeeding pinnules, though very stiff. The first two segments are slightly broader than long, the third and fourth half again as long as broad, and the following gradually increase in length, becoming distally three or four times as long as broad; distal edges of the third and following segments are produced and spinous. The pinnule resembles the lower pinnules in certain of the more slender species of Colobometra. P, is similar to P,, 11 mm. long, with 17 segments. P, is similar to P;, 8 mm. long, with 18 segments. P, is 8 mm. long, with 18 segments, proportionately slenderer than P, and with much shorter segments, which become as long as broad on the third and twice as long as broad basally on the tenth or eleventh. P, and P, are 7.5 mm. long, with 18 segments, and resemble the preceding pinnules; the distal pinnules are 12 mm. long, with 20-24 segments, which, beyond the fifth, are twice, and distally three times, as long as broad, with very finely spinous distal ends. The deposits along the ambulacra in this species are essentially the same as in E. junceus. Notes —The preceding description was drawn up from the type specimen from Siboga station 310. The arms in the six specimens from the Danish Expedition to the Kei Islands station 54 are up to 55 mm. in length. The specimen from off Jolo in about 46 meters may be thus described: The centrodorsal is thin discoidal with a broad flat dorsal pole; the cirrus sockets are arranged in a single regular marginal row. The cirri are XV, 17, moderately stout and strongly curved. The first two segments are more than twice as broad as long, and those following are all about as long as broad, the fourth being sometimes, but not always, slightly longer than broad. The third and fourth segments are rather strongly constricted centrally, and the two following have somewhat flaring distal ends. The opposing spine is prominent, conical, with the apex median, and rises to a height equal to half the width of the penultimate segment. The terminal claw is slightly longer than the penultimate segment, rather stout, and strongly and usually evenly curved. The radials have a slightly and very finely roughened distal border. The brachial structure resembles that of EZ. indivisus. The distal edges of the bra- chials are very finely spmous but are not produced. Pe is about 6 mm. long and is composed of 15 segments;it tapers evenly from a fairly broad base to the tip. The first two segments are much broader than long, the fourth is about as long as broad, and the remainder are about half again as long as broad. The pinnule is not obviously prismatic, the dorsal ridge being so broadly rounded as scarcely to be evident. Each of the first five segments bears a conspicuous sharp carinate process with the crest straight and parallel to the longitudinal axis of the pinnule. On the first segment this crest occurs only in the distal half, but it occupies the entire side of the other segments. It is highest on the second-fourth segments, lower on the fifth, and very low on the sixth. P, is very slightly longer than P, and also stouter. It is composed of 14 segments, which become about as long as broad on the fourth and half again as long as broad distally. The first four A MONOGRAPH OF THE EXISTING CRINOIDS 155 segments bear conspicuous carinate processes resembling those on the corresponding segments of Pc. P,, situated on the epizygal of the first brachial syzgyial pair, is 12 mm. long, stiffened and more or less spinelike, stouter basally than P,, and tapering very slowly. It is composed of 16 segments of which the first is trapezoidal, some- what broader than long, the second is nearly oblong, not quite twice as broad as long, the third is somewhat longer than broad, and those following increase in length be- coming nearly 3 times as long as broad near the tip. The pinnule is not carinate but is strongly rounded dorsally. The third and following segments have the distal edge bordered with fine spines, which increase in size on succeeding segments. There is no trace of carination on the earlier segments. P, resembles P,. This pair of pinnules stands out in marked contrast to the other pinnules. P, is 7 mm. long, with 17 segments. It is less stout basally than the preceding pinnules, tapers to a delicate tip, and is flexible. Only the five terminal segments are longer than broad. The segments have produced and spinous distal ends. The following pinnules are very small and weak. In the specimen from Willebrord Snellius station 123* the cirri are XVII, 13-14, 5mm. long. The first segment is very short, the second is twice as broad as long, the third is half again as broad as long, and the remainder are about as long as the width of their distal ends. The segments are slightly constricted centrally, this constriction becoming obsolete in the distal fourth of the cirri. The opposing spine is slender, sharp, erect, and subterminal, and is equal in height to about half the width of the penultimate segment. The terminal claw is about as long as the penul- timate segment and is rather stout and strongly curved. The radials have their distal border strongly and irregularly tuberculated. All the segments of P, and P, have high thin-lobed or coarsely serrate carinate processes that decrease in height toward the pinnule tip. P, and P, are 3.5 mm. long, with 11 segments, and are slender. The first segment is crescentic, almost semicircular, twice as broad as long, the second is oblong, twice as broad as long or even broader, the third is half again as long as broad, and those following increase in length, becoming twice as long as broad dis- tally. The second and following segments have the distal edges spinous, especially at the prismatic angles, this spinosity increasing distally. P, resembles P,. P, and the following pinnules are somewhat shorter and weaker, though only very slightly less stout. This specimen appears to represent H. pinnatus, differing from the type specimen only in features attributable to its smaller size. Localities —Siboga station 310; north of the eastern end of Sumbava, Lesser Sunda Islands (lat. 8° 30’ S., long. 119° 07’ 30’ E.); 73 meters; sand with a few pieces o1 dead coral; February 12, 1900 [A. H. Clark, 1912, 1918] (1, Amsterdam Mus.). Danish Expedition to the Kei Islands station 53; Dr. Th. Mortensen; 85 meters; sand and coral; May 9, 1922 (1). Danish Expedition to the Kei Islands station 54; Dr. Th. Mortensen; 85 meters; sand and coral; May 9, 1922 (6). Dr. Th. Mortensen’s Pacific Expedition, 1914-1916; off Jolo (Sulu); about 46 meters; sand and coral; March 19, 1914 (1). Willebrord Snellius station 123*; lat. 10° 29’ 00’’ S., long. 126° 44’ 00” E.; about 100 meters; October 28, 1929 (1, L. M.). Geographical range-—From the Lesser Sunda to the Kei and Philippine Islands. 156 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Bathymetrical range.—From 46 to 85 meters. Remarks.—This species is as yet known only from the 10 specimens listed above. EUDIOCRINUS LOVENI Gislén Eudiocrinus loveni Gisuf#x, Nova Acta Reg. Soc. Sci. Upsaliensis, ser. 4, vol. 5, No. 6, 1922, p. 4 (145 m.), p. 6 (Bonin Islands), p. 7, p. 72 (description; Bock’s station 47; comparisons), p. 182 (listed), figs. 53-56, p. 70, pl. 1, fig. 6; Zool. Bidrag Uppsala, vol. 9, 1924, p. 41 (measurements; details of arm structure), p. 42 (brachial angles), p. 285 (Bonin Islands, 144 m.; details), fig. 9, p. 45 (arm base). Diagnostic features.—The cirri are short and stout, with none of the component segments longer than broad; P, and P, are elongated and stiffened, but slender like the pinnules succeeding and not enlarged, with the third and fourth segments half again as long as broad and the distal segments three or four times as long as broad; these pinnules are of the same length as the succeeding pinnules and are composed of 11 segments. The short, stout, short-segmented cirri distinguish this species from all the others in the genus except FE. variegatus and E. pinnatus. From E£. variegatus it is distin- guished by the slenderness of P, and P:, which are composed of longer segments, and from £. pinnatus it is distingushed by the fact that these pinnules are of the same length as those succeeding and are composed of only 11 segments. Description.—As described by Gislén the centrodorsal is 1.8 mm. in diameter. The cirri are XX, 20-21, from 9 to 10 mm. long. The first two segments are broader than long, the third is about as long as broad, and the fifth-seventh are one- quarter again as long as broad. The segments are smooth, slightly constricted cen- trally, and rather stout. The height of the opposing spine is two-thirds the width of the penultimate segment. The terminal claw is curved, and is as long as the penulti- mate segment. The radials are visible as narrow smooth bands about 5 times as broad as long. The syzygial pair formed by the elements of the IBr series is as long as broad with rather pronounced ventrolateral ledges. The five arms are 40-45 mm. long and are smooth. The first brachial is 1.4 mm. in width. Pe is 3.5 mm. long and is composed of nine segments. The first segment tapers strongly distally, and the second segment is the broadest. The first three or five segments have a distally directed crest of which the height is equal to one-third the width of the segments. P, is 4 mm. long, with 10 segments and is similar to Pg. P, is 5 mm. long, with 11 segments, of which the third is half again as long as broad and the distal are two or two and one-half times as long as broad with a somewhat spiny distal edge. P, is 5 mm. long, with 11 segments. P, is 5 mm. long with 12 segments, P; is 5 mm. long with 11 segments of which the distal have a small spiny wreath. The distal pinnules are 4.5 mm. long, with 16 or 17 segments which are three times as long as broad and of which the last three bear dorsal hooks. Each seg- ment bears four pairs of large sacculi. As preserved in alcohol the arms are dorsally brown with white transverse spots, and the pinnulars are whitish with the sacculi darker. The cirri are white. Notes.—Dr. Gislén said that at first he had some doubt whether this specimen represented a distinct species or was merely a fully grown individual of the type he A MONOGRAPH OF THE EXISTING CRINOIDS 157 described as E. gracilis pulchellus. But he pointed out that some of the differences between the two are of such a nature as to oblige him to believe that they cannot be due simply to age. He listed the most important differences as follows: The cirrus segment are short, not very much longer than broad, and are not at all or only slightly constricted centrally, as in Z. pinnatus and E. variegatus. The proximal segments of Pc and P, bear prominences that are lacking on the corresponding pinnules in E. gracilis pulchellus. In the possession of these prominences it approaches E. venustulus. Eudiocrinus loveni further differs from E. gracilis pulchellus in having the arms, though of about the same length, much stouter, in the more spiny borders of the distal segments of P, and P3;, and in the relatively short distal pinnules. Locality.—Dr. Sixten Bock’s Expedition to Japan station 47; Bonin Islands, east of the Channel; 146 meters; August 1, 1914 [Gislén, 1922]. Remarks.—This species is as yet known only from the single specimen described by Gislén. EUDIOCRINUS VARIEGATUS A. H. Clark PuaTE 11, Figure 41 [See also vol. 1, pt. 2, fig. 127 (arm base), p. 79; fig. 266 (arm), p. 207.] Eudiocrinus variegatus A. H. CuarKk, Proc. Biol. Soc. Washington, vol. 21, 1908, p. 134 (nomen nudum; belongs to same type as E. indivisus and EF. granulatus; arm structure); Proc. U. S. Nat. Mus., vol. 34, 1908, p. 277 (listed), fig. 8, p. 277 (centrodorsal and base of an arm), pp. 278, 279, figs. 9-11 (description; 34°59’ N., 139°34’ E., 60 fathoms), p. 307 (recorded); Proc. Biol. Soc. Washington, vol. 25, 1912, p. 27 (compared with E. pinnatus); Crinoids of the Indian Ocean, 1912, p. 102 (synonymy; habitat); Journ. Washington Acad. Sci., vol. 5, 1915, p. 214 (southern Japanese species; range and its significance); Amer. Journ. Sci., vol. 40, 1915, p. 62 (listed) ; Unstalked crinoids of the Siboga-Exped., 1918, p. 1x (relationship with E. pinnatus), p. 64 (in key; range).—Guistin, Nova Acta Reg. Soc. Sci. Upsaliensis, ser. 4, vol. 5, No. 6, 1922, pp. 68, 73 (compared with E. lovenz). Eudiocrinis variegatus A. H. Cuark, Smithsonian Misc. Coll., vol. 60, No. 10, 1912, p. 12 (no locality; description). Diagnostic features —The cirri are short and stout, with none of the component segments longer than broad; P, and P, are enlarged and are composed of 11 or 12 segments of which the first four are broader than long, and the outer are about twice as long as broad. This species is distinguished from FE. pinnatus and E. loveni, which have similar short, stout, short-segmented cirri, by the stout and short-segmented P, and P). Description The centrodorsal is discoidal, rather thick, with somewhat con- vergent sides and a broad, flat polar area. The cirrus sockets are arranged in two crowded and more or less irregular rows. The cirri are about XXV, 15, from 7 to 9 mm. long. The first segment is about twice as broad as long, and the remainder are about as long as their distal width. The fourth and fifth segments have their distal ends considerably swollen and projecting somewhat dorsally, this feature becoming gradually less and less marked distally, where the segments are somewhat compressed laterally. There are no dorsal spines. The opposing spine is prominent, erect, and median, arising from the entire dorsal surface of the penultimate segment. The terminal claw is rather stout and well curved and is about as long as the penultimate segment. 158 BULLETIN 82, UNITED STATES NATIONAL MUSEUM The distal edges of the radials are even with the rim of the centrodorsal. The IBr, are short, oblong, about four times as broad as long, with a well-marked median tubercle. The IBr, are slightly longer than the IBr,, though similar. These two ossicles together form a syzygial pair, which is nearly twice as broad as long, with the lateral edges slightly turned outward. The five arms are 75mm. long. The first and second brachials are slightly over twice as broad as long with the distal edges rather prominent and with a rounded median tubercle; their lateral edges, like those of the IBr, and IBr,, are slightly turned outward, forming a narrow lateral border. The first syzygial pair (composed of brachials 3-4-4) is not quite so long as broad, slightly longer on one side than on the other, and bears a low rounded median keel. The next two brachials are wedge- shaped, not so long as broad, and those following are triangular, about as long as broad, soon becoming obliquely wedge-shaped and longer than broad, and distally elongate and less obliquely wedge-shaped. The brachials as far as the sixth or seventh have alow rounded median keel marked by aseries of longitudinally elongate tubercles, one on each brachial. From this point onward the keel becomes less evident and assumes a zigzag course. It is traceable to at least the distal third of the arm. The distal ends of the brachials in the proximal portion of the arm project somewhat, but this feature dies away after about the middle of the arm. Syzygies occur between brachials 3+-4, 8+-9, and 13+14, and distally at inter- vals of 3 muscular articulations. P. is 4 mm. long, stout and prismatic, composed of nine segments, which are about as long as broad. P, is similar to P. but much stouter, 6 mm. long with 12 segments, of which the first three are about as long as broad and the remainder are longer than broad, though not elongated; the pinnule is rounded and all the segments overlap strongly and have serrate distal ends. P, resembles P,. The pinnules following are about 4.5 mm. long, much more slender than those preceding, with about 12 segments of which the first is very short, the next two are about as long as broad, and the re- mainder are elongate with the articulations somewhat swollen and the lower segments with overlapping and spinous distal ends. The distal pinnules are 9 mm. long, slender, with the first segment short, the second about as long as broad, and the remainder much elongated with expanded articulations. The disk is lacking. Sacculi are abundant and large on the pinnules, less abun- dant end smaller along the brachial ambulacra. The color is purplish brown, lighter dorsally, the arms and pinnules banded with purple and whitish, the cirri white, and the enlarged pinnules (P; and P,) purple. Notes.—A rather small and immature specimen in the Hamburg Museum may be described as follows: The centrodorsal is thin discoidal with a flat dorsal pole 1 mm. in diameter. The cirri are arranged in a single regular marginal row. The cirri are XVII, 14-15, 5 mm. long. The fourth and fifth segments are the longest, very slightly longer than the proximal width. The second-fifth segments are centrally constricted with widely flaring distal ends which project strongly dorsally; this fea- ture rapidly decreases distally and disappears after the eighth. The distal border of the radials is slightly swollen, and is smooth or finely beaded. The proximal border of the IBr, is slightly everted and bears a small but prominent median tubercle; the distal edge is also everted, the eversion being thickened and slightly produced in the A MONOGRAPH OF THE EXISTING CRINOIDS 159 center. The five arms are 40 mm. long. The proximal oblong brachials have the dis- tal edges very slightly turned outward and slightly thickened in the center. The arms beyond the proximal discoidal brachials have a smooth dorsal line and appar- ently lack the usual zigzag carination. The pinnules of the first pair are 3 mm. long and consist of nine segments, which at first are short, becoming about as long as broad on the third and nearly three times as long as broad distally. The first segment has a high rounded carinate process, which is nearly half as high as the width of the segment. The second has a carinate process, which is not quite so high and which has a straight distal border parallel to the longitudinal axis. The third has a carinate process similar to that of the second, but smaller. The following segments are without carinate processes and therefore appear more slender. The pinnules of the second pair are 4 mm. long and consist of 11 or 12 segments. They are stiffened but are only slightly enlarged. The first segment is twice as broad as the median length, and has the distal angle very slightly produced. The second is slightly broader than the median length, but not quite so broad as the first. The third is half again as long as broad, only about half as broad basally as the first. The fourth is similar to the third but longer, twice as long as broad. The following segments slowly increase in length to about three times as long as the proximal width, the terminal one or two being small as usual. The distal edges of the third and following are produced and spinous, the spines being especially long on the prismatic angles. The pinnules as a whole are rounded pris- matic and taper regularly from the third segment to the tip. The pinnules of the third pair are 3 mm. long and are composed of 11 segments. They resemble those of the second pair, but are more slender with more elongated distal segments. The pinnules of the fourth pair are very slender, weak, and delicate, not stiffened, about 3 mm. long, with 10 or 11 segments, which at first are short, becoming half again as long as broad on the third, twice as long as broad on the fourth, and four times as long as broad distally. The third and following have produced and spinous distal ends. The distal pinnules are exceedingly slender, 7 mm. long with 15 or 16 segments of which the third and following are greatly elongated with produced and spinous distal edges. The disk is completely plated. Localities —Sagami Bay, Japan (lat. 34°59’ N., long. 139934’ E.); 110 meters [A. H. Clark, 1908, 1912, 1915, 1918; Gislén, 1922] (1, U.S. N. M., 25326 [Owston Coll. No. 6931]). No locality; C. Eberstein [A. H. Clark, 1912] (1, H. M.). History—This species was first mentioned, as a nomen nudum, in my paper on new genera of unstalked crinoids, published on April 11, 1908. The occurrence of a single specimen in the Owston collection of Japanese crinoids, which had been pur- chased and presented to the U. S. National Museum by Frank Springer, had enabled me to understand the proximal arm structure and to show that in this species, and also in the related E. indivisus, which I had not then seen, the first two ossicles fol- lowing the radials are homologous with the elements of the IBr series in 10-armed forms. Having at hand specimens of the species described by Dr. P. H. Carpenter as Eudiocrinus varians and E. japonicus, as well as of my own recently described EF. tuber- culatus, I was able to demonstrate that in these species the IBr series are wholly 160 BULLETIN 82, UNITED STATES NATIONAL MUSEUM lacking, the first two ossicles following the radials being the equivalent of the first two ossicles following the [Br axillary in 10-armed forms. The species in which IBr series are present although the IBr2 is not axillary (variegatus and indivisus) I retained in the genus Hudiocrinus, while those species that lack the [Br series altogether I removed to the new genus Pentametrocrinus, assigning to this new genus the species previously known as Eudiocrinus atlanticus, E. japonicus, E. semperi, E. tuberculatus, and £. varians. The species was first described, and figures showing the lower pinnules and a cirrus were given, in a paper written before that just mentioned, although it was not published until June 20, 1908. In this paper a very detailed comparison of the struc- ture of the species of Hudiocrinus and of those of Pentametrocrinus is given. In 1912 I published a description of a rather small and immature specimen without locality that I had examined in the Hamburg Museum. EUDIOCRINUS VENUSTULUS A. H. Clark Puate 12, Fiaures 45, 46 (See also vol. 1, pt. 2, fig. 191 (side view), p. 112.] Eudiocrinus venustulus A. H. Cuark, Proc. Biol. Soc. Washington, vol. 25, 1912, p. 27 (description; Siboga station 289); Amer. Journ. Sci., vol. 40, 1915, p. 62 (listed); Unstalked crinoids of the Stboga-Exped., 1918, p. 65 (in key; range), p. 68 (references; detailed description; Stas. 289, 294; Albatross station 5355), p. 275 (listed), pl. 17, figs. 27, 28.—Guistin, Nova Acta Reg. Soc. Sci. Upsaliensis, ser. 4, vol. 5, No. 6, 1922, pp. 7, 73 (compared with ZH. loveni).—A. H. Cuarx, Temminckia, vol. 1, 1936, p. 295 (listed), p. 302 (Snellius station 60*; notes). Diagnostic features.—This small and slender species is readily distinguished from all the other species in the genus by the large spatulate or fan-shaped process on the first segment of Pe. Description.—The centrodorsal is thin-discoidal, with the bare dorsal pole flat, finely papillose, 1.0 mm. in diameter. The cirrus sockets are arranged in a single marginal row. The cirri are XII, 15-16, 6.5 mm. long, rather slender. The first segment is short, the second is longer, the third is about as long as the median diameter, and the fifth and sixth are the longest, about as long as the distal diameter or slightly longer. The segments after the eighth are subequal, slightly longer than broad. The third- seventh segments are constricted centrally with strongly expanded distal ends, which overlap the bases of the succeeding segments, especially dorsally; beyond the seventh this character gradually dies away. The distal border of the radials is just visible beyond the edge of the centrodorsal, and is ornamented with a row of small regular tubercles. The ossicles of the IBr series (united in a syzygial pair) taken together form an element which is oblong, not quite twice as broad as long, with the proximal, distal, and lateral edges everted; the lateral edges are beaded like the distal edge of the radials; the proximal edge is faintly scalloped and bears a prominent median tubercle; the distal edge has the median third of the eversion thickened and standing up vertically as a high transverse ridge; the syzygial line is finely beaded. The five arms are 60 mm. long. The first brachial is oblong, about three times as broad as long, with the proximal edge slightly everted and bearing a prominent, though A MONOGRAPH OF THE EXISTING CRINOIDS 161 small, median tubercle, and the distal edge is strongly everted and thickened, this thickened and everted border being more or less divided in the middle. The second brachial is very slightly larger than the first, about twice as broad as long, with the distal edge everted, the central third of this eversion being thickened and produced. The first syzygial pair (composed of brachials 3-4) is about as long as broad or slightly longer than broad, with the proximal edge slightly everted and bearing a & minute median tubercle, and the distal edge slightly everted with a somewhat larger, more or less transversely elongate, median tubercle. The following brachials have finely spinous distal ends, which are not produced or everted; a slight median tubercle is visible on the proximal border of the brachials as far as the first or second beyond the second syzygy; there is a very low and faint median carination on the syzygial pair and on the following brachials, which is accentuated by being light in color bordered with dark on each side; on the triangular brachials this becomes zig- zag as in the other species of the genus. Pc is 3 mm. long and is composed of 10 segments, of which the first bears a very large fan-shaped, rounded, or distally truncated carinate process about as high as the lateral height of the segment, and the second bears a high carinate process half as high as the lateral height of the segment, of which the crest is parallel to the longitudi- nal axis of the pinnule. The following segments are similarly, but diminishingly, carinate. P, is similar. P, is 5 mm. long, with 11 or 12 segments, of which the first is short, the second is nearly as long as broad, the third is about as long as broad, and the distal are twice as long as broad; the pinnule is rather slender, not greatly enlarged, and rather strongly prismatic; the distal edges of the third and following segments are slightly produced and finely spinous, with prominent spines at the angles of the prism; the ventral borders bear very numerous fine spines; the first segment has a strongly rounded carinate process, and the second and third are narrowly, though sharply carinate. P, is similar to P,. P, is 4 mm. long, with 13 segments, slightly slenderer than P,; the first segment is short, the second is slightly longer, and the third is about as long as broad; the distal segments are much elongated with a few long spines on the distal edges, which are turned outward; the proximal segments are not carinate. The following pinnules are similar, though weaker and more slender with the distal seg- ments slightly longer and bearing a few conspicuous spines on their overlapping distal ends. The distal pinnules are exceedingly slender, 7 mm. long, with 17 segments, of which the third and following are greatly elongated; the third bears a narrow carination. The perisomic deposits along the ambulacra in this species do not differ essentially from those in EL. junceus. Notes.—The three specimens from the Danish Expedition to the Kei Islands sta- tion 54 have the arms up to 45 mm. in length. In the specimen from Willebrord Snellius station 60* the production of the ends of the ossicles of the division series and proximal brachials is exaggerated, and on the radials and Br, is resolved into numerous closely set prominent tubercles which are expecially long in the interradial angles. The rim of the centrodorsal is also tubercular. Localities —Albatross station 5355; north Balabac Strait; Balabac Light bearing S. 61° W., 16.6 miles distant (lat. 8°08’10’’ N., long. 117°19’15’’ E.); 80 meters; coral and sand; January 5, 1909 [A. H. Clark, 1918] (2, U.S. N. M., 36009). 162 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Danish Expedition to the Kei Islands, station 54; Dr. Th. Mortensen; 85 meters; sand and coral; May 9, 1922 (3). Danish Expedition to the Kei Islands, station 57; Dr. Th. Mortensen; 85 meters; sand and coral; May 9, 1922 (2). Siboga station 289; off the southern coast of Timor (lat. 9°00/18’’ S., long. 126°24’30’’ E.); 112 meters; mud, sand, and shells; January 20, 1900 [A. H. Clark, 1912; 1918] (3, Amsterdam Mus.). Siboga station 294; off southwestern Timor (lat. 10°12’12’’ S., long. 124°27/18’" E.); 73 meters; soft mud with very fine sand; January 23, 1900 [A. H. Clark, 1918] (55, U.S. N. M., E. 475; Amsterdam Mus.). Willebrord Snellius station 60*; lat. 6°58’00’’ N., long. 121°52’30’’ E.; 72-80 meters; September 5, 1929 (1, L. M.). Geographical range-—From the Philippines to the Kei and Lesser Sunda Islands. Bathymetricat range—From 73 to 112 meters. History.—This species was first described by me in 1912 from three specimens from Siboga station 289. It was redescribed and figured in 1918, when it was recorded also from Siboga station 294 and Albatross station 5355. In 1936 it was recorded from Willebrord Snellius station 60*. EUDIOCRINUS EOA, sp. nov. Diagnostic features —The slender and strongly recurved cirri composed of only 11 or 12 segments, of which the longest are about two and one-half times as long as the median width and the antepenultimate is half again as long as broad, distinguish this species from all the others in the genus. Description.—The centrodorsal is thin discoidal with a broad flat dorsal pole. The cirrus sockets are arranged in a single fairly regular marginal row. The cirri are XV, 11-12, very slender, strongly recurved, and composed of seg- ments all of which except the basal are longer than broad. ‘The first segment is about twice as broad as long, the second is about twice as long as the median width, the third-sixth or -seventh are about two and one-half times as long as the median width, and those following slowly decrease in length so that the antepenultimate is half again as long as broad and the penultimate is slightly longer than broad. The second segment is rather strongly constricted centrally, the third much less so, and on those following the constriction is still less, finally disappearing at about the middle of the cirri. The opposing spine is represented by a minute terminal tubercle. The terminal claw is somewhat longer than the penultimate segment, and is rather strongly and evenly curved. The IBr series are rather deeply constricted at the syzygial line. In the earlier portion of the arms the central part of the ends of the brachials is slightly thickened, but this disappears after the basal third. P, is short, stout, strongly prismatic, and tapers rapidly and evenly to the tip; it is 2.5 mm. long and is composed of eight segments. P, resembles P,. P, is 4mm. long, about as stout basally as P., but tapers more gradually to a slender tip. Locality—Dr. Th. Mortensen’s Pacific Expedition, 1914-1916; off Jolo (Sulu); about 46 meters; sand and coral; March 19, 1914 (1, C. M.). Remarks.—As yet this species is known only from the type specimen. A MONOGRAPH OF THE EXISTING CRINOIDS.- 163 EUDIOCRINUS INDIVISUS (Semper) Puatse 11, Figure 42 Ophiocrinus indivisus Semper, Archiv fiir Naturg., 1868, vol. 1, p. 68 (description; Pandanon, 30 fathoms).—A. H. Cuark, Proc. U.S. Nat. Mus., vol. 32, 1907, p. 573 (type of Hudiocrinus), p. 574 (first reference and type locality) ; vol. 34, 1908, p. 271 (history; original description quoted) ; Crinoids of the Indian Ocean, 1912, p. 3 (history). Eudiocrinus indivisus P. H. CARPENTER, Journ. Linn. Soc. (Zool.), vol. 16, 1882, p. 495.—Prrnrizr, Compt. Rend. Acad. Sci., vol. 96, 1883, p. 726; Ann. Mag. Nat. Hist., ser. 5, vol. 11, 1883, p. 294 (comparison with EH. [Pentametrocrinus] atlanticus).—P. H. Carpenter, Challenger Re- ports, Zoology, vol. 11, pt. 32, 1884, pp. 47, 48, 58, 127, 187 (discussion).—Prrrier, Mém- oire sur l’organisation et le développement de la comatule de la Meditérranée, 1886, p. 106 (sacculi).—P. H. Carprnter, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, pp. 76, 79-81, 83, 367, 374 (discussion).—Brtu, Proc. Zool. Soc. London, 1894, p. 392 (type in the collection of W. Percy Sladen), p. 397 (comparison with E. granulatus)—Wattuer, Hinleitung in die Geologie als historisch. Wiss., 1894, p. 299 (from Carpenter).—Prirrer, Abh. Senck. Naturf. Ges., vol. 25, 1900, p. 85 (Ternate)—A. H. Cuarx, Proc. U. 8. Nat. Mus., vol. 32, 1907, p. 574 (original reference and type locality)—Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1576 (listed) —A. H. Cuarx, Proc. Biol. Soc. Washington, vol. 21, 1908, p. 134 (belongs to same type as H. granulatus and E. variegatus; arm structure) ; Proc. U. 8. Nat. Mus., vol. 34, 1908, p. 271 (history; original description; structure; relation- ships), p. 277 (listed); vol. 35, 1908, pp. 118, 114, 117 (arm structure); Smithsonian Misc. Coll., vol. 52, pt. 2, 1908, p. 211 (compared with EF. serripinna); Proc. U. S. Nat. Mus., vol. 39, 1911, p. 540 (Albatross station 5856); Mem. Australian Mus., vol. 4, 1911, p. 715 (belongs to a family peculiarly charactristic of the Jurassic rocks of Europe); Crinoids of the Indian Ocean, 1912, p. 3 (history), p. 32 (identity), p. 38 (HE. granulatus a synonym), p. 102 (synonymy; localities); Smithsonian Misc. Coll., vol. 61, No. 15, 1913, p. 21 (published reference to speci- men in the B.M.; Macclesfield Bank, 34-40 fathoms; characters); Amer. Journ. Sci., vol. 40, 1915, p. 62 (listed; Unstalked crinoids of the Siboga-Exped., 1918, p. 65 (in key; range), p. 71 (synonymy; notes; stations 65a, 144, 257, 260, 305); Albatross station 5356); pp. 272, 273, 275, 276 (listed).—Gistin, Nova Acta Reg. Soc. Sci. Upsaliensis, ser. 4, vol. 5, No. 6, 1922, p. 4 (182 m.), p. 6 (Bonin Islands), p. 68 (Bock’s station 59; detailed notes), p. 183 (listed), fig. 52, p. 70; Zool. Bidrag Uppsala, vol. 9, 1924, p. 46 (obliquity of joint faces), p. 77 (syzygies), fig. 68, p. 75 (sygygial face).—A. H. Ciarx, Temminckia, vol. 1, 1936, p. 301 (Snellius Station 60*; notes).—GisL&£N, Kungl. Fysiogr. Sallsk. Lund Férh., vol. 7, No. 1, 1936, p. 3 (Mac- clesfield Bank), p. 6 (range). Eudiocrinus granulatus Bru, Proc. Zool. Soc. London, 1894, pp. 396, 397 (detailed description and comparisons; Macclesfield Bank, 34-40 fms.), pl. 23, opposite p. 392.—A. H. Ciarx, Proc. U.S. Nat. Mus., vol. 32, 1907, p. 573 (original reference and type locality) Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1576 (listed) —A. H. CLarx, Proc. Biol. Soc. Washington, vol. 21, 1908, p. 134 (belongs to same type as EL. indivisus and E, variegatus; arm structure); Proc. U. S. Nat. Mus., vol. 34, 1908, p. 271 (history), p. 272 (rela- tionships), p. 273, footnote (discrepancies between the original figure and description noted), p. 277 (listed); vol. 35, 1908, p. 117 (arm structure); Smiths. Misc. Coll., vol. 52, 1908, p. 211 (compared with E. serripinna; correction of original diagnosis); Crinoids of the Indian Ocean, 1912, p. 38 (identity); Unstalked crinoids of the Siboga-Exped., 1918, p. 65, footnote 2 (synonym of indivisus).—GisLin, Nova Acta Reg. Soc. Sci. Upsaliensis, ser. 4, vol. 5, No. 6, 1922, pp. 68, 69 (discussion). Diagnostic features.—The cirri have 18-20 segments of which the longest proxi- mal are about twice as long as broad and the outermost are about as long as broad. The arms are 80-90 mm. long. The edges of the elements of the IBr series and the proximal and distal ends of the brachials are scarcely or not at all modified. P, and P, are enlarged and prismatic with up to 26 segments. The distal ends of the 164 BULLETIN 82, UNITED STATES NATIONAL MUSEUM segments of the enlarged proximal pinnules may be finely spinous but are not other- wise modified. Description.—The centrodorsal is discoidal, small, 2-3 mm. in diameter, with the dorsal pole flat or somewhat convex, up to 2 mm. in diameter. The cirrus sockets are arranged in a single, or more or less completely double, marginal row. The cirri are XVI-XX, 16-20 (averaging 18), 9-15 (usually about 12) mm. long. The first two segments are short, about as long as broad or slightly broader than long, the third is slightly longer, about as long as broad or a trifle longer, and the fourth- or fifth-sixth or -seventh are the longest, half again to twice as long as broad. The segments following decrease in length, the distal being about as long as broad or slightly longer than broad. The longer earlier segments are strongly constricted cen- trally with swollen and prominent ends, this feature diminishing on the distal segments, which are somewhat compressed laterally with the distal ends more promi- nent on the dorsal than on the ventral side. The opposing spine is stout and promi- nent, and its height is equal to about one-third the width of the penultimate segment. The radials are almost entirely concealed by the centrodorsal; their distal border may be plain, or it may be adorned with small tubercles. The syzygial pair repre- senting the IBr series is about three times as long as the radials. The elements of the IBr series bear a weak median prominence. The five arms are 25-95 (averaging 70) mm. in length and are composed of about 120 brachials. The first brachial is short and oblong and bears no pinnule, and the next four or five are nearly oblong, slightly shorter on one side than on the other, the longer side being shorter than the width and bearing a pinnule. The eighth and following brachials have more markedly unequal sides, the longer of which is longer than the width. The brachials in the middle of the arm are nearly square, and the terminal brachials are elongated. Feeble articular tubercles are sometimes evident. There may be ventrolateral rows of small tubercles as far as the fourth brachial. Syzygies occur between brachials 3+4, 8+-9, and 13+14 (rarely between brachials 12+13) and distally at intervals of 3 muscular articulations. P, is 4 mm. long with 10-12 short segments of which the basal are rather broad. P, is similar to P, but usually slightly longer and stouter. P, is 6-7.5 mm. long, with 10-17 (usually 10-12) segments, very much longer than the preceding pinnules, with longer and more massive segments. P, is 6-8 mm. long and resembles P,. P, and P; are much slenderer than the preceding pinnules and have the outer segments from two and one-half to three times as long as broad. The distal pinnules are about 9 mm. long with 15 segments. The pinnules from P, to P, are extraordinarily stout and thick and are sharply triangular in cross section. The third segment of P; and P, is about as long as broad, and the distal segments are three-quarters again as long as broad, without distal spiny edges. The disk is completely plated, entire, and up to about 3 mm. in diameter. Sac- culi are fairly closely set on the arms, but much larger and more closely set along the sides of the pinnule ambulacra. Notes.—In the specimen from Bock’s station 59, as described by Gislén, the cen- trodorsal is 3 mm. in diameter with the bare dorsal pole, which is somewhat convex, 2mm. in diameter. The cirri are XTX, 19-21, from 11 to 15 mm. long, arranged in a single or double row. The first-third segments are short, the fourth-seventh are A MONOGRAPH OF THE EXISTING CRINOIDS 165 from half again to twice as long as the median breadth, and the distal segments are shorter. The antepenultimate is not very much longer than broad. The fourth- seventh and following segments are strongly constricted centrally, the fourteenth- eighteenth becoming less swollen at the ends. The opposing spine equals in height one-third the width of the segment. The terminal claw is somewhat longer than the penultimate segment. The radials are almost entirely concealed; the visible distal border is adorned with small tubercles. The elements of the IBr series have a weak median prominence. The arm bases are smooth, without the longitudinal promi- nences seen in EL. variegatus. There is a prominence between the second and third brachials, formed in the oblique articulation on the right. There is a similar promi- nence on the left between the fourth and fifth brachials, and one on the right between the fifth and sixth brachials; the prominences then become less pronounced. There are small ventrolateral rows of tubercles as far as the fourth brachial. The width of the first brachial is 1.9 mm. Syzygies occur between brachials 3+4 and 8+9. P, is 4mm. long with 10 segments. P, is also 4mm. long with 10 segments. P, is 7.5mm. long with 10 segments. P;is8mm.long with 11 segments. These pinnules are extraordinarily stout and thick and are sharply triangular. The third segment of P; and P, is about as long as broad. The distal segments are three-quarters again as long as broad, without distal spiny borders. P, and P; are much more slender, the latter having 17 segments and reaching 8 mm. in length; the segments are from two and one-half to three times as long as broad. The disk has been thrown off, but it was probably about 3 mm. in diameter. The color, as preserved, is yellow-brown. Dr. Gislén said that this specimen is closest to Hudiocrinus indivisus, although the longest cirrus segments are rather short for this species. But it has cirri con- siderably longer than those described for E. indivisus, and in this respect approaches the form described as E. granulatus by Bell. Professor Bell said that his new species EZ. granulatus is altogether much stouter than £. indivisus with longer and stronger cirri, much stronger pinnules, and a granular covering to the segments. It has an arm length of 120 mm., and the cirri are about 12mm. long. Bell said that the arrangement and number of the cirri are as described for E. indivisus, but in E. granulatus the first two pinnules have more massive seg- ments than the third and fourth and are quite as long, the second, indeed, being longer than the third. The other striking difference is the granulation of the surface of the arms. The ambulacral surface of the pinnules is a purplish brown, the rest of the animal being yellowish white. Bell wrote that from the descriptions of Semper and of P. H. Carpenter, bearing in mind that they had only one specimen and he had only one, he was inclined to regard the Macclesfield Bank specimen as belonging to Semper’s species. But when he put the two specimens side by side it was easy to see that the two could not be united. In Bell’s figure the cirrus shown has 20 (+1) segments of which the longest are about twice as long as broad and the distal are about as long as broad or slightly broader than long. In the figure of an arm base the lowest pinnule shown, which is apparently P,, is about 6 mm. long and is composed of 16 or 17 segments; P, is 10 mm. long, with 26 segments; P, is about 8 mm. long, with 16 segments; and Pyis 7.5 mm. long. Bell figured the seventh pinnule as 8 mm. long with 17 segments, and “‘one of the most distal pinnules” as about 9 mm. long with 14 (-++1) segments of which all but the basal 166 BULLETIN 82, UNITED STATES NATIONAL MUSEUM are greatly elongated. Gislén has pointed out that in Bell’s figure of the arm base while the pinnules are on the left side they are on brachials corresponding to P, P2, and soon. P, is not shown, and it is somewhat uncertain whether the lowest pinnule should be interpreted as P, or P;. In 1908 I pointed out that in Professor Bell’s description of E. granulatus he uses “first”? and “second” pinnule in the sense of the two first pinnules on the same side of the arm, whereas Professor Semper used the same terms strictly, taking the pinnules first on one side of the arm and then on the other in their actual sequence considering the arm as a whole. Thus Professor Bell found a greater difference between the “first”? and “second” pinnules in his /. granulatus than Professor Semper did in his EZ. indivisus, although in reality such a difference is nonexistent. I examined the type specimen of Z. granulatus at the British Museum in 1910. It is a large specimen, the arms measuring between 85 and 90 mm. in length. There are 18-20 cirrus segments. I could see no difference whatever between this specimen and others undoubtedly referable to EZ. indivisus from the Philippines that I had examined. The characters of the type specimen from Bohol were given by Professor Semper as follows: The centrodorsal is small and flat, 2 mm. in diameter, with the cirri ar- ranged in a single marginal row. The cirri are XVI, 18-20, 9 mm. long, and very knobby, especially proximally where the expansion is equal to the length of the segment. The first two segments are short, as long as broad, and the third-sixth are the longest, twice as long as broad. The outermost segments are slightly compressed laterally and smooth. The last bears, in addition to the terminal claw, a strong tooth (that is, opposing spine). The five arms are 80 mm. long. The IBr series appear to arise directly from the centrodorsal; the two elements are united by syzygy and the IBr, bears a pinnule. The following ossicle (first brachial) is without a pinnule. The second brachial bears a pinnule, and the third and fourth are united by syzygy, the fourth bearing a pinnule. The brachials are almost twice as broad as long, and in the middle of the arm are very obliquely wedge-shaped. The two first pinnules are small, but rather stout basally, and the third and fourth are the longest, 8 mm. long. Then follow shorter pinnules, and subsequently pinnules which gradually become longer and at the same time slenderer. The disk was lacking in the only specimen secured. The pinnules are blotched with light and dark yellow brown, and the dorsal surface of the arms is marked with two parallel zigzag lines. Dr. P. H. Carpenter’s redescription of Professor Semper’s type specimen is as follows: The centrodorsal is small, convex, bearing cirri in two marginal rows, the dorsal pole being free from them. The cirri are about XX, about 20, 9 mm. long. The third segment is about as long as broad, or a trifle longer than broad. The fifth segment is the longest, and those following are nearly all longer than broad, and overlap on the dorsal side. The penultimate bears an opposing spine. The radials are partially visible, and are about one-third the length of the syzygial pair formed by the elements of the IBr series. The IBr,. bears a pinnule on the left side. The five arms are 75 mm. long. The first brachial (third ossicle beyond the centrodorsal) is short and oblong and bears no pinnule. The next four or five brachials are nearly oblong, slightly shorter on one side than on the other, the longer side being shorter than the width and bearing a pinnule. The eighth and following brachials have A MONOGRAPH OF THE EXISTING CRINOIDS 167 more markedly unequal sides, the longer of which is longer than the width. There are about 120 brachials of which the middle ones are nearly square and the terminal are elongated. Syzygies occur between brachials 3+-4, 8+9, and 13-+14 (in one case 12+13) and distally at intervals of 3 muscular articulations. Carpenter expressed this as follows: ‘Second syzygy on the fifth brachial, the next on the ninth, and the next on the thirteenth (once on the twelfth) brachial; after this an interval of two joints between successive syzygia.”’ Pg is quite small, consisting of about a dozen short segments of which the basal are rather broad. P, is somewhat longer and stouter. P, and P, are very much longer and stouter, with longer and more massive segments. The next three or four pinnules on each side gradually decrease in size, and the following ones increase again, the terminal ones being very long and slender so as to give the arms a very feathery appearance. The disk has been lost, but it must have been somewhat less than 3 mm. in diameter. The sacculi are tolerably close on the arms, but much larger and more closely set at the sides of the pinnule ambulacra, which have only the very slightest trace of any superficial limestone deposits. Carpenter said that this is a very slender and graceful little species, differing altogether in appearance from those dredged by the Challenger (that is, Pentametro- crinus varians, P. semperi, and P. japonicus). This specimen subsequently came into the possession of W. Percy Sladen. It was later examined by Dr. P. H. Carpenter, who mentioned that the color of the skele- ton was brownish white, and still later, in January 1894, it was examined by Prof. F. Jeftrey Bell, who found that it had become quite white. The five specimens dredged by the Albatross at station 5356 all have the arms about 85 mm. long. In the specimen from Willebrord Snellius station 60* P, and P, are 7 mm. long with 12-14 segments. The specimens from Siboga station 257 are small. The brachials, except the basal, show a strong production of the distal edge. The individuals from Siboga station 260 have the arms 35 and 40 mm. long. The three specimens from the Danish Expedition to the Kei Islands station 26 have the arms up to 50 mm. long. There is a tendency to an eversion of the central portion of the distal ends of the elements of the IBr series and the earlier brachials. The two specimens from Siboga station 144 are small and, like those from station 257, have the brachials, except the basal, with a strong production of the distal edge. Of the three specimens from Siboga station 305 one has the arms 80 mm. long, the cirri XVIII, 17, 11 mm. long, and the pinnules of the second pair 6 mm. long with 11 or 12 segments. Another has the arms 95 mm. long, the cirri XVII, 16-17, 13 mm. long, and the second pair of pinnules 7 mm. long with 12 or 13 segments. The third specimen is similar to the first. The two specimens from Siboga station 65a are small with the arms 25 to 30 mm. long. Localities —Dr. Sixten Bock’s Expedition to Japan station 59; Bonin Islands, eastnortheast of Anojima; 183 meters; August 15, 1914 [Gislén, 1922, 1924]. Macclesfield Bank; 62-73 meters [Bell, 1894; A. H. Clark, 1907, 1908, 1912, 1913] GiB eM): 168 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Pandanon, near Bohol, Philippines; 55 meters [Semper, 1868; P. H. Carpenter, 1882, 1884, 1888; Perrier, 1883, 1886; Bell, 1894; A. H. Clark, 1907, 1908, 1912]. Albatross station 5355; north Balabac Strait; Balabac Light bearing S. 61° W., 16.6 miles distant (lat. 8°08’10’” N., long. 117°19’15’’ E.); 80 meters; coral and sand; January 5, 1909 (1, U.S.N.M., 35979). Albatross station 5356; north Balabac Strait; Balabac Light bearing S. 64° W., 15.5 miles distant (lat. 8°06’40’’ N., long 117°18’45’’ W.); 106 meters; sand and shells; January 5, 1909 [A. H. Clark, 1911, 1912, 1918] (4, U.S.N.M., 35138, 35139, 35140, 35143). Albatross station 5141; in the vicinity of Jolo (Sulu), Philippines; Jolo Light bearing S. 17° E., 5.5 miles distant (lat. 6°09’00’’ N., long. 120°58’00’’ E.); 53 meters; coral sand; February 15, 1908 (1, U.S.N.M., 36033). Dr. Th. Mortensen’s Pacific Expedition 1914-1916; off Jolo; about 36 meters; lithothamnion; March 17, 1914 (1); 36-55 meters; sand and coral; March 19, 1914 (1). Ternate (west of north central Halmahera), Moluccas [Pfeffer, 1900). Willebrord Snellius station 60*; lat. 6°58’00’’ N., long. 121°52’30’’ E.; 72-80 meters [A. H. Clark, 1936] (1, L. M.). Siboga station 257; Kei Islands, in Du-Roa Strait; down to 52 meters; coral bottom; December 11, 1899 [A. H. Clark, 1918] (2, Amsterdam Mus.). Siboga station 260; Kei Islands (lat. 5°36’30’’ S., long. 182°55’12’’ E.); 90 meters; sand, coral and shells; December 16 and 18, 1899 [A. H. Clark, 1918] (2, U.S.N.M., E. 403; Amsterdam Mus.). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; station 26; about 90 meters; sand; April 16, 1922 (3). Siboga station 144; anchorage north of Salomakieé (Damar) Island; 45 meters; bottom coral and lithothamnion; August 7-9, 1899 [A. H. Clark, 1918] (2, Amsterdam Mus.). Siboga station 305; mid-channel in Solor Strait (east of Flores); off Kampong Menanga; 113 meters; stony bottom; February 8, 1900 [A. H. Clark, 1918] (3, U.S.N.M., E. 417; Amsterdam Mus.). Siboga station 65a; Kambaragi Bay, Tanah Djampeah; 400-120 meters; pale gray mud, changing to coral bottom during the haul; May 6, 1899 [A. H. Clark, 1918] (2, Amsterdam Mus.). Geographical range-—From the Bonin and Philippine Islands and the Macclesfield Bank to the Moluccas and the Kei and Lesser Sunda Islands. Bathymetrical range.-—From 36 to 183 meters; the average of 11 definite records is 82 meters. History—Among the many comatulids that he collected near Bohol in the Philippines, Prof. C. Semper found one specimen with only five arms. At first he believed it to be a young individual of some species having when fully developed 10 or more arms, but since it did not agree in its more significant characters with any other Philippine species it appeared advisable to him to describe it as a distinct form. He considered that the unique feature of the possession of only five undivided arms entitled it to recognition as a representative of a distinct genus, so he described it under the name of Ophiocrinus indivisus. A MONOGRAPH OF THE EXISTING CRINOIDS 169 In the Challenger collection Dr. P. H. Carpenter found three species of 5-armed comatulids that he considered congeneric with Semper’s Ophiocrinus indivisus. In a paper published in 1882, first pointing out that Ophiocrinus Semper, 1868, is pre- occupied by Ophiocrinus Sultan 1853, and suggesting Hudiocrinus as a pabsaitwtes he redescribed the only known specimen of Ophiocrinus (Eudiocrinus) indivisus, which was now in the collection of Dr. W. B. Carpenter, his father, and also described the three new species from the Challenger collection. These last, which he called Eudiocrinus varians, E. semperi, and E. japonicus, are now referred to the genus Pentametrocrinus. Carpenter in the Challenger reports published in 1884 and 1888 mentioned various features connected with this species but without bringing out any new in- formation. In 1894 Prof. F. Jeffrey Bell described and figured in detail a supposedly new species from the Macclesfield Bank, which he called Eudiocrinus granulatus. At that time the type specimen of Semper’s indivisus was in the collection of W. Percy Sladen, and Professor Bell was able to compare it directly with his single specimen of granulatus. In 1900 Prof. Georg Pfeffer recorded E. indivisus without comment from Ternate. In 1911 I recorded and gave notes upon five specimens dredged by the Albatross at station 5356 in the Philippines, and in 1913 I published notes upon Bell’s type specimen of &. granulatus, which I had examined in the British Museum in 1910. In 1918 I recorded and gave notes upon 11 specimens dredged by the Siboga at five stations in the Kei and Lesser Sunda Islands. Dr. Torsten Gislén in 1922 recorded and gave notes upon a specimen from Dr. Sixten Bock’s station 59 in the Bonin Islands, discussing the structure of the species, based upon a study of this specimen, at considerable length in 1924. EUDIOCRINUS SERRIPINNA A. H. Clark PuaTe 12, Figure 49 [See also vol. 1, pt. 2, fig. 267 (arm), p. 207.] Eudiocrinus serripinna A. H. Cuarx, Smithsonian Misc. Coll., vol. 52, 1908, p. 211 (description; Albatross station 5136); Crinoids of the Indian Ocean, 1912, p. 102 (synonymy; locality); Amer. Journ. Sci., vol. 40, 1915, p. 62 (listed); Unstalked crinoids of the Siboga-Exped., 1918, p. 65 (in key; range), p. 71 (references; notes; stations 49a, 310; possibly only a variety of E. indivisus), pp. 271, 276 (listed). Diagnostic features ——Resembling EH. indivisus, but the proximal pinnules are very sharply triangular in section; on the proximal pinnules, especially on P, and P2, the distal edges of the segments are very strongly produced at the prismatic angles dor- sally, so that these pinnules have a strongly serrate dorsal profile. The arms are 40-65 mm. long, and there are 12-16 (usually 14-16) cirrus segments. Description —The centrodorsal is discoidal, with the rather broad dorsal pole flat. The cirri are arranged in a partially double marginal row. The cirri are XIX, 12-16 (usually 14-15), 5-10 mm. in length. The first seg- ment is short, the second is about as long as broad, and those following increase in 20824440 12 170 BULLETIN 82, UNITED STATES NATIONAL MUSEUM length so that the fourth and those immediately succeeding are from about as long as the median width to half again as long as the basal width. The second-seventh or -eighth segments are centrally constricted with expanded distal ends, this expan- sion of the ends being especially marked dorsally so that the dorsal profile of the cirri is rather strongly serrate, but this dies away distally so that after about the tenth segment the dorsal profile becomes smooth and straight. The fifth is a more or less marked transition segment. The cirri are basally rounded in cross section, later becoming laterally compressed, the distal portion consequently appearing broader in lateral view. The opposing spine is prominent, central in position, and in height does not reach one-half the width of the penultimate segment. The termi- nal claw is longer than the penultimate segment and is abruptly curved basally. The five arms are 45 to 65 mm. long. The IBr, and the fifth and following brachials have the distal edge produced and overlapping and scalloped, the scallops being composed of groups of fine spines, so that the proximal portion of the arms is strongly serrate in lateral view. After the proximal third of the arms these scallops merge into a uniformly finely spinous distal edge. The brachials have a narrow and distinct, though low and rounded, median carination, which is most prominent in the basal third of the arms. The proportions of the lower pinnules are the same as in L£. individus, but they have the distal ends on the dorsal side very strongly produced giving them a strongly serrate profile, this feature being especially marked on those pinnules that are enlarged. Notes.—The largest specimens collected by Dr. Th. Mortensen off Jolo in about 46 meters have the arms 45-55 mm. in length; the cirri have 14-16 segments, of which the longest are only slightly longer than broad to half again as long as broad at the base. P, has 11 or 12 segments. One of the specimens from the Danish Expedition to the Kei Islands station 24 has the arms 60 mm. long; the other two are smaller. One of the specimens from the Danish Expedition to the Kei Islands station 26 has the arms about 45 mm. long. The cirri have 12 segments. The brachials have a narrow and distinct, though low and rounded, carination which is most prominent in the basal third of the arms. The IBr, (second postradial ossicle) and the lower brachials have produced and everted distal ends so that the basal portion of the arms is strongly serrate in lateral view. The fifth and following brachials (the seventh and following ossicles beyond the radials) have the distal edge produced and overlapping and scalloped, the scallops being composed of groups of fine spines. After the proximal third of the arms these scallops merge into a uniformly finely spinous distal edge. The specimens from the Danish Expedition to the Kei Islands station 53 have the arms up to 65 mm. in length. Those from station 54 also have the arms up to 65 mm. long. The specimen from Siboga station 310 has the cirri XIX, 14-15 (usually the latter), 10 mm. long. The first segment is short and those following gradually increase in length so that the fourth and those succeeding are about as long as the median width. In the longest cirri the fifth-eighth may be one-third again as long as the basal width. The fourth-seventh or -eighth segments are centrally con- A MONOGRAPH OF THE EXISTING CRINOIDS 171 stricted with expanded distal ends, this expansion of the ends being especially marked dorsally so that the dorsal profile of the cirrus is rather strongly serrate, but this dies away distally so that after the tenth segment the dorsal profile becomes smooth and straight. The fifth is a more or less marked transition segment. The five arms are 55 mm. long. P, is composed of 12-13 segments as in the type specimen. This individual differs from the type only in its slightly greater size. The specimen from Siboga station 49a has the ornamentation of the brachials slightly more accentuated than in the preceding. Parasite —The largest individual from the Danish Expedition to the Kei Islands station 24 has a small parasitic gastropod of the genus Melanella (see Part 2, pages 645-649) attached to it. Localities—Albatross station 5136; in the vicinity of Jolo (Sulu); Jolo Light bearing S. 37° E., 0.7 mile distant (lat. 6°04’20’’ N., long. 120°59’20’’ E.); 40 meters; sand and shells; February 14, 1908 [A. H. Clark, 1908, 1912, 1915] (1, U.S.N.M., 25437). Dr. Th. Mortensen’s Pacific Expedition 1914-1916; off Jolo (Sulu); about 46 meters; sand and coral; March 19-20, 1914 (11); same, 36-55 meters; March 19, 1914 (1). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; station 24; 100 meters; hard bottom; April 15, 1922 (3). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; station 26; about 90 meters; sand; April 16, 1922 (4). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; station 53; 85 meters; sand and coral; May 9, 1922 (11). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; station 54; 85 meters; sand and coral; May 9, 1922 (10). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; station 57; 85 meters; sand and coral; May 9, 1922 (4). Siboga station 310; north of the eastern end of Sumbava, Lesser Sunda Islands (lat. 8°30’ S., long. 119°07’30’’ E.); 73 meters; sand with a few pieces of dead coral; February 12, 1900 [A. H. Clark, 1918] (1, Amsterdam Mus.). Siboga station 49a; Sapeh Strait, between Sumbava and Komodo (lat. 8°25/30/’ S., long. 119°04’36’” E.); 69 meters; coral and shells; April 14, 1899 [A. H. Clark, 1918] (1, Amsterdam Mus.). Geographical range-—From the Philippines to the Kei and Lesser Sunda Islands. Bathymetrical range-—From 40 to 100 meters; the average of 10 records is 69 meters. Remarks.—This species is known from the single specimen dredged by the Albatross in 1908 and described later in the same year; from two specimens dredged by the Siboga in 1899 and 1900 and recorded in 1918; and from 33 specimens dredged by Dr. Th. Mortensen in the Philippines in 1914 and in the Kei Islands in 1922 and herein first recorded. 172 BULLETIN 82, UNITED STATES NATIONAL MUSEUM EUDIOCRINUS ORNATUS A. H. Clark Puare 11, Ficure 40; Prats 12, Ficure 47 {See also vol. 1, pt. 1, fig. 250 (centrodorsal), p. 253, figs. 465, 466 (centrodorsal), p. 359; pt. 2, figs. 29-31 (radial pentagon), p. 20, fig. 190 (side view), p. 111, figs. 4836-439 (pinnule tips), p. 261, fig. 712 (disk), p. 346.) Eudiocrinus ornatus A. H. Cuarx, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 76 (nomen nudum; arms and pinnules compared with those of E. minor); Proc. U. S. Nat. Mus., vol. 36, 1909, p. 633 (14°04’30’" N., 93°51'00"’ E.; description); Crinoids of the Indian Ocean, 1912, p. 99 (synonymy; detailed description; localities), fig. 4, p. 100; Amer. Journ. Sci., vol. 40, 1915, p. 62 (listed); Unstalked crinoids of the Siboga-Exped., 1918, p. 65 (in key; range), p. 70 (syn- onymy; detailed description; Sta. 294; possibly a variety of HE. indivisus), p. 275 (listed), pl. 17, fig. 29; Rec. Indian Mus., vol. 34, pt. 4, 1932, p. 551 (listed), p. 554 (Investigator station 387; notes), p. 556 (compared with Z. philenor). Eudiocrinus minor A. H. Ciarx, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 75 (description; Andaman Islands); Crinoids of the Indian Ocean, 1912, p. 102 (synonymy; detailed description; Andaman Islands), fig. 5, p. 103; Amer. Journ. Sci., vol. 40, 1915, p. 62 (listed) ; Unstalked crinoids of the Siboga-Exped., 1918, p. 65, footnote (synonym of E. ornatus). Diagnostic features—Resembling E. indivisus, but the brachials have very prominent, everted, and overlapping distal ends, giving the animal a curiously ornate appearance. Description.—The centrodorsal is a thin disk, with the bare polar area flat, 2.5 mm. in diameter. The cirrus sockets are arranged in a single marginal row. The cirri are XVIII, 17-18, 20 mm. long. The first segment is twice as broad as long, the second is nearly or quite as long as broad, the third-fifth are twice as long as the proximal diameter, the sixth is slightly shorter, a more or less marked transition segment, and the following gradually decrease in length so that the ter- minal segments are only slightly longer than broad. The penultimate segment is about as long as broad. The third-sixth segments are very strongly constricted centrally with the distal edge all around produced, except on the dorsal side; from the seventh onward both these features become less marked, and the cirrus becomes somewhat compressed laterally. There are no dorsal processes. The opposing spine is sharp, prominent, arising from the entire dorsal surface of the penultimate segment, equal to about one-half of the lateral width of that segment in height. The ter- minal claw is equal in length to the penultimate segment, and is stout and strongly curved. The ends of the basal rays are visible as small tubercles in the angles of the calyx. The distal borders of the radials project slightly beyond the rim of the centro- dorsal, and are slightly concave. The IBr, and IBr, form a syzygial pair, which is oblong, from one-third to one-half again as broad as long, with the lateral edges straight, barely in apposition basally, and the ventrolateral border slightly produced. The five arms are 85 mm. long. The first brachial is oblong, about three times as broad as long, the second is slightly wedge-shaped, about the same size, and the first syzygial pair (composed of brachials 3+-4) is slightly longer on one side than on the other, half again as broad as the median length. The next three brachials are approximately oblong, two and one-half times as broad as long, and those following becoming triangular, as broad as long, and after the proximal fourth of the arm wedge-shaped, as long as broad, and in the terminal portion somewhat longer. The A MONOGRAPH OF THE EXISTING CRINOIDS 173 lower brachials have on each side, as far as the lowest pinnule, a slightly produced ventrolateral edge, corresponding with that on the IBr series. Viewed in profile the brachials have a somewhat concave dorsal surface and very prominent distal ends which are strongly overlapping and give the animal a curiously ornate appearance. Syzygies occur between the third and fourth brachials (the fifth and sixth post- radial ossicles), again between the eighth and ninth, and distally at intervals of three, more rarely four, muscular articulations. Po is 5.5 mm. long, moderately stout basally, tapering evenly to the tip, rather strongly prismatic, with 12 segments, of which the first is short, the second is not quite so long as broad, the third and fourth are squarish, and the following gradually increase in length, becoming nearly or quite twice as long as broad terminally. P, is similar to Po, with the same number of segments, but somewhat stouter and not tapering so rapidly. P, is 8.5 mm. long, much stouter than Po, gradually tapering from the base to the tip, with 12 or 15 segments, of which the first three are about as long as broad and these following very gradually become elongated and about twice as long as broad distally; the pinnule is rounded prismatic. P, is similar to P,. P, is 6 mm. long, slender, cylindrical, less stout basally than P,, gradually tapering and becoming very delicate in the terminal portion, with 15 or 16 segments, of which the first is short, the second and third are about as long as brcad, and the following gradually increase in length, becoming nearly or quite three times as long as broad in the terminal portion. P; is similar to P,. The following pinnules are similar, gradually decreasing in length to 5mm., then very slowly increasing, reaching a length of 10 mm. distally. The distal ends of the segments of the lower pinnules are more or less produced and spinous. The disk bears a few rather large plates along the ambulacra and is well plated in the anal area. Notes.—The preceding description is based upon specimens from Jnvestigator station 61. In the specimen from Investigator station 387, P, reaches 12 mm. in length and is composed of 18 segments, of which the distal are three to four times as long as broad. The single specimen collected by the Investigator and labeled Andaman Islands with no further data was described as the type specimen of a new species, Hudiocrinus minor. In this specimen the centrodorsal is a thin disk, with the bare flat dorsal pole 1 mm. in diameter. The curus sockets are arranged in a single marginal row. The cirri are XII, 12, 5 mm. long. The first two segments are twice as broad as long, the third is half again as broad as long, and the fourth is twice as long as broad. The following segments very gradually decrease in length, the antepenultimate being about one-third again as long as broad. The third and fourth segments are strongly con- stricted centrally, the fifth and sixth are slightly constricted, but the following have practically straight profiles. The cirri are proximally almost circular in cross section, after the fifth segment becoming laterally compressed and therefore broader in lateral view. There are no dorsal spines or projections. The opposing spine is median, small, scarcely equaling one-fourth the width of the penultimate segment in height. The arms and pinnules resemble those of EL. indivisus, the overlapping of the brachials and pinnule segments being moderately marked. The arms are 15 mm. long. The color in alcohol is white. 174 BULLETIN 82, UNITED STATES NATIONAL MUSEUM One of the specimens from Siboga station 294 may be described as follows: The centrodorsal is discoidal, with the cirrus sockets arranged in two rows. The bare dorsal pole is 2.5 mm. in diameter. The cirri are XXI, 18, 12 mm. long, and compara- tively slender. The segments increase in length to the fourth-sixth, which are about twice as long as the proximal width, and then decrease in length so that the distalmost seven or eight are about as long as broad. The longer segments are centrally con- stricted with widely flaring distal ends. The distal edge of the radials and the proxi- mal border of the IBr, are slightly thickened and everted, and there is a slight tubercle in the middle of the latter. The distal edge of the [Br series is everted, the middle fourth being thickened into a transversely elongate tubercle. The arms are about 50 mm. long. The central half of the distal border of the proximal oblong brachials is everted and spinous, and the distal edge of the following brachials is slightly produced and finely spinous. The lateral borders of the [Br series are finely dentate. Po is 5 mm. long and is composed of 12 segments. P; is 7.5 mm. long with 15 segments. P, is 12 mm. long with 16-18 segments, which at first are short, becoming about as long as broad on the fourth, twice as long as broad on the eighth or ninth, and slightly longer terminally. The pinnule is only moderately stout. The segments overlap slightly along the prismatic crest, where there is a slight development of spines on their distal ends. In profile the prismatic crest is slightly notched. The other specimen from Siboga station 294 is similar to the one described. The four specimens from the Danish Expedition to the Kei Islands station 26 have the arms up to 50 mm. long. Localities.—Investigator station 61; eastnortheast of Preparis Island, between the Andamans and the delta of the Irrawadi river (lat. 14° 54’30’’ N., long. 93° 51/00’ E.); 75 meters; November 30, 1889 [A. H. Clark, 1909, 1912, 1915] (6, U.S. N. M., 35133, 35135, 35136, 36244; I. M.). Investigator station 387 ; off Cape Negrais, Burma (lat. 15° 25’ N., long. 93°45’ E.); 73-90 meters; sand and coral; November 16, 1909 [A. H. Clark, 1932] (1, I. M.). Investigator; 8 miles west of Interview Island, Andamans; 493-82 meters [A. H. Clark, 1912] (2, U. S. N. M., 35144, 35163). Investigator; Andaman Islands [A. H. Clark, 1912] (1, I. M.). Siboga station 294; off southwestern Timor (lat. 10°12/12’’ S., long. 124°27/18’” E.); 73 meters; soft mud with very fine sand; January 23, 1900 [A. H. Clark, 1918] (2, Amsterdam Mus.). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; station 26; about 90 meters; sand; April 16, 1922 (4). Geographical range-—From southwest of the delta of the Irrawadi river south- ward and southeastward to the Lesser Sunda and the Kei Islands. Bathymetrical range.—According to the definite records, from 73 to 90 meters. History.—Eudiocrinus minor was described from a single specimen taken by the Royal Indian Marine Surveying steamer Investigator in the Andaman Islands in a paper published on April 17, 1909. The last paragraph of the account of this new species reads: “Arms and pinnules as in Eudiocrinus ornatus, the overlapping of the brachials and pinnulars being moderately marked .. .” On June 19, 1909, the detailed de- scription of Hudiocrinus ornatus, based upon a single specimen from Investigator station 61, appeared. This description had been written and submitted for publica- A MONOGRAPH OF THE EXISTING CRINOIDS 175 tion before the description of EZ. minor, which accounts for the appearance of Eudio- crinus ornatus, © new name, without any explanation in the description of E. minor. In 1912 both Eudiocrinus ornatus and E. minor were redescribed and figured. Of E. ornatus there were recorded six specimens from the type locality and one from another locality in the Andamans, while of E. minor the type was the only specimen mentioned. In 1918 Eudiocrinus minor was placed, without explanation, in the synonym of E. ornatus. The name ornatus was used in preference to minor for the reason that it is based upon a fully developed individual with all its characters developed, whereas the name minor is based upon a small and young specimen concerning the exact determination of which it is possible to entertain a certain amount of doubt. Technically the mention of the overlapping of the brachials and pinnulars of E. ornatus in the comparison with E£. minor may be considered as an ‘‘indication’’ of its characters as that term is used in nomenclature, and therefore the names minor and ornatus may be regarded as of the same date. In 1918 two specimens were recorded from Siboga station 294. In 1932 I recorded and gave notes upon a specimen from Investigator station 387. EUDIOCRINUS GRACILIS A. H. Clark Puate 12, Figurn 48 [See also vol. 1, pt. 2, fig. 189 (side view), p. 109.] Eudiocrinus gracilis A. H. Cuarx, Crinoids of the Indian Ocean, 1912, p. 315 (description; Investi- gator station 387); Unstalked crinoids of the Siboga-Exped., 1918, p. 65 (in key; range).— Gistin, Nova Acta Reg. Soc. Sci. Upsaliensis, ser. 4, vol. 5, No. 6, 1922, p. 7 (Burma; also represented by a somewhat divergent form [pulchellus] in the Bonin Islands), p. 72 (compared with E. g. var. pulchellus) —A. H. Cuarx, Temminckia, vol. 1, 1936, p. 295 (listed), p. 301 (Suvadiva Atoll; notes); John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1936, p. 99 (range), p. 103. Diagnostic features —In general resembling EL. indivisus, but P, and P, are about twice as long as Po and P,, relatively slender, with much elongated segments of which the third is half again as long as broad and the distal are about three times as long as broad. Description.—The centrodorsal resembles that of E. indivisus. The cirri are X, 15-16, 9 mm. long. The first segment is about three times as broad as long, the second is about twice as broad as long, the third is about as long as the width of its expanded distal end, and the fourth and fifth are about twice as long as their proximal width, the fifth being slightly longer than the fourth. The following segments are about as long as their distal width. The cirri taper gradually to the fifth segment, which is a well marked transition segment, and are more delicate from that point onward. The longer earlier segments are strongly constricted centrally with prominent flaring distal ends. Beyond the fifth segment the ventral profile becomes straight, but the distal dorsal edge is produced so that the dorsal profile of the cirrus as a whole is strongly serrate. The antepenultimate segment is slightly longer than broad, without any production of the distal dorsal edge. The penul- timate segment is wedge-shaped, about as long as the greater (ventral) length. The opposing spine is prominent, arising from the entire dorsal surface of the penultimate 176 BULLETIN 82, UNITED STATES NATIONAL MUSEUM segment, with the apex terminal, and is equal in height to about one-half the width of the segment. The terminal claw is about as long as the penultimate segment and is rather stout; it is strongly curved proximally, but becomes more slender and straighter distally. The arms are essentially like those of E. indivisus and measure 55 mm. in length. Pe is 3.5 mm. long with 9 or 10 segments. It is small and weak, rounded-pris- matic, and tapers evenly from the base to the tip. The segments are at first short, becoming about as long as broad on the third and twice as long as broad, or even longer, terminally. P, is 4.5 mm. long with 9 or 10 segments, resembling Pe but proportionally larger and stouter. P, is 7 mm. long, much larger and stouter than the preceding or succeeding pinnules, with 10 segments of which the first is slightly over twice as broad as long, the second is about half again as broad as long, and the third is half again as long as broad; the following segments gradually increase in length and become about 3 times as long as broad distally. The pinnule is much more slender than the corresponding pinnule in the other species of the genus. The second and following segments have rather strongly produced distal edges which are armed with fine spines. P, is similar but slightly longer and larger. P, is 5 mm. long, slender, with 12 segments, in general resembling the preceding pinnules. P, is 3 mm. long with 9 or 10 segments which after the third become much elongated, small, and weak. The distal pinnules are very slender, 6 mm. long. The color in alcohol is a uniform dark purple, the cirri beyond the transition segment being nearly white. Notes.—The second specimen from the type locality has the cirri XIII, 16-18 (usually 18), 10 to 12 mm. long; the sixth is a transition segment. Pg is 4 mm. long, with 10 segments. P, is 9 mm. long, with 11 segments. P, is 6.5 mm. long, with 14 segments. The individual from the Kei Islands is apparently of this species, although the identification is not quite positive. In the larger specimen from the Maldive Archipelago the cirri are XVI, 13-16. The first segment is very short, the second is somewhat more than twice as broad as long, the third is nearly twice as long as the median width, the fourth is about twice as long as the width of the proximal end, the fifth is not quite so long, and those follow- ing slowly decrease in length to the tenth or eleventh, which is half again as long as the width of the proximal end, then increase in length slightly so that the antepen- ultimate is from half again to twice as long as broad. The second-seventh segments are rather strongly constricted centrally, this feature decreasing distally so that the last two or three segments before the penultimate are oblong. The opposing spine is small, sharp, erect, and subterminal. The terminal claw is about as long as the penultimate segment, and is stout, rapidly tapering, and moderately curved. Pg is 2.5 to 3mm. long with 9 segments. It is stout basally, rapidly tapering to a fine tip, and is strongly prismatic. The first segment is short, the second is twice as broad as long, the third varies from nearly as long as broad to one-third again as broad as long, the fourth is from one-third to one-half again as long as broad, and those follow- ing are about twice as long as broad, becoming longer distally. The first three or four segments are rather strongly carinate, the crest of the carination being parallel with the axis of the pinnule. P, is 3.5 mm. long with 10 segments, similar to Pg but A MONOGRAPH OF THE EXISTING CRINOIDS 177 larger. The second-fourth segments are strongly carinate, and the fifth is slightly so. P, is 5 mm. long with 11 segments, somewhat broader basally than Pg, tapering evenly from the base to a slender and delicate tip, slender and rather flexible. The first segment is about twice as broad as long, the second is slightly broader than long, the third is half again as long as broad, the fourth is twice as long as broad, the fifth is nearly or quite three times as long as broad, and those following are still longer, the distal about six times as long as broad and very slender. The pinnule is slightly, though distinctly, carinate, and the first three segments have slight dorsal processes on the side toward the arm tips. The ends of the segments are very slightly and finely spinous, the spines being longest at the prismatic angles. P, is 5 mm. long and re- sembles P,. Syzgyies occur between the elements of the IBr series, and between brachials 3+-4, 8+9, 13+14, and 17+18, and distally at intervals of 3 muscular articulations. Localities —Investigator station 387; off Cape Negrais, Burma (lat. 15° 25’ N., long. 93° 45’ E.); 73-89 meters; sand and coral; November 16, 1909 [A. H. Clark, 1912] (2, U. S. N. M., 35141; I. M.). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; station 54; 85 meters; sand and coral; May 9, 1922 (1, C. M.). Suvadiva atoll, Maldive Archipelago; lagoon; 80 meters or less; Waillebrord Snellius, May 4, 1929 [A. H. Clark, 1936] (2, L. M.). Remarks.—This species is as yet known only from the two specimens dredged by the Royal Indian Marine Surveying steamer Investigator in 1909 and described in 1912, another from the Kei Islands collected by Dr. Th. Mortensen and not previously recorded, and two from the Maldive Islands dredged by the Willebrord Snellius in 1929. EUDIOCRINUS PULCHELLUS Gislén Eudiocrinus gracilis var. pulchellus GisLin, Nova Acta Reg. Soc. Sci. Upsaliensis, ser. 4, vol. 5, No. 6, 1922, p. 4 (127-182 m.), p. 69 (Bock’s stations 45, 46, 47, 48, 53; description and discus- s10N). ei, gracilis pulchellus Gisthn, Nova Acta Reg. Soc. Sci. Upsaliensis, ser. 4, vol. 5, No. 6, 1922, p. 6 (Bonin Islands), pp. 8, 73, 182, 183 (listed), figs. 60-63, p. 70, pl. 1, fig. 5 —Gistén, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, p. 17. Eudiocriuus gracilis pulchellus GisL&N, Nova Acta Reg. Soc. Sci. Upsaliensis, ser. 4, vol. 5, No. 6, 1922, p. 182 (listed). Diagnostic features —This species is closely related to EH. gracilis and possesses similar long-segmented P, and P;; but these pinnules are scarcely longer than the other proximal pinnules instead of being markedly longer as is the case in EL. gracilis. Description — According to Gislén the centrodorsal is flattened, 1.5 mm. in di- ameter, with the free dorsal pole 1 mm. in diameter. The cirri are XIX, 13-15, from 5.5 to6.5 mm. in length. The first two segments are broader than long, the third is as long as broad, and the fourth-sixth are about twice as long as the median width. The following segments become shorter again, the antepenultimate being about as long as broad. The radials are smooth. The IBr series is as long as broad. The five arms are 40 mm. in length. The brachials do not overlap, even in the distal portion of the arms. The width of the first brachial is 0.8 mm. 178 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Syzygies occur between brachials 3-+4, 9+10, and 13+14, and distally at intervals of 3 muscular articulations. Po is 2.5 mm. long, with 10 segments. P, is 3 mm. long, with 10 segments. P, is 4 mm. long, with 10 segments. P, is 4 mm. long, with 12 segments. The second segment of P, and P, is not quite so long as broad, the third segment is one- third to one-half again as long as broad, and the distal segments are three times as long as broad, somewhat expanded at the distal end but only slightly spiny. P, is 3.5 mm. long, with 11 segments. P, is 3.5 mm. long, with 12 segments. The distal pinnules are 5.5 mm. long with 15 or 16 segments, which are three to four times as long as broad, and of which the outermost bear dorsal hooks. There are three pairs of large sacculi on each segment. The disk has been lost; it was about 2 mm. in diameter. The proximal muscular articulations are dark colored. From about the seventh the brachials are wedge-shaped, each with a pair of dark longitudinal bands. Notes —The specimen described was from station 46. Another specimen from the same station showed the following characters: The cirri are XVII, 10-14, from 4.5 to6.5mm.long. The arms are 30 mm.long. Pg, is 2 mm. long, with 8 segments. P, is 2 mm. long, with 8 segments. P, is 3.5 mm. long, with 10 segments. P, is 3.2 mm. long, with 10 segments. P, has 12 segments. The distal pinnules are 5 mm. long, with 14 segments. The third segments of P; and P, are one-third again as long as broad. The disk is 1.5 mm. in diameter, and the anal tube is 0.8 mm. high. In a specimen from Bock’s station 47 the cirri are XVI, 13-16, from 6.5 to 8 mm. long. The fifth segment is twice as long as broad and is centrally constricted. The arms are 40 mm.long. Pg is 2.5 mm. long, with 8 segments. P, has 8 segments. The third segments of P,; and P, are one-third again as long as broad. P, is 5 mm. long with 13 segments. P, is 4.5 mm. long, with 12 or 13 segments. The distal pinnules are 7 mm. long, with 18 segments, which are three to three and one-half times as long as broad. The disk is 2 mm. in diameter. In another specimen from station 47 the cirri are XVI, 12-15, from 5 to 6 mm. long. The cirrus segments are a little shorter than in the preceding specimen, one-half to three-quarters again as long as broad. The arms are 35 mm. long. Pg is 2 mm. long, with 7 segments. P, has 8 segments. P, is 4.2 mm. long, with 13 segments, of which the third is half again as long as broad. P, has 12 segments. P, is 4 mm. long, with 12 segments. The disk is 1.8 mm. in diameter. In a specimen from station 48 the cirri are XIV, 14-16, from 6 to 7 mm. long. The arms are 30 mm. long. Pg is 2 mm. long, with 8 segments. P, is 2.2 mm. long, with 9 segments. P, is 3.7 mm. long, with 13 segments, of which the third is one- fourth again as long as broad. P, is 3.5 mm. long, with 11 segments. The distal pinnules are 4.5 mm. long, with 14 segments, which are twice as long as broad. In a small specimen from station 53 the cirri are XV, 8-12, from 2 to 4 mm. long. The arms are 10+ mm. long. There are strong prominences on the proximal ossicles as in the smallest specimen from station 45. Po is 1.2 mm. long, with 6 segments. P, is 1.2 mm. long, with 7 segments. P, has 7 segments, of which the third is one-half to three-quarters again as long as broad. In a specimen from station 45 the cirri are XV, 14-15, from 5 to 6.5 mm. long. The arms are 40 mm. long. Pg is 2.2 mm. long, with 8 segments. P, is 2.7 mm. long, A MONOGRAPH OF THE EXISTING CRINOIDS 179 with 9 segments. P, is 4.5 mm. long, with 10 segments. P, is 4.5 to 5 mm. long, with 11 segments. The distal pinnules are 6.2 mm. long, with 18 segments, which are three times as long as broad. The disk is 2 mm. in diameter. In another specimen from station 45 the centrodorsal is 1 mm. in diameter with the free dorsal pole 0.5 mm. in diameter. The cirri are XIX, 13-15, from 4 to 6.5 mm. long. The arms are 35 mm. long. Pg is 2 mm. long, with 7 or 8 segments. P, is 2.5 mm. long, with 9 segments. P, is 3.5 mm. long, with 11 segments. The third segment of P, and P, is one-third again as long as broad. P, has 12 segments. The distal pinnules are 6 mm. long, with 18 segments. The disk is 2 mm. in diameter. In a third specimen from station 45 the cirri are XVIII, 12-13, from 3.5 to 5 mm. long. The arms are 25 mm. long. P, is 2.2 mm. long, with 9 segments. P, is 2.4 mm. long, with 9 segments. P, is 3.3 mm. long, with 10 segments. P, is 3 mm. long, with 10 segments. P, is 3.2 mm. long, with 12 segments. P; is 3.2 mm. long, with 13 segments. The disk is 2 mm. in diameter. In a fourth specimen from station 45 the cirri are about XX, 13-14, from 4 to 5 mm. long. The arms are 23 mm. long. The proximal border of the IBr, is very slightly everted. Pc is 1.8 mm. long, with 6 or 7 segments. P, is 2.2 mm. long, with 8 segments. P, is 3.2 mm. long, with 9 segments. P, is 2.7 mm. long, with 9 segments. P, is 3 mm. long, with 10 segments. P3 is 3.5 mm. long, with 11 seg- ments. P, and P; have rather serrate distal segments. The disk is 2 mm. in diameter. This specimen, according to Gislén, is transitional between the larger specimens and the very small ones from stations 45 and 53. In a small specimen from station 45 the cirri are XX, 11-12, from 2.5 to 3 mm. long. The length of the third segment is one-fourth more than the width of the expanded distal end. The fourth segment is twice as long as the distal width and is strongly constricted centrally. The fifth and sixth segments are proximally a little narrower. The segments from the seventh to the antepenultimate are half again as long as broad and are laterally compressed. The opposing spine is sharp, its length equaling about half the width of the penultimate segment. The terminal claw is pointed, curved, and somewhat longer than the penultimate segment. The five arms are about 20 mm.long. The radials are visible as a narrow band with a median promi- nence; they are separated at the interradial angles by the basals. The IBr, are proximally everted with a median keel which in height, as viewed laterally, equals one-third the thickness of the ossicle. There is a similar prominence on the distal part of the IBr, and the first two brachials where laterally, however, it has no distinct border. The ends of the segments are slightly everted. The brachials from the eighth onward have oblique ends. Syzygies occur between brachials 3-+4 and 13+14, and distally at intervals of three (or four) muscular articulations. Po, is 1.5 mm. long with 6 segments. P, has 6 segments of which the second-fourth are half again as long as broad. P, is 2 mm. long with 8 segments of which the third-fifth are twice as long as broad. The distal ends of the pinnule segments are spiny and are somewhat everted. The segments of the distal pinnules are very long, four or five times as long as broad. The disk is 1 mm. in diameter. The anal tube is smooth and subcentral. The mouth is subcentral. The sacculi are dark. The color in alocohol is yellowish white. 180 BULLETIN 82, UNITED STATES NATIONAL MUSEUM The small specimens from stations 45 and 53 are, according to Gislén, evidently young individuals of this form although of an ornate type. Localities.—Dr. Sixten Bock’s Expedition to Japan station 46; Bonin Islands, east of the Channel; 128 meters; August 1, 1914 [Gislén, 1922]. Dr. Sixten Bock’s Expedition to Japan station 47; Bonin Islands; east of the Channel; 146 meters; August 1, 1914 [Gislén, 1922]. Dr. Sixten Bock’s Expedition to Japan station 48; Bonin Islands; east of Chichi- jima; 183 meters; August 1, 1914 [Gislén, 1922]. Dr. Sixten Bock’s Expedition to Japan station 53; Bonin Islands, 2 miles east of Higashijima; 164 meters; sand and broken shells; August 8, 1914 [Gislén, 1922]. Dr. Sixten Bock’s Expedition to Japan station 45; Bonin Islands, east of Chichi- jima; 146 meters; July 31, 1914 [Gislén, 1922]. Geographical range-—Known only from the Bonin Islands. Bathymetrical range——Apparently from 128 to 183 meters; but the depths as given represent the length of dredging wire out, not the actual depths. Remarks.—This species is known only from the 11 specimens collected by Dr. Sixten Bock in 1914 and described by Dr. Torsten Gislén in 1922. EUDIOCRINUS gp. Eudiocrinus sp. A. H. Cuarx, Temminckia, vol. 1, 1936, p. 302. Locality —Binonko, Toekay Besi Island; about 10 meters; Willebrord Snellius, March 7-10, 1930. Remarks.—A small undeterminable individual was dredged at this locality. Family HIMEROMETRIDAE A. H. Clark Savignyii group P. H. Carpenter, Journ. Linn. Soc. (Zool.), vol. 16, No. 95, 1882, p. 508. Basicurva group (part) P. H. Carpenter, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 102 (Antedon denticulata only).—Harrtiavs, Bull. Mus. Comp. Zool., vol. 27, 1895, p. 130.—A. H. Cuark, Proc. U. 8S. Nat. Mus., vol. 33, 1907, pp. 127, 128; Smithsonian Misc. Coll., vol. 50, 1907, pp. 343-345, 348, 359, 361; Proc. U. S. Nat. Mus., vol. 36, 1909, p. 365. Milberti group (part) P. H. Carpenter, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 192 (Antedon pinniformis, A. milberti, A. laevissima [part], A. tessellata, A. anceps, and A. variipinna).— P. H. Carpenter, Journ. Linn. Soc. (Zool.), vol. 21, 1889, p. 304 (species collected at Mergui).— Harrtiavs, Nova Acta Acad. German., vol. 58, No. 1, 1891, pp. 13, 80; Bull. Mus. Comp. Zool., vol. 27, p. 137 (structural position).—A. H. Crarx, Smithsonian Misc. Coll., vol. 50, 1907, pp. 343, 344 (characters differentiating this group not diagnostic), p. 345 (heterogeneous char- acter of the group), p. 349 (Tropiometra, gen. nov., includes part of the species), p. 357 (Cyllo- metra, gen. nov., includes part of the species)—HartLtaus, Mem. Mus. Comp. Zool., vol. 27, 1912, p. 370. Palmata group (part) P. H. Carpenter, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 223 (Antedon clemens). (For further references to this group see page 391.) Savignyi group P. IT. Carpenter, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 252 (all the included species except Antedon angustiradia)—Hartiaus, Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 13 (diagnosis; range; 16 species), p. 15 (diagnosis; discussion), p. 16 (key to the species included); Bull. Mus. Comp. Zool., vol. 27, 1895, p. 137 (structural position). Savignii group A. H. Ciark, Smithsonian, Mise. Coll., vol. 50, 1907, p. 343 (characters differentiating this group not diagnostic), p. 355 (Himerometra, gen. nov., includes part of species), p. 364 (one species [angustiradia] referred to Adelometra, gen. nov.). A MONOGRAPH OF THE EXISTING CRINOIDS 181 Himerometridae A. H. Ciark, Proc. Biol. Soc. Washington, vol. 21, 1908, p. 126 (covering Oligo- metra, gen. nov.), p. 135 (including Himerometra, Pontiometra, Cylometra, and Oligometra); Proc. U. S. Nat. Mus., vol. 34, 1908, p. 210 (in key), p. 211 (includes Oligometra, Cyllometra, Himerometra, and Pontiometra; range; not represented in the Hawaiian Islands), p. 212 (range of included genera); vol. 35, p. 119 (footnote), fig. 20, p. 120 (arm structure); Amer. Nat., vol. 42, 1908, p. 541 (characteristic of the Indo-Pacific-Japanese faunal area), p. 725 (color) ; Geogr. Tourn., vol. 32, 1908, p. 602 (characteristic of the Indo-Pacific-Japanese faunal area), p. 605 (ecology); Proc. Biol. Soc. Washington, vol. 22, 1909, p. 1 (revision; description of new genera); Amer. Nat., vol. 43, 1909, p. 254 (represented in the Red Sea); Proc. U. S. Nat. Mus., vol. 36, 1909, p. 361 (part of Antedon as understood by P. H. Carpenter), p. 365 (mouth often more or less eccentric in certain species); Proc. Biol. Soc. Washington, vol. 22, 1909, p. 144 (Mariametra, gen. noy., described in this family), p. 174 (referred to the Comatulida Oligo- phreata), p. 175 (divided into three subfamilies, Himerometrinae, Stephanometrinae, and Mariametrinae, and Pontiometra removed to a separate family, Pontiometridae); Vid. Medd. Nat. Foren. Kgbenhavn, 1909, p. 119 (occurs at Singapore; has large eggs that presumably develop rapidly; ecology), p. 148 (arm tips in Comatula pectinata normally as in species of this family), p. 152 (cirri compared with those of Zygometra fluctuans [=comata]); Proc. U. S. Nat. Mus., vol. 40, 1911, p. 6 (12 species in Africa), p. 8 (species on southeast coast), p. 9 (species on northeast coast), p. 10 (West Indian and corresponding East Indian genera), p. 649 (referred to the Oligophreata); Die Fauna Siidwest-Australiens, vol. 3, Lief. 13, 1911, p. 438 (3 genera and 7 species in Australia); Mem. Australian Mus., vol. 4, 1911, p. 717 (proportion of species of this family in Australia as known to P. H. Carpenter), p. 720 (proportion of species in Aus- tralian fauna), p. 725 (Himerometra absent from Australia), p. 728 (3 subfamilies, Himero- metrinae, Mariametrinae, and Stephanometrinae, raised to family rank), p. 730 (in key), p. 731 (key to Australian genera)—Hartiavs, Mem. Mus. Comp. Zool., vol. 27, 1912, p. 371.— A. H. Cuark, Crinoids of the Indian Ocean, 1912, p. 6 (family exclusively confined to East Indian region; number of included genera and species), p. 11 (represented in the Ceylon region by Amphimetra, Himerometra, and Heterometra), p. 12 (represented in Red Sea region by Cras- pedometra and Heterometra; represented in southeast African region by Amphimetra, Cras- pedometra, and Heterometra), p. 21 (range in detail; exclusively littoral, not extending deeper than 50 fathoms), p. 42 (family discarded, the 3 included subfamilies raised to family rank), pp. 46, 49, 53 (in keys), p. 56 (key to included genera); Die Fauna Siidwest-Australiens, vol. 4, Lief. 6, 1913, p. 311 (discussion, and comparison with Zygometridae); Bull. Inst. Océanogr. Monaco, No. 294, 1914, pp. 7, 8 (relation to temperature) ; Journ. Washington Acad. Sci., vol. 4, 1914, pp. 559-563 (correlation of geographical and bathymetrical ranges), p. 582 (relation to temperature of habitat).—ALEXANDER, Rec. Western Australian Mus., vol. 1, pt. 8, 1914, p. 108.—A. H. Cuarx, Rec. Western Australian Mus. vol. 1, pt. 3, 1914, p. 114 (species collected by Endeavour in Western Australia); Amer. Journ. Sci., vol. 40, 1915, p. 67 (detailed philo- sophical discussion of bathymetrical range); Internat. Rev. gesamt. Hydrobiol. und Hydrogr., 1915, p. 223 ff. (detailed account of distribution of Australian species) ; Smithsonian Mise. Coll., vol. 65, No. 10, 1915, p. 45 ff. (phylogenetical study); Unstalked crinoids of the Siboga Exped., 1918, p. 72 (key to the included genera).—GisLEN, Kungl. Svenska Vet.-Akad. Handl., vol. 59, No. 4, 1919, p. 23.—A. H. Cuarx, Univ. Iowa Stud. Nat. Hist., vol. 9, No. 5, 1921, p. 14 (not represented in West Indies); Smithsonian Mise. Coll., vol. 72, No. 7, 1921, p. 3—Gus.L&n, Zool. Bidrag Uppsala, vol. 9, 1924, pp. 19, 35, 42, 55, 77, 84, 89, 99, 118, 230, 231, 235, 240, 280, 287; Arkiv for Zool., vol. 19, No. 32, 1928, p. 4A. H. Cuark, Journ. Linn. Soe. (Zool.), vol. 36, 1929, p. 638; Rec. Indian Mus., vol. 34, pt. 4, 1932, p. 556; Treubia, vol. 14, livr. 2, 1933, pp. 207, 211.— Gustin, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 8, 18, 22, 23, 27, 28, 43, 44, 45, 46, 47, 48, 49, 53.—A. H. Cuarx, Temminckia, vol. 1, 1936, p. 302.— Gistfin, Kungl. Fysiogr. Sallsk. Lund Férh., vol. 7, No. 1, 1936, p. 8—A. H. Crarx, John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1936, pp. 99, 103.— Gistin, Kung]. Svenska Vet.-Akad. Handl., ser. 3, vol. 17, No. 2, 1938, pp. 4, 10.—H. L. Crarxk, Mem. Mus. Comp. Zool., vol. 55, 1938, p. 33.—Boons, Bull. Vanderbilt Marine Mus., vol. 7, 1938, p. 122. 182 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Himerometrinae A. H. Cuark, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 175 (includes Amphi- metra, Himerometra, Craspedometra, and Heterometra); Mem. Australian Mus., vol. 4, 1911, p. 728 (raised to family rank), p. 730 (in key, through inadvertence) ; Journ. Washington Acad. Sci., vol. 5, 1915, pp. 126-134 (bathymetrical range; phylogenetical and palaeontological significance). Himerométres A. H. Cuarx, Bull. Mus. Hist. Nat., Paris, No. 4, 1911, p. 250. Himorometridae A. H. Cuarx, Mem. Australian Mus., vol. 4, 1911, p. 728 (typographical error). Diagnosis.—A family of the superfamily Mariametrida in which the elements of the [Br series are united by synarthry and the arms are 10 or more in number; if there are more than 10 arms the IIBr series are typically 4(3+4), though in any individual some of these may be replaced by IIBr series of two elements; the brachials are always short and are frequently—always in the 10-armed species—exceedingly short with almost or quite parallel ends. The persiome of the disk is usually naked, rarely bearing small scattered concretions, which are especially evident in the anal area (see Part 2, p. 228). The lateral perisome of the pinnules usually contains simple, forked, or multiradiate spicules, very rarely delicate filmy plates of from 2 to 15 meshes, between which and the edge of the pinnulars are small straight or rarely forked spicules or small rods; but calcareous deposits are frequently wholly absent (see Part 2, pp. 239-241). Geographical range-—From southern Japan southward to the Philippines, Tonga and Fiji, Port Molle, Queensland, and to between Fremantle and Geraldton, Western Australia, and westward to east Africa, from the Red Sea southward to Zanzibar and Bagamoyo. Bathymetrical range.—From the intertidal zone down to 111 (?183) meters. Remarks.—The 41 forms included in the family Himerometridae fall into five genera—Himerometra, Craspedometra, Homalometra, Heterometra, and Amphimetra. Amphimetra is rather sharply differentiated from the other four, which together form a closely knit unit. The six species of Himerometra are easily distinguished by their narrow and well- separated division series combined with the very short brachials and the much en- larged proximal pinnules, which decrease in size from the lowest outward. But in Himerometra bartschi and in H. persica the proximal pinnules are no stouter than they are in many species of Heterometra and, as in the latter, have the lower segments strongly carinate, while the lowest is sometimes shorter than the next. Also in Heterometra variipinna P2, though by no means greatly enlarged, is of quite the same type as the enlarged proximal pinnules in Himerometra magnipinna and in H. sol. Thus Himerometra shows a very close approach to two apparently different genetic lines within the genus Heterometra, although the two genera seem not actually to intergrade. An interesting feature of the species of Himerometra is that typically the two outer III Br series on each ray are 4(3+-4) and the two inner are 2, though this arrange- ment is by no means invariable. This distribution of IIIBr series of 4(3-+-4) and of 2 is not confined to Himerometra, for it occurs also in Craspedometra and is frequently more or less strongly evident in the large species of Zygometra (microdiscus and elegans) and in Heterometra philiberti. Although when typically developed its appearance is very distinctive, the single species of Craspedometra is very closely allied with certain species of Heterometra. A MONOGRAPH OF THE EXISTING CRINOIDS 183 Indeed, it is probably more closely allied with them than they are with other species included within the same genus. But there is a more definite break between Craspe- dometra and Heterometra than occurs anywhere within the species groups included in the latter genus. The single species of Homalometra, like that of Oraspedometra, is merely an offshoot from one of the several genetic lines included in thegenus Heterometra; but there is a considerable gap between Homalometra and the most closely related species of Heterometra. The eight species and forms included in the genus Amphimetra form a closely knit unit that is rather abruptly distinct from the other types included in the family Himerometridae. Though in its arm structure, particularly in the very short bra- chials, Amphimetra closely approaches one of the groups within the genus Heterometra, and also the genus Himerometra, it shows certain features which indicate an approach to the Colobometridae. Thus through individual variation specimens of Amphimetra tessellata tessellata develop I1Br series, which are always of 2 elements and not 4(3-++4) as in the other genera of Himerometridae, and other specimens have the dorsal processes on the cirrus segments much broadened and somestimes forked. Like the species of the Zygometridae, the species of Himerometridae are espe- cially characteristic of the region from southern Japan to northern Australia and the eastern portion of the Bay of Bengal. One genus (Amphimetra) extends westward to the Maldive archipelago, one (Himerometra) is represented by a single species in the Persian Gulf, and a third (Heterometra) is represented by several species on the east African coast. All the genera but one (Homalometra) are represented at the low-tide mark, and the included species are chiefly to be found in very shallow water. None of the species have been dredged from a greater depth than 111 (or possibly 183) meters. History—The family Himerometridae was first mentioned in my paper on new genera of unstalked crinoids published on April 11, 1908, as a simple heading under which the new genus Oligometra was described. Later in the same paper the genera of free crinoids belonging to the Comatulida were grouped in several families, which were not defined. Among these was the family Himerometridae, including the genera Himerometra, Pontiometra, Cyllometra, and Oligometra. In a paper of mine published on May 14, 1908, a key to the families of Comatulida having recent representatives was given, in which the family Himerometridae was included. The family Himerometridae was said to contain the genera Oligometra, Cyllometra, Himerometra, and Pontiometra, and its range was given. In my paper published on October 30, 1908, the arm structure and brachial homologies of the family Himerometridae were discussed in detail. In a revision of the family Himerometridae published on January 9, 1909, the following new genera, including species previously assigned to the large and hetero- geneous genus Himerometra, were added to it — Colobometra, Amphimetra, Cenometra, Craspedometra, Stephanometra, Heterometra, and Dichrometra—and the new genus Ozxymetra was mentioned but not defined. In a paper on the new genus Comatilia published on April 27, 1909, I remarked that I had become convinced that the division of the recent comatulids into two contrasted groups, called by Carpenter Antedon and Actinometra, was very artificial and could 184 BULLETIN 82, UNITED STATES NATIONAL MUSEUM not stand the test of modern systematic methods. Instead of these two genera I therefore proposed to recognize five great divisions, the families Comasteridae, Zygometridae Himerometridae, Tropiometridae, and Thalassometridae, which covered exactly the same ground except that the number of included species was nearly, if not quite, doubled. I remarked that each of these families appeared to me to be separated from the others by characters of just as great importance as those separating the Comasteridae (the old genus Actinometra) from any one of them. In the same paper I noted that in the species of Himerometridae the mouth is often more or less eccentric. In a paper on new recent Indian crinoids published on June 25, 1909, I used the family name Colobometridae as a heading under which two new species of Cenometra (herdmani and insueta) and one of Cyllometra (soluta) were described, thus restricting the family Himerometridae by the removal of Cenometra, Cyllometra, and Colobometra, and also Oligometra, although the last was not mentioned. In a paper published on September 14, 1909, I still further restricted the family Himerometridae by removing the genus Pontiometra to the new family Pontiome- tridae. At the same time I divided the family Himerometridae into three subfamilies, Himerometrinae, including the genera Amphimetra, Himerometra, Craspedometra, and Heterometra; Stephanometrinae, including the genus Stephanometra; and Mariame- trinae, including the genera Mariametra and Dichrometra. In a paper on the recent crinoids of Australia published on August 17, 1911, I said that recent discoveries have shown that the Zygometridae are not nearly so sharply differentiated from the Himerometridae as was previously supposed, and it has therefore seemed best te discard the latter family altogether, raising the three subfamilies previously included within it to family rank. The three families into which the Himerometridae are divided are the Himerometridae, Stephanometridae, and Mariametridae. Since that time the family Himerometridae has remained unchanged. The group now known as the family Himerometridae was somewhat indefinitely indicated by Dr. P. H. Carpenter in 1882 as the Savignyii group of Antedon. In Carpenter’s report on the Challenger comatulids published in 1888 the Savignyi group is essentially the equivalent of the family Himerometridae, except for the inclusion of Antedon (Adelometra) angustiradia. But the 10-armed species, and 10-armed individuals of multibrachiate species, are included in the Milberti group. Thus the species of Amphimetra are assigned to the Milberti group, and Antedon anceps (= Heterometra quinduplicava) and Antedon variipinna (= Heterometra crenulata) occur both in the Milberti and in the Savignyi groups. One species, Antedon (Homalometra) denticulata was placed by Carpenter in the Basicurva group, and another, Antedon clemens (= Heterometra quinduplicava), in the Palmata group. KEY§TO THE GENERA OF THE FAMILY HIMEROMETRIDAE a‘. Pp and P, longer and stouter than P;, which in turn is longer and stouter than P;; division series strongly rounded dorsally and well separated; usually more than 25 arms; the IIIBr series are usually 4(3-+-4) on the outer side of each IIBr series and 2 on the inner side (Philippine Islands and Macclesfield Bank southward to Admiralty Islands, Great Barrier Reef, and Kei Islands, and westward to Maldive Archipelago and Persian Gulf; 0-57 [766] meters) ___-____ Himerometra (p. 185) A MONOGRAPH OF THE EXISTING CRINOIDS 185 a?, P; (and Pp when present) shorter and more slender than P3. b!. Cirri gradually tapering to a sharp point distally, the distal cirrus segments proportionately as long as, or longer than, the proximal, twice as long as broad or even longer, and entirely without dorsal processes; no opposing spine; terminal claw long and nearly straight; outer portion of cirri very slightly, sometimes not at all, recurved. c!, Cirri very long and stout, one-third to almost half as long as the arms, with 40-60 (usually 45-50) segments; proximal portion of the cirri light in color, each segment with a ventral purple saddle, the distal portion brown; division series usually broad, well rounded dorsally, not in lateral contact; large and robust with 20-36 arms 90-130 mm. long (Philippines and Hongkong southward to Amboina and westward to Andaman Islands and Gulf of Martaban; 0-33 meters) 22202 6 Bid tae eh be va Craspedometra (p. 214) c. Cirri small and very slender, less than one-third the arm length, with 25-30 segments, and entirely white; slender species with 10 or 11 arms 70-90 mm. long (Arafura Sea to western endiofiNew Guinea :!90-95: meters) pene) see ee Homalometra (p. 337) b?, Distal cirrus segments always with dorsal tubercles or spines, and the terminal claw always strongly curved; distal cirrus segments usually shorter than the proximal, or both distal and proximal cirrus segments broader than long. c!. More than 10 arms; enlarged lower pinnules much elongated, plain, or with the lower seg- ments carinate, or the outer segments bearing flange-like processes or having very spinous distal ends; in some species 10-armed individuals occur; these are easily distinguished by the strongly carinate or otherwise modified proximal pinnules (southern Japan southward to Port Curtis, Queensland, and Baudin Island, northwestern Australia, and westward to Madagascar and east coast of Africa from Red Sea to Zanzibar and Bagamoyo; 0-111 Meters) 22) Pk 2. ape hohe? bioertoreeny iO) yb) moat Heterometra (p. 225) c?, Arms 10; enlarged lower pinnules but slightly differentiated from those following, not much enlarged and but little elongated; occasional individuals may have more than 10 arms; in these the IIBr series are 2 instead of 4(3+4) (southern Japan southward to the Philip- pines, Fiji, and Tonga, Port Molle, Queensland, and to between Fremantle and Geraldton, Western Australia, and westward to Maldive Archipelago; 0-109 [7183] meters)________ Amphimetra (p. 340) Genus HIMEROMETRA A. H. Clark Actinometra (part) P. H. Carpenter, Notes Leyden Mus., vol. 3, 1881, p. 201. Antedon (part) Hartiaus, Nachr. Ges. Géttingen, May 1890, p. 182, and following authors. Himerometra A. H. Cuarxk, Smithsonian Mise. Coll., vol. 50, 1907, p. 350 (variation in the component species), p. 355 (diagnosis; genotype Antedon crassipinna Hartlaub, 1890); Bull. Mus. Comp. Zool., vol. 51, 1908, p. 247 (characters; genotype Antedon crassipinna Hartlaub, 1890); Proc. Biol. Soc. Washington, vol. 21, 1908, p. 135 (referred to Himerometridae); Proc. U. S. Nat. Mus., vol. 34, 1908, p. 211 (referred to Himerometridae), p. 212 (occurs in Japan [error]), pp. 308, 317; vol. 35, 1908, fig. 20, p. 120 (arm structure); Amer. Nat., vol. 42, 1908, p. 722 (ecology); Proc. Biol. Soc. Washington, vol. 22, 1909, p. 1 (revision), p. 7 (included species, as restricted); Amer. Nat., vol. 43, 1909, p. 256 (absence from Red Sea); Zool. Anz., vol. 34, 1909, p. 367 (arm division similar to that of Zygometra elegans); Proc. Biol. Soc. Washington, vol. 22, 1909, p. 175 (referred to Himerometrinae); Vid. Medd. Nat. Foren. Kgbenhayn, 1909, p. 152 (brachials compared with those of Zygometra fluctuans [comata]), p. 160 (brachials compared with those of Craspedometra acuticirra); Die Fauna Siidwest-Australiens, vol. 3, Lief. 13, 1911, p. 455 (a specimen of Comanthina belli recorded with the arm division characteristic of this genus); Mem. Australian Mus., vol. 4, 1911, p. 725 (absent from Australia); Crinoids of the Indian Ocean, 1912, p. 9 (absent from Australia), p. 10 (absent from Japan; reason), p. 11 (represented in Ceylon region; this the western limit of the large and highly multibrachiate species), p. 12 (occurs in Persian Gulf but not in Red Sea), p. 21 (range in detail), p. 56 (in key), p. 114 (original reference; genotype).—Harriaus, Mem. Mus. Comp. Zool., vol. 27, 1912, pp. 370, 371 (discussion).—A. H. Cuarx, Die Fauna Siidwest-Australiens, vol. 4, Lief. 6, 1913, p. * 311 (comparison with the large species of Zygometridae); Proc. U. 8S. Nat. Mus., vol. 46, 1913, pp. 279-289 (detailed account, with revision of the included species); Unstalked crinoids of the 208244—40——13 186 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Siboga-Exped., 1918, p. 72 (in key; range), p. 73 (key to the included species).—Gis.6n, Zool. Bidrag Uppsala, vol. 9, 1924, p. 35 (brachial homologies), p. 51 (obliquity of brachials), pp. 92, 99 (pinnule articulations)—Gist6n, Kungl. Fysiogr. Siillsk. Hand!., new ser., vol. 45, No. 11, 1934, pp. 40, 45, 46, 48, 52, 53, 54, 57. Heterometra (part) A. H. Cuarx, Vid. Medd. Nat. Foren. Kgbenhavn, 1909, pp. 164, 165, 193; Crinoids of the Indian Ocean, 1912, p. 127. Craspedometra (part) Gisu6N, Kungl. Fysiogr. Sillsk. Handl., new ser., vol. 45, No. 11, 1934, p. 22. Diagnosis.—A genus of Himerometridae in which Pp is much longer and stouter than P,, which in turn is much longer and stouter than P.; the division series are narrow, strongly rounded dorsally, and widely separated laterally; the II[Br serieson the outer side of each I[Br series are 4(3-++-4), but on the inner side usually 2; the brachials are exceedingly short, discoidal, with produced and overlapping distal ends. The species are large and robust with 20-62 (but seldom less than 30) arms, which are 100 to 200 (commonly 130-150) mm. in length. Geographical range-——From the Philippine Islands and the Macclesfield Bank south to the Admiralty Islands, the Moluccas, the Kei Islands, and the Great Barrier reef, Queensland, and westward to the Maldive Islands and the Persian Gulf. Bathymetrical range.—From the shoreline down to 57 (?66) meters; chiefly littoral and sublittoral. Remarks.—The species of the genus Himerometra are very easily distinguished from all the other forms in the family Himerometridae by the fact that Pp is the longest and largest pinnule, the pinnules succeeding gradually decreasing in size to P;. In addition to the characteristic pinnule structure, the strongly rounded, narrow, and well-separated division series and exceedingly short discoidal brachials are features that aid in the prompt generic determination of the included forms. The six species of the genus recognized herein are divisible into three groups of two each. The first group, including H. bartschi and H. persica, is rather sharply separated from the others. The slender proximal pinnules of these two species, which have carinate basal segments, resemble in structure the corresponding pinnules in certain species of Heterometra very closely—much more closely than they do the pinnules of the other species of Himerometra—and it is interesting to note that although Pp is almost invariably longer than Pp and Pp is longer than P,, in a specimen of H. bartschi from the Kei Islands (see page 211) Pp is slightly longer and usually somewhat stouter than Pp. In these two species also the pinnules following the enlarged proximal pinnules have the basal segments strongly carinate, a feature common in the species of Heterometra but not found in the other species of Himerometra. Another peculiarity of these two forms is that their outer cirrus segments bear more or less prominent spines instead of being smooth or bearing obsolescent tubercles. While these two species can only be regarded as belonging to the genus Himerometra, with the other species of which they agree completely in the structure of their division series and arms, they are to a certain extent intermediate between the species in the second and third groups within the genus Himerometra and the species of Heterometra. The species in these last two groups are abruptly separated from all the other forms in the family Himerometridae by the enormous stoutness of the proximal pin- nules. Indeed, pinnules of the type characteristic of these species are found else- where only in the genus Cenometra of the family Colobometridae in which P, is simi- A MONOGRAPH OF THE EXISTING CRINOIDS 187 larly modified. These two groups are distinguished from each other by the number of the segments in the proximal pinnules. In one group (including H. sol and H. magnipinna) the stout proximal pinnules have about 30 segments and taper slowly distally to a delicate tip, while in the other (including H. robustipinna and H. martensi) they are composed of about 20 segments and taper very rapidly. These two groups seem quite distinct; but it may be mentioned that in a specimen of H. robustipinna from the Kei Islands Reichensperger found Pp to be composed of 22-24 segments. Each of the three groups is represented in the Malayan region from the Philip- pines to the Kei Islands—by H. bartschi, H. magnipinna, and H. robustipinna. But the two species included in each of the three groups always have distinct ranges that do not overlap. Thus H. persica is the western representative of the East Indian H. bartschi, H. sol is the western representative of H. magnipinna, and H. martensi repre- sents the widely spread H. robustipinna at Singapore and in British North Borneo. With the exception of H. robustipinna, specimens of all the species from any one locality are very uniform in their characters. But H. robustipinna, which appears to have the most extensive range of all the included forms, seems to be very variable. This variation does not seem to be geographical, for Reichensperger has pointed out that a specimen from the Philippines (the type of my H. robustipinna) is unquestion- ably the same as Hartlaub’s specimen (the type of Antedon kraepelini) from Burma, and at the same time he recorded two specimens from the Kei Islands in one of which Pp has 22-24 segments while in the other it has only 18-20 segments. The species of the genus Himerometra are among the very largest of the tropical comatulids, being exceeded in size, when fully developed, only by the largest species of Comasteridae and perhaps by Pontiometra andersoni in the Colobometridae. They are all inhabitants of very shallow water, none having been recorded from a depth greater than 57, or possibly 66, meters. Specimens of all of them have been taken in shore collecting, although they are commonest a few meters beneath the surface. History —The genus Himerometra was first proposed, with the genotype Antedon crassipinna Hartlaub, 1890, in a revision of the old genus Antedon published by me in 1907. At that time it was considered as including all the multibrachiate species of comatulids with unplated ambulacra except those in the family Comasteridae and those assigned to the genera Zygometra, Pontiometra, and Cyllometra, which were described at the same time; a few 10-armed species, now placed in the genera Amphi- metra and Oligometrides, were also assigned to it. In a revision of the family Himerometridae published in 1909 the genus Himero- metra, by the removal of various species to the new genera Amphimetra, Cenometra, Craspedometra, Stephanometra, Heterometra, and Dichrometra, was restricted to its present limits, except only that Himerometra philiberti (J. Miiller) was retained in it with a query. KEY TO THE SPECIES IN THE GENUS HIMEROMETRA a!, Enlarged proximal pinnules very stout with all, or nearly all, of the component segments broader than long, or at least as broad as long; none of the segments in the proximal pinnules are carinate; following pinnules without carinate processes on the earlier segments. b!, Enlarged proximal pinnules with about 30 segments, very stout basally and distally tapering gradually to a delicate and more or less flagellate tip; the distal edges of the segments in the middle half or proximal two-thirds of the proximal pinnules are swollen and may be strongly everted, but they are always smooth, never spinous. 188 BULLETIN 82, UNITED STATES NATIONAL MUSEUM c!, Cirri very stout, stouter than in any other species of the genus; the enlarged proximal pinnules have the segments in the basal two-thirds with strongly produced and everted distal edges CPanel baa 1s) en a es ree ee sol (p. 188) @. Cirri rather short and weak, without, or with only slight traces of, dorsal processes on the distal segments; the enlarged proximal pinnules have the segments in the middle half with slightly swollen distal edges (from Cochinchina and Philippines south to southern Celebes and Kei Islands and eastward to Admiralty Islands; 0-38 meters) _--magnipinna (p. 189) b?, Enlarged proximal pinnules with about 20 segments, distally tapering more or less abruptly and without a flagellate tip. c!. Segments of the enlarged proximal pinnules entirely smooth; the distal edges of the com- ponent segments may be swollen, or they may be unmodified, but they are never spinous; distal edges of the proximal brachials smooth, or only very slightly produced (from Philip- pine Islands and Macclesfield Bank southward to Moluccas, Kei Islands, and Great Barrier Reef and westward to Ceylon; 0-57 [766] meters) ---..----- robustipinna (p. 193) c. Segments of the enlarged proximal pinnules with the distal ends projecting abruptly beyond the bases of the segments succeeding (but not overlapping) and armed with very numerous fine spines (in large specimens on the distal segments only); distal ends of the proximal brachials very strongly produced and everted (Singapore and North Borneo; littoral) martensi (p. 203) a. Enlarged proximal pinnules slender, flagellate distally, and very long (slightly more than half the length of the cirri), composed of 36-40 perfectly smooth segments most or all of which are longer than broad; a few of the earlier segments are narrowly, but prominently, carinate; earlier segments of the following pinnules very strongly carinate; segments in the outer half of the cirri with prominent dorsal spines. b'. Arms 35-61 in number; about 40 cirrus segments; a few of the basal segments in the proximal pinnules are broader than long (from Philippine Islands and Singapore south to Amboina pridwWeulelands|O0—bL Meters) =. =< a> 2a ae ee eee ane bartschi (p. 209) . Arms 20-25 in number; about 35 cirrus segments; all the segments in the proximal pinnules are longer thar broad’ (Persian Gulf) = 2222222022 se ea eee ie eee persica (p.212) HIMEROMETRA SOL A. H. Clark Antedon palmata Bet, in Gardiner, Fauna and geography of the Maldive and Laccadive Archi- pelagoes, vol. 1, pt. 3, 1902, p. 224 (Kolumaduli, Maldives; 38 fathoms). Himerometra sol A. H. Cuarx, Crinoids of the Indian Ocean, 1912, p. 40 (identity of Bell’s record), p. 115 (synonymy; detailed description, based on Bell’s specimens; locality); Proc. U. S. Nat. Mus., vol. 46, 1913, p. 285 (in key), p. 288 (synonymy; range); Smithsonian Misc. Coll., vol. 61, No. 15, 1913, p. 26 (published reference to specimens in B. M.; locality; detailed description of 1 specimen and shorter description of another); Unstalked crinoids of the Siboga-Exped., 1918, p. 73 (in key; range); John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1936, pp. 99 (range), 103. Diagnostic features—The enlarged proximal pinnules are very stout basally, distally gradually tapering to a delicate and flagellate tip; they are composed of 30-32 segments nearly all of which are broader than long, or at least not longer than broad; the distal edges of the segments in the middle half or proximal two-thirds are strongly produced and everted, but are always smooth and never spinous; none of the segments of the proximal pinnules are carinate, and the following pinnules are without carinate processes on the earlier segments. The cirri are very stout, stouter than those in any other species of the genus, 25-30 mm. in length, with 25-30 seg- ments, of which the outermost 10-12 have a small and low dorsal tubercle. The 41-51 arms are 140 mm. long. Description.—The centrodorsal is thick discoidal, with a strongly concave dorsal pole 4 mm. in diameter. A MONOGRAPH OF THE EXISTING CRINOIDS 189 The cirri are XXI, 25, 27, 28, and 30, from 25 to 30 mm. long; the longest seg- ments, in the basal third, are nearly or quite as long as broad; the distal segments are slightly broader than long, sometimes as much as one-third again as broad as long; the terminal 10 or 12 segments have a small and low median dorsal tubercle, some- times scarcely noticeable until near the end of the cirrus. The opposing spine is well developed and conspicuous. The cirri as a whole are stout, stouter than in the other species of the genus, with approximately subequal segments. The IIBr series are 4(3-+-4). The IIIBr series are 4(3-+-4) externally, 2 internally. IVBr series, when present, are 4(3-++4). The division series resemble those of the other species of the genus. The 41 arms of the type specimen are 140 mm. long, resembling those of related species. Ppis 18 mm. long, with 30 to 32 segments, which are nearly twice as broad as long in the proximal half, becoming squarish in the distal third, and terminally twice as long as broad. After the fourth or fifth the segments develop strongly everted and produced distal edges, this character gradually dying away in the distal third; this eversion is smooth and not serrate. The proximal pinnules are very stout, but also very long, and taper distally to a flagellate tip as in H. magnipinna; but the eversion of the distal ends of the segments is much greater than in that species, and the cirri are much stouter. Notes.—Another specimen has the cirri XX XV, 27-29. There are 51 arms. As in the type specimen the division series are strongly convex dorsally and well sepa- rated. Only one of the division series, an internally developed ILIBr series, is 2, all the others being 4(3-+4). The proximal pinnules are exactly as in the type, and are 18-21 mm. long. Locality. Kolumaduli, Maldive Archipelago; 69 meters; Prof. J. Stanley Gar- diner [Bell, 1902; A. H. Clark, 1912, 1913] (2, B. M.). History.—The two specimens upon which this species is based were originally recorded by Prof. F. Jeffrey Bell under the name of Antedon palmata in 1902. I ex- amined them at the British Museum in 1910 and described them as representing a new species, Himerometra sol, in 1912. They were redescribed in my notes on the crinoids of the British Museum in 1913. HIMEROMETRA MAGNIPINNA A. H. Clark Puats 15, Ficures 54, 55; Puats 16, Figure 56; Puate 17, Ficurss 61, 62. [See also vol. 1, pt. 2, fig. 260 (arm), p. 205; fig. 271 (arm and pinnules), p. 207; fig. 715 (disk) p. 346.] Antedon crassipinna (part) Hartiaus, Nachr. Ges. Géttingen, May 1890, p. 185 (specimen from Cochinchina, but not described specimen from Amboina [= H. robustipinna]); Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 32 (specimen from Cochinchina, but not those from Amboina), p. 34 (characters of proximal pinnules), not pl. 1, figs. 1, 5, 10=H. robustipinna.—Gis.in, Kungl. Fysiogr. Sallsk. Lund Férh., vol. 7, No. 1, 1936, p. 2 (Cochinchina), p. 4 (French Indochina). Himerometra crassipinna A. H. Cuark, Smithsonian Mise. Coll., vol. 50, 1907, p. 356 (in part; listed); Proc. Biol. Soc. Washington, vol. 22, 1909, p. 7 (listed); Crinoids of the Indian Ocean, 1912, p. 116 (record from Cochinchina). 190 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Himerometra magnipinna A. H. Crarx, Smithsonian Mise Coll., vol. 52, 1908, p. 214 (description; Albatross stations 5139 and 5147); Proc. Biol. Soc. Washington, vol. 22, 1909, p. 7 (listed); Proc. U. S. Nat. Mus., vol. 39, 1911, p. 540 (Ulugan Bay, Palawan); Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p. 2 (Antedon crassipinna Hartlaub, in part), p. 17 (Cochinchina; detailed description), p. 18 (additional localities; descriptions); Proc. U. S. Nat. Mus., vol. 48, 1912, p. 394 (St. Mathias Island); Crinoids of the Indian Ocean, 1912, p. 114 (synonymy; Philippine Islands; St. Mathias Island; 0-21 fathoms); Proc. Biol. Soc. Washington, vol. 26, 1913, p. 179 (range in east Asia); Proc. U. S. Nat. Mus., vol. 46, 1913, p. 285 (in key), p. 287 (synonymy; range); Journ. Washington Acad. Sci., vol. 5, 1915, p. 214 (Malayan species; range and its significance); Unstalked crinoids of the Siboga-Exped., 1918, p. 74 (in key; range; synonymy; notes; station 209), p. 274 (listed) —Gistin, Zool. Bidrag Uppsala, vol. 9, 1924, p. 41 (details of arm structure), pp. 44, 51 (reversion), p. 77 (syzygies), p. 283, fig. 8, p. 45 (arm base), fig. 98, p. 87 (synarthrial face), figs. 144-148, p. 95 (pinnule articulations), fig. 149, p. 98 (same).— A. H. Cuark, Treubia, vol. 14, livr. 2, 1933, p. 211 (Nassi Besar); Proc. Biol. Soc. Washington, vol. 47, p. 9, 1934 (first record from Singapore), p. 11 (Singapore; notes).—Gistin, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 41, 46, 47, 50, 52—A. H. Cuarx, Temminckia, vol. 1, 1936, p. 302 (Ternate; notes) —GistiNn, Kungl. Fysiogr. Sallsk. Lund Férh., vol 7, No. 1, 1936, p. 2 (Cochinchina), p. 4 (French Indochina), p. 6 (range), p. 8 (Pulo Condor; notes). Diagnostic jeatures.—The enlarged proximal pinnules are long, very stout basally, and gradually taper distally to a delicate and flagellate tip; they are composed of 28-42 (usually 30-35) segments, nearly all of which are broader than long, or at least as broad as long; the distal edges of the segments in the middle half of the proximal pinnules are swollen but are smooth and not spinous; none of the segments of the proximal pinnules are carinate, and the following pinnules are without carinate processes on the earlier segments. The cirri are rather short, 25-41 (usually 30-35) mm. in length, and are composed of 28-40 (usually 30-35) segments of which the distal are wholly without, or bear only very slight, dorsal processes. There are 33-62 (usually about 45) arms which are 130-184 (usually about 145) mm. long. Characters of the type specimen.—The cirri are XV, 28-32 (usually 28-30), 30 mm. long. The distal segments are smooth, or bear only very faint dorsal tubercles. The 62 arms are 120 mm. long. The proximal pinnules are very stout and very long, tapering evenly from the base to the tip; they are composed of 28-29 segments of which the distal ends are slightly swollen but are not spinous or overlapping. Notes.—Hartlaub said that the specimen from Cochinchina, which he referred to H. crassipinna, is of a beautiful purple-violet color. In contrast to the conditions in the three specimens of H. crassipinna he described, the IBr, are entirely free laterally. There are 45 arms of which some on the outer side of the postradial series arise directly from a IJBr axillary. Compared with the specimens of H. crassipinna P, is relatively very long, reaching about 28 mm., and, in common with the pinnules following, it is somewhat more slender than the corresponding pinnules in HZ. crassipinna. I examined this specimen in the Hamburg Museum. The cirri are about XXX (with some undeveloped), 28-32, from 35 to 41 mm. long. The cirri are moderately stout. The longest proximal segments are slightly broader than long. There are no dorsal spines, but the dorsal surface of the short distal segments is convex. The antepenultimate segment bears a small tubercle which quickly becomes obsolete on the one or two preceding segments. The 45 arms are 160 mm. long. P, is 28 mm. long, with 33-34 segments, very stout basally and tapering evenly to a delicate tip. The earlier segments are twice as broad as long, those following gradually increasing A MONOGRAPH OF THE EXISTING CRINOIDS 191 in length and becoming about as long as broad on the fourteenth, and twice as long as broad on the small terminal segments. The segments in the middle half of the pinnule have the distal dorsal edge slightly swollen and prominent. P, is 26 mm. long, with 33 segments, and is similar in every way to P,. P, is 18 mm. long, with 25 segments, similar to the preceding pinnules but proportionately smaller. The cirri, centrodorsal, and division series are pinkish flesh color, the arms are slaty, and the proximal pinnules and the outer portion of the middle and outer pinnules are pinkish. Of the specimens from Port Galera, Mindoro, one (original No. 202) has 58 arms 140 mm. long; another (original No. 224) has 51 arms;a third (original No. 150) has 40 arms and 35-38 cirrus segments; in this individual the proximal pinnules are very large and stout, long, with 23-28 segments, and without a flagellate tip but tapering gradually distally; a fourth specimen has 33 arms; this last in life had the bases of the arms whitish or light straw color, the arms transversely barred with brown, the pin- nules barred whitish and brown, and the anal tube straw color spotted with light brown. In its present dry condition it is a uniform purplish brown. In the specimen from Pulo Condor, Cochinchina, as described by Gislén, the arms are broken and the cirri have been lost. The centrodorsal bears sockets for XXXIV cirri. The IIBr series are 4(3-++-4) and the IIJBr series 2, the latter usually developed internally. There are 28+ arms. P, is about 15 mm. long, with 19 seg- ments, very stout, tapering distally, composed of smooth and short segments of which the longest are scarcely as long as broad; there is a slight constriction at the articula- tions, which makes the pinnule appear somewhat knobby. ‘The pinnules are mostly broken at the tips. The color is violet. Gislén says that because of the rather small size and few arms this must be considered as a young individual. The specimen from Ulugan Bay, Palawan, is a magnificent example of the species with 42 arms 184 mm. long and the cirri XX XIII, 29-35, 30 mm. long. The color in alcohol is a uniform deep violet. One of the specimens from Bantayan Reef, Cebu, has 46 arms. In life the center of the dorsal surface was dull ochre, and the arms were crimson shading to rich chocolate at the tips. In the present dried condition it is a uniform brown. The single broken specimen from Isabela, Basilan, has 40 arms. The cirri are 32 mm. long and are composed of 32 segments of which the terminal 15 have a minute submedian tubercle, which increases in size distally. The distal edges of the segments of the proximal pinnules, which are large and very stout, are everted but not spinous. The IIBr series are 4 (3-++4). The IIIBr series are 4 (3+-4) externally and 2 internally, or all 4 (3+4). The following division series, when present, are 4 (3-4-4). In the fragmentary specimen from Ekalin, St. Matthias Island, the cirri and lower pinnules are very slightly stouter than they are in the specimen just preceding, and the segments of the lower pinnules have less prominent distal ends. The three fine specimens from Pitilu, Admiralty Islands, have 36, 44, and 47 arms 130 mm. long. The specimen from Ternate has 45 arms. The specimen from the Danish Expedition to the Kei Islands station 11 has about 40 arms 140 mm. long. The cirri are XXVII, 39-40, 35 mm. long. Pp has 42 seg- ments, and P, has 42 segments and is 25 mm. long. 192 BULLETIN 82, UNITED STATES NATIONAL MUSEUM One of the specimens from Toeal has 46 arms 135 mm. long and the cirri XXVI, 28-30, 25-30 mm. long, mostly about 30 mm. long. One external IVBr 4 (3+4) series bears externally an arm and internally a first brachial from which arises a pinnule the base of which is about half as broad as the distal face of the first brachial. The other specimen is similar, with 44 arms 130 mm. long and the cirri XXVI, 26-27, 25 mm. long. The specimen from Siboga station 209 is small and immature with 12 arms 45 mm. long. Two IIBr 4 (3+4) series are developed on a single postradial series. The cirri are XV, 18-21, 17 mm. long. The longer proximal cirrus segments are half again as long as broad and are constricted centrally so that the cirri, which are rather slender, have very much the appearance of the cirriof Stephanometra protectus or of Lamprometra palmata. The dorsal pole of the centrodorsal is slightly concave. On the arms arising directly from the 1Br axillaries P; is much larger than P., the relative proportions being the same as on the free undivided arms of the adults. Localities Singapore, 1899 [A. H. Clark, 1934] (1, Raffles Mus.). Singapore [A. H. Clark, 1934] (2, U.S.N.M., E. 3135). Nassi Besar; 9 meters; otter trawl; June 4, 1908 [A. H. Clark, 1933] (1, Buitenzorg Mus.). Cochinchina [Hartlaub, 1891; A. H. Clark, 1912] (1, H. M.). Pulo Condor, Cochinchina; 15 meters; Dr. C. Dawydoff [Gislén, 1936]. Port Galera, Mindoro, Philippines; Dr. Lawrence E. Griffin (7, M. C. Z., 631 {original Nos. 88, 91, 100, 202], 681 [original No. 224], 682, 683 [original No. 150] ). Albatross, Ulugan Bay, Palawan; December 28, 1908 [A. H. Clark, 1911] (1, U.S.N.M., 35200). Bantayan Reef, Cebu, Philippines; Dr. Lawrence E. Griffin (8, M. C. Z., 386, 394, 395). Isabela, Province of Mindanao, Philippines; Dr. H. Hallier, January 22, 1904 [A. H. Clark, 1912] (1, H. M.). Albatross station 5139; in the vicinity of Jolo (Sulu); Jolo light bearing S. 51° W., 3.6 miles distant (lat. 6°06’00’’ N., long. 121°02’30’’ E.); 36 meters; coral sand; February 14, 1908 [A. H. Clark, 1908] (1, U.S.N.M., 25440). Albatross station 5147; Sulu Archipelago, in the vicinity of Siasi; Sulade Island (E.) bearing N. 3° E., 8.4 miles distant (lat. 5°41’40’” N., long. 120°47/10’’ E.); 38 meters; coral sand and shells; February 16, 1908 [A. H. Clark, 1908] (1, U.S.N.M., 35198). Ekalin, St. Matthias Island, Admiralty Islands; Dr. G. Duncker [A. H. Clark, 1912] (1, H. M.). St. Matthias Island, Admiralty Islands; Herr Schéde [A. H. Clark, 1912] (1, Berl. M.). Pitilu, Admiralty Islands; Dr. G. Duncker [A. H. Clark, 1912] (3, H. M.). Ternate; 2-4 meters; Willebrord Snellius, June 6, 1930 (1, L. M.). Danish Expedition to the Kei Islands station 11; Dr. Th. Mortensen; 20 meters; sand; April 9, 1922 (1). Danish Expedition to the Kei Islands station 61; Dr. Th. Mortensen; about 50 meters; May 15, 1922 (1). A MONOGRAPH OF THE EXISTING CRINOIDS 193 Danish Expedition to the Kei Islands; Dr. Th. Mortensen; Vatek van Toeal; about 2 meters; rocky coast; March 23, 1922 (2). Siboga station 209; anchorage off the south point of Kabaéna Islands (off south- eastern Celebes); reef [A. H. Clark, 1918] (1, Amsterdam Mus.). Geographical range-—From Singapore, Cochinchina, and the Philippines south- ward to Kabaéna Island and the Kei Islands, and eastward to the Admiralty Islands. Bathymetrical range.—Littoral and down to 38 meters. IMstory.—The first known specimen of this species was recorded by Dr. Clemens Hartlaub from Cochinchina in 1891. He referred it to a new species from Amboina, which he described under the name Antedon [Himerometra] crassipinna, but he menr- tioned certain features whereby it differed from the three specimens he had from Amboina. In 1908 I described the species from a specimen dredged by the Albatross in the Philippines at station 5139 and at the same time recorded another specimen from station 5147. In 1911 I recorded an additional example, which had been collected by the Albatross at Palawan. In my monograph of the crinoids of the Indian Ocean published in 1912, I gave this species from the Philippines and from St Matthias Island, the latter locality being taken from the labels of unrecorded specimens in the Hamburg and Berlin Museums. Hartlaub’s specimen from Cochinchina I included with a query under H. crassipinna. Ina paper on the crinoids of the Hamburg Museum published in the same year, I referred Hartlaub’s Cochinchina specimen definitely to H. magnipinna, and recorded one specimen from Isabela, Basilan, another from Ekalin, St. Matthias Island, and three from Pitilu in the Admiralty Islands. In a paper on the crinoids of the Berlin Museum published two weeks later, I recorded another specimen from St. Matthias Island. In 1918 [ recorded and gave notes upon a very small specimen which had been dredged by the Siboga at station 209. Dr. Torsten Gislén in 1924 described and figured various structural features of this species, but did not give the origin of his specimen. In 1933 I recorded a specimen from Nassi Besar in the collection of the Buitenzorg, Java, Museum, and in 1934 I recorded three specimens from Singapore, one dated 1899, in the collection of the Raffles Museum, Singapore. In 1936 Dr. Gislén recorded and gave notes upon a young specimen from Pulo Condor, Cochinchina. HIMEROMETRA ROBUSTIPINNA (P. H. Carpenter) Piate 16, Figure 60; Pats 17, Fiaure 63; PLate 18, Fiaures 68, 69 [See also vol. 1, pt. 2, fig. 270 (arm and pinnules), p. 207; fig. 714 (disk), p. 346.] Actinometra robustipinna P. H. Carpenter, Notes Leyden Mus., vol. 3, 1881, p. 201 (description; Moluccas); Challenger Reports, Zoology, vol. 26, pt. 60, 1888, pp. 59, 330, 345, 367, 383.— A. H. Cuark, Proc. U. S. Nat. Mus., vol. 37, 1909, p. 30 (appears to be a synonym of Alecto bennetti); Notes Leyden Mus., vol. 33, 1911, p. 176 (identified as Himerometra, sp.; doubtfully referred to H. crassipinna from Singapore [martensi]}); Crinoids of the Indian Ocean, 1912, pp. 116, 117 (same). 194 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Antedon kraepelini Hantuaup, Nachr. Ges. Géttingen, May 1890, p. 183 (description; Akyab, Burma); Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 22 (detailed description; Akyab), pl. 2, figs. 15, 21—A. H. Criark, Smithsonian Misc. Coll. vol. 60, No. 10, 1912, p. 18 (type seems to be a small robustipinna). Antedon crassipinna Harriaus, Nachr. Ges. Géttingen, May 1890, p. 185 (description; Amboina {specimen from Cochinchina is H. magnipinna]); Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 11 (collected by Brock at Amboina), p. 19 (in key), p. 32 (detailed description and comparisons; Amboina [specimen from Cochinchina is H. magnipinna]), p. 113 (in Gottingen Mus.), pl. 1, figs. 1, 5, 10.—A. H. Crarx, Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p. 2 (—H?merometra crassipinna, in part); Crinoids of the Indian Ocean, 1912, p. 37 (same). Antedon inopinata Bxrxt, Proc. Zool. Soc. London, 1894, p. 398 (description; Macclesfield Bank, 31-36 fathoms).—A. H. Cuark, Proc. U.S. Nat. Mus.,vol. 34, 1908, p. 477 (not recognizable from descrip- tion; placed in wrong “group”’); Crinoids of the Indian Ocean, 1912, p. 38 (of Bell, 1894—= Himerometra inopinata); Smithsonian Mise. Coll., vol. 61, No. 15, 1913, p. 83 (of Bell, 1894— Himerometra robustipinna) . Antedon crassispina Koruter, Mém. Soc. Zool. France, vol. 8, 1895, p. 420 (Sunda Islands; Biliton; notes). Himerometra crassipinna A. H. Cuark, Smithsonian Misc. Coll., vol. 50, 1907, p. 356 (listed) ; Proc. Biol. Soc. Washington, vol. 22, 1909, p. 7 (listed); Notes Leyden Mus., vol. 38, 1911, p. 183 (possibly species represented by type of Actinometra robustipinna); Crinoids of the Indian Ocean, 1912, pp. 36, 37 (identity), p. 116 (Amboina [records from Singapore and Pulau Ubin are H. martensi, and those from Cochinchina are H. magnipinna])—ReicHENSPERGER, Abh. Senck. naturf. Ges., vol. 35, Heft 1, 1913, p. 83 (Kei Islands), p. 98 (Nuhu-Tawan, north coast of Little Kei; characters of the 4 specimens).—A. H. Cruark, Unstalked crinoids of the Siboga- Exped., 1918, p. 74 (of Reichensperger—J. robustipinna). Himerometra kraepelint A. H. Cuark, Smithsonian Mise. Coll., vol. 50, 1907, p. 356 (listed) ; Proc. Biol. Soc. Washington, vol. 22, 1909, p. 7 (listed); Crinoids of the Indian Ocean, 1912, p. 116 (synonymy; locality)—ReicHEeNspPerGER, Abh. Senck. naturf. Ges., vol. 35, Heft 1, 1913, p. 83 (Ceylon), p. 96 (Ceylon, Sarasin; characters; discussion), p. 98 (synonymy; further discussion) .— A. H. Cuark, Unstalked crinoids of the Siboga-Exped., 1918, p. 74 (identity) —Gistin, Kungl. Fysiogr. Sillsk. Handl., new ser., vol. 45, No. 11, 1934, p. 48. Comaster robustipinna A. H. Cuark, Proc. U. S. Nat. Mus., vol. 33, 1908, p. 686 (listed) . Phanogenia robustipinna A. H. Cuark, Proc. U.S. Nat. Mus., vol. 35, 1908, p. 124 (listed). Himerometra robustipinna A. H. Cuark, Smithsonian Mise. Coll., vol. 52, 1908, p. 213 (new species; description; Albatross Sta, 5165), p. 214 (compared with H. magnipinna); Proc. Biol. Soc. Wash- ington, vol. 22, 1909, p. 7 (listed); Smithsonian Misc. Coll., vol. 60, No. 10, 1912, p. 10 (A. krae- pelini possibly a synonym of this ); Crinoids of the Indian Ocean, 1912, p. 114 (of A. H. Clark, 1908, renamed H. pulcher).—Rei1cHENsPERGER, Abh. Senck. naturf. Ges., vol. 35, Heft 1, 1913, pp. 97, 98.—A. H. Crarx, Proc. U. S. Nat. Mus., vol. 46, 1913, p. 285 (in key), p. 286 (of P. H. Carpenter, 1881; synonymy; range); Smithsonian Misc. Coll., vol. 61, No. 15, 1913, p. 25 (refer- ence to the specimen in B.M. [inopinata of Bell, 1894]; locality; detailed description) ; Unstalked crinoids of the Siboga-Exped., 1918, p. 74 (in key; range).—H. L. Cuarx, Rec. Australian Mus., vol. 15, No. 2, 1926, p. 184 (off Ellison reef; 5-15 fathoms).—GisLGn, Kungl. Fysiogr. Sillsk. Handl., new ser., vol. 45, No. 11, 1934, p. 48; Kungl. Fysiogr. Sallsk. Lund Forh., vol. 7, No. 1, 1936, p. 5 (range). Himerometra sp. A. H. Cuarx, Notes Leyden Mus., vol. 33, 1911, p. 176 (Actinometra robustipinna), p. 182 (Moluccas; description; appears to be H. crassipinna [martensi]); Crinoids of the Indian Ocean, 1912, p. 117 (synonymy; remarks on the type of Actinometra robustipinna P. H. Carpenter, 1881; possibly H. crassipinna [martensi]); Smithsonian Misc. Coll., vol. 60, No. 10, 1912, p. 2 (Antedon kraepelini), p. 18 (locality; appears to be a small robustipinna [of A. H. Clark, 1908]).— REICHENSPERGER, Abh. Senck. naturf. Ges., vol. 35, Heft 1, 1913, p. 98. Himerometra pulcher A. H. Cuarx, Crinoids of the Indian Ocean, 1912, p. 114 (new name for Himero- metra robustipinna A. H. Clark, 1908, not Actinometra robustipinna [Himerometra, sp.] P. H. Carpenter, 1881; synonymy; locality); Proc. U. S. Nat. Mus., vol. 46, 1913, p. 287 (synonym A MONOGRAPH OF THE EXISTING CRINOIDS 195 of Carpenter’s robustipinna)—REICHENSPERGER, Abh. Senck. naturf. Ges., vol. 35, Heft 1, 1913, p. 97 (synonym of kraepelini), p. 98 (further discussion). Himerometra inopinata A. H. Cuarx, Crinoids of the Indian Ocean, 1912, p. 114 (type of Antedon tmopinata Bell, 1894, redescribed and shown to be a Himerometra; correction of the original depth record). Diagnostic features.—The enlarged proximal pinnules are very stout and taper rapidly to the tip; they are composed of 17-24 (but almost always 20) segments of which the distal ends may be unmodified or swollen, but are never spinous; the distal edges of the proximal brachials are smooth, or only slightly produced. There are 37-56 (usually about 45) arms, which are 100-200 (usually 130-140) mm. long. Description.—The centrodorsal is as in the other species of the genus. The cirri are XXX, 26-34, stout, 30-35 mm. long. The sixth or seventh seg- ments are the longest, from slightly broader than long to half again as broad as long. The distal segments are slightly carinate, the carination on the last six or seven ending distally in a small spine. Nine of the IIBr series are 4 (3-++-4) and two are 2; the I] Brseries are all 4 (3-++4) except one, internally developed; the remaining division series are all 4 (3+-4) except two IVBr series, which are 2. The division series are strongly convex and well sepa- rated as is usual in the genus. The 46 arms are 140 mm. long and resemble those of the other species of the genus. The proximal pinnules are very stout but nearly smooth, the distal edges of the segments being only very slightly swollen. All the segments are short, about twice as broad as long in the proximal half, becoming squarish toward the tip. The proximal pinnules are about 20 mm. long, with 17-20 segments. Notes.—The specimen from Albatross station 5165, which was described as the type of a new species Himerometra robustipinna, has 37 arms 100 mm. long. The cirri are XVIII, 30-40, 35 mm. long; they are without dorsal spines, though the last three segments may have small median tubercles. The proximal pinnules are very stout, with most of the segments, which number about 20, broader than long and smooth. Pp is 15 mm. long. In the type specimen of Antedon inopinata from the Macclesfield Bank, according to Bell, the centrodorsal is large, hollowed in the center, which is bare of cirrus pits; the cirri are arranged in three irregular rows on the side. The cirri are long and stout, about XLV, about 40, up to 28 mm. in length; the terminal segments are faintly spinous. The radials and the IBr; are broad and stout, and the IBr, (axillaries) are short at the side. The IIBr series are 4 (3-+4). There are nearly always IIIBr 4 (8+4) series present. In rare cases there may be IVBr 2 series. The arms are about 45 in number, and are about 100 mm. long. The brachials are wide, low, and very regular and are much compressed laterally. Syzygies occur between brachials 3+4, then not again for about 26 muscular articulations. The pinnules generally are pretty stout and stiff, the basal ones being very markedly stout. The color in alcohol is light brown, with the ambulacral surface of the pinnules somewhat darker. Dr. P. H. Carpenter described Actinometra robustipinna as follows: The centro- dorsal is a thick disk slightly flattened at the dorsal pole and bearing about 40 cirrus sockets in 2 marginal rows. The radials are partially visible all around the calyx, but they diverge slightly at each angle so as to leave a gap. This is bridged over by 196 BULLETIN 82, UNITED STATES NATIONAL MUSEUM the proximal portions of the short IBr, which have lateral extensions meeting those of their fellows on either side. The IBr, (axillaries) are pentagonal, quite free laterally, and rather more than twice as long as the IBr,, with moderately sharp distal angles. The post-radial series may divide three times. Six of the I[Br series are 4(3+-4) and three are 2; the remaining one is broken away. The IIIBr series, when present, are 4(3+4). The first ossicles beyond each axillary are closely united laterally. The total diameter between IIIBr axillaries is 25 mm. There are 17+ arms. The pinnules borne by the second ossicles after each axillary are enormously large and stout, so much so that the ossicles bearing them have almost the appearance of being axillaries themselves. Each pinnule contains 20+ massive segments the ventral edges of which are produced so as to stand up as plates sloping slightly inward toward the ambulacral groove. The disk is lost. The diameter of the centrodorsal is 7 mm. The color is light brownish white. Carpenter said that, although no traces remain either of the disk or of a terminal comb on the oral pinnules, the flattened calyx and the wide funnel in the center of the radial pentagon indicate the generic position of this specimen, which is markedly different from most species of Actinometra that he had seen. I examined the type specimen of Carpenter’s Actinometra robustipinna at Leyden in 1910, and in 1911 I wrote that it proved to be a specimen of a typical species of Himerometra. All the IIIBr series are 4(3-++4). The proximal pinnules are very large and stout, the tips ending bluntly after a considerable recurve. So far as they are preserved the segments are broader than long. The distal ends are not thickened or produced, though appearing slightly prominent and a trifle swollen. One P, with 18 segments appears to be complete. The three specimens collected by Dr. J. Brock at Amboina were described by Hartlaub as a new species, Antedon crassipinna, in the following terms: The centro- dorsal is broad and thick, rather deeply concave in the middle; the cirri are arranged in three marginal rows. The cirri are about XX XVII, 30-40, reaching 46 mm. in length. None of the segments are longer than broad. Toward the ends the cirri become compressed laterally and the outermost segments bear small spines. The radials are entirely concealed, or only in small part visible. The IBr, are partially united laterally. The IBr, (axillaries) are pentagonal. There are 2 or 3 division series beyond the IBr series. The IIBr series are 4(3-+-4), or some of them may be 2. The IIIBr series are 2 and 4(3+-4), sometimes the inner on a post-radial series being 2 and the outer 4(3+-4), while in other cases they are mostly 4(3+4). All the IVBr series are 4(3-++-4). The axillaries of the division series are rather large. The ossicles following the axillaries are always partially united laterally. On the articulation between the two first elements of each division series there is a dorsal tubercle, though this may be scarcely noticeable; a similar tubercle occurs on the articulation between the first two brachials, though here it is less marked. The axillaries of the division series of 4(3-++-4) form with the segment succeeding sometimes a small lateral boss over the base of the pinnule. The 46-56 arms are about 200 mm. long. They have a narrow dorsal surface and a serrate dorsal profile, and are composed of short uniform brachials. From the very base of the arm the brachials have produced distal ends, this feature becoming less and less marked toward the arm tips so that here the surface of the arms is fairly smooth. The six or seven earliest brachials, especially the first A MONOGRAPH OF THE EXISTING CRINOIDS 197 4, are somewhat longer than those following, which are shortly discoidal and in general very uniform. Syzygies occur between brachials 3+4, again from between brachials 23-+24 to between brachials 31+32, and distally at intervals of 13-15 muscular articulations. The pinnules on the division series and P, in their great thickness, size and stiff- ness contrast strikingly with the pinnules following. They have thick cylindrical segments with especially thickened distal ends which with smooth edges overlap the bases of the segments succeeding. Pp and Pp» are of about equal size, about 20 mm. in length and composed of about 20 segments. P» on the inner IIIBr series is some- what smaller than on the outer. The pinnule on the IVBR series is somewhat shorter. P,; on the outer arms is about 16 mm. long, but on the inner arms markedly shorter. The following pinnules vary a little in their length, but are always much smaller and weaker. Their length decreases to the pinnule of the eleventh brachial. The pinnules succeeding become larger again, reaching up to about 10 mm. in length. P, is 10 mm. long, about as long as P,. The segments of these pinnules have sometimes short lateral processes on their distal ends. The disk is incised, and is 32 mm. in diameter. Sacculi are crowded on the pinnules. The color is purple-violet or chocolate-brown. While the preceding description is based wholly upon Professor Brock’s three specimens from Amboina, Hartlaub included in his account as a representative of this species a specimen from Cochinchina in the Hamburg Museum, which in reality represents H. magnipinna. Hartlaub remarked that his new species is the largest of the known forms in the Savignyi group and has the largest number of arms. It is easily distinguished from all the others by its stout and massive lower pinnules. He said that it is especially interesting in possessing a feature that heretofore has not been observed in any endocyclic comatulid, although it occurs in several comasterids, that is, the presence in equal numbers of IJ1Br 2 and II/Br 4(3+-4) series. He pointed out that while in the comasterids the inner IJIBr series are 4(3-++-4) and the outer are 2, the reverse is the case in this species. He also noticed that both in this form and in the comasterids this arrangement is not wholly constant. He mentioned that while one or two IVBr series may occur on the inner IIIBr series, on the outer I11Br series there is usually none, although there may be one. He discussed the presence of IIBr 2 series, of which one of the specimens from Amboina possessed three, and said that this variation in the number of elements in the I] Br series seems not to be unusual, as it occurs in Craspedometra acuticirra, Heterometra nematodon, and Adelometra angustiradia (the last, although a macrophreate form, was placed in the Savignyi group by Car- penter). But he noted that in these species it is not so developed as it is in Comanthus parvicirra, which is characterized by the greatest irregularity in this respect. Speaking of the specimens from Little Kei, Dr. August Reichensperger said that Dr. Merton obtained four individuals of this well-characterized and fairly stable species with 36, 39, 45, and 47 arms 130 to 140 mm. in length. There is little correla- tion between the arm number and the size of the individuals. Except for one, all have a smaller arm number than was given by Hartlaub. Reichensperger remarked that in his specimens he confirmed the alternation of II]Br 2 and II/Br 4(3+4) series noted by Hartlaub. He said that in general the outer III Br series are 2 and the inner 4(3+4). This, however, is a slip; he meant to say just the opposite, as is shown by 198 BULLETIN 82, UNITED STATES NATIONAL MUSEUM his details of the division of two postradial series given below. The number of IlIBr 2 and I1JBr 4(3+-4) series he found to be approximately the same, and only in a single specimen did the IIIBr 4(3+4) series strongly predominate. He found the IIBr series 2 in a few cases. He gave two examples of the arm division in his speci- mens. In one of these one of the IIBr series is 4(3-++4) and the other is 2. The IlIBr series are 4(3+-4) externally and 2 internally. The two inner IIIBr series bear next to the midradial line of the postradial series as a whole a IVBr series, both of these adjacent IVBr series being 4(3+4). The outer IIIBr series following the I[Br 2 series bears a IVBr 2 series on the outermost side of the postradial series as a whole. In the other example of arm division both IIBr series are 4(3+-4), and the I11Br series are 4(3-+4) externally and 2 internally. There are two IVBr series, one next to the midradial line of the postradial series as a whole on either side; one of these is 4(3-+-4) and the other is 2. The cirri of these specimens are up to 42 mm. in length and consist of 28-38 segments of which the distal are mostly quite without dorsal spines. Pp and P, are almost equal. Pp in the largest specimen has 22-24 segments and in the smaller 18-20. The segments are thick discoidal, broader than long, being longer than broad only in regenerating pinnules. These pinnules measure 19-22 mm. in length. The lower pinnules to and including P; and P, have entirely smooth overlapping distal ends. Reichensperger noted that I had said that in smaller animals with Pp only 13 mm. long the segments of that pinnule have produced and spinous distal ends. My re- mark referred to specimens from Singapore (H. martensi), which at the time (1909) I believed to be conspecific with Hartlaub’s specimens from Amboina. In Reichensperger’s specimens the diameter of the disk is about 25 mm. The color in alcohol is dark red-brown. One of the specimens from the Great Barrier Reef, as described by Dr. H. L. Clark, has 29 arms and the cirri about XXX, 27-29. The other has 36 arms and the cirri are XXXII, 29-33. In both specimens the cirri have the dorsal spines confined to the outermost segments; seldom more than six segments show the spines clearly, and often only one. The specimen from Biliton recorded by Professor Koehler is not completely developed. There are 36 arms, which do not exceed 80 mm. in length, and IVBr series are not present on all the rays. One of the IIBr series and two of the IIIBr series are 2; all the other [[Br, [1] Br, and IVBr series being 4(3+-4). The specimen from Akyab on the Burmese coast was described by Hartlaub as a new species under the name Antedon kraepelini. The centrodorsal is thick, strongly concave in the middle of the dorsal surface, with the sides swollen. There were about XXX cirri of which only two stumps remain. The radials are only slightly visible; they are separated laterally. The IBr, are entirely free laterally. The IBr, (axillaries) are rather short, and are pentagonal. The IIBr series are 4(3+4). The inner II Br series are always 2, and the outer are often 4(3-+4). The articulations between the elements of the division series are smooth. The two ossicles immediately following each axillary are partially united interiorly. There are 33 rough and rather short arms from 40 to 50 mm. long which taper rapidly. From about the ninth on- ward the brachials are very short and wedge-shaped with rather strongly produced distal ends. In the distal half of the arms they again become more discoidal. The A MONOGRAPH OF THE EXISTING CRINOIDS 199 proximal brachials are discoidal and have smooth articulations. The first brachial syzygy is between brachials 3+4, the second is at about brachials 24-++-25, and the distal intersyzygial interval is usually 9 muscular articulations. Pp and P» are thick and massive. Pp is about 13 mm. long and is composed of about 18 approximately squarish segments which gradually become somewhat more slender toward the end of the pinnule. Pp is of the same form and sometimes also of almost the same length. P, on the outer arms has a similar form at the base but it is much shorter and after the basal segments quickly becomes slender. On the inner arms P, is markedly weaker. P, is extraordinarily small. The following pinnule pairs, especially from the third onward, are very small and are of about the same length. From the fifteenth brachial onward they very gradually become larger though they do not reach more than 5 mm. in length. The sacculi on the pinnules are small and scattered. The disk is lacking. Hartlaub said that this specimen is unfortunately much broken, the cirri and most of the arms having been lost. He said that the species represented is remarkable for the thickness of the proximal pinnules, which is the more noticeable as the more distal pinnules are unusually small and slender. He regarded the shortness and rapid taper of the arms as noteworthy, this being in contrast to the central portion of the animal which has a stout centrodorsal and relatively strong first division series. He said that the new species seems to stand nearest to Antedon (Himerometra) martensi from which it differs, however, in the form of the centrodorsal, the arms, and the lower pinnules. The last are very massive in both species, but more so in martensi in which there is a special enlargement of the basal segments. Besides, in martensi the thick pinnules have an uneven surface due to the abrupt decrease in the size of their com- ponent segments, which is not the case in kraepelini. Also in martensi the strong production of the distal border of the brachials begins at the third brachial. When at Hamburg in 1910 I examined this specimen. It appears to be a small specimen of robustipinna (that is, the form which I described as robustipinna in 1908) with which it agrees in the absence of a thickening or eversion of the distal ends of the segments of the enlarged proximal pinnules. The lack of the pinnule tips and of the cirri prevent a wholly satisfactory determination. Under the name of Himerometra kraepelini Reichensperger recorded two well- preserved specimens from Ceylon. In the larger individual the centrodorsal is thick with a large concave dorsal pole. The cirri are arranged in two rows. ‘The cirri are XXX, 25-30, from 27.5 to 35 mm. long. The first four or five seryments are markedly broader than long, and that following is about as long as broad; the next five or six are somewhat longer than broad, but toward the end of the cirri the segments become again markedly broader than long. There are no dorsal spines, but the last three or four segments bear a very faint dorsal swelling. The antepenultimate segment has a small blunt spine. The terminal claw is short and curved. There are 46 arms 95 mm. long. There are up to four divisions. In their structure the arms agree com- pletely with Hartlaub’s description. The basal portion of the arms is smooth. From the last division onward the arms taper rapidly and are very rough with entirely discoidal brachials. Pp is the thickest and longest pinnule, stiff, the easily broken point turned somewhat outward. It is 14-15 mm. long and is composed of 14-18 segments, all of which are more or less squarish. The distal ends never overlap. 200 BULLETIN 82, UNITED STATES NATIONAL MUSEUM P, is similar in form but slightly shorter, with about 15 segments. P3, in common with P,, after the third or fourth segment begins to taper rapidly and becomes more flagellate, but it is rather rigid and is almost as long as P2; it is composed of about 22 segments. P, is small and slender, only about one-third as long as P;. The other characters agree entirely with those given by Hartlaub. The close placing of the short distal pinnules is especially striking. The color, as preserved, is ventrally dark red-brown, P,; to P, being lighter. The cirri and the proximal portion of the arms is light brown, the arms becoming almost black distally. In the second specimen from Ceylon the centrodorsal is thick with the dorsal pole rather large and concave. The cirri are arranged in two rows. ‘The cirri are XXXII, 28-30, up to 35 mm. in length. As in the other individual the segments in the middle portion of the cirri are somewhat longer than broad—trelatively slightly longer in this specimen than in the other. The 38 arms are about 80 mm. long. Their structure is the same as in the other. A number of arms are regenerating at the syzygy between brachials 3+4. In three cases an axillary is regenerating, and in five cases a new arm without an axillary. Pp and P, are relatively a little more slender than in the other specimen and not quite so straight and stiff; they are com- posed of 18 and 15 segments, respectively. The other characters are as given in the other individual. Unfortunately the pinnules are in part defective. The color is dark brown, almost blackish. Localities.—Albatross station 5165; Tawi Tawi group, Sulu (Jolo) Archipelago; Observation Island bearing N. 70° W., 6.4 miles distant (lat. 4° 58’ 20’’ N., long. 119° 50’ 30’ E.); 16 meters; coral; February 24, 1908 [A. H. Clark, 1908, 1912, 1913; Reichensperger, 1913] (1, U.S.N.M., 25439). Macclesfield Bank; 57-66 meters; H. M. S. Penguin [Bell, 1894; A. H. Clark, 1908, 1912, 1913] (1, B.M.). Albatross station 5644; Buton Strait, off southeastern Celebes; Makasser Island (E.) bearing N. 4° E., 1.3 miles distant (lat. 5° 27’ 24’’ S., long. 122° 38’ 00” E.); 40 meters; December 16, 1909 (1, U.S.N.M., 35971). Moluccas; H. C. Macklot [P. H. Carpenter, 1881; A. H. Clark, 1911, 1912, 1913] (1, L.M.). Amboina; Dr. J. Brock [Hartlaub, 1890, 1891; A. H. Clark, 1912, 1913). Kei Islands; Nuhu-Tawan, northern coast of Little Kei; Dr. H. Merton, June 16, 1908 [Reichensperger, 1913; A. H. Clark, 1918]. Off Ellison Reef, outer Great Barrier Reef, Queensland; 9-27 meters; Surgeon Lt. Comdr. W. E. J. Paradice, H. M. S. Geranium, August 1924 [H. L. Clark, 1926]. Biliton, Sunda Islands; M. Korotneyv, 1885 [Koehler, 1895; A. H. Clark, 1913). Akyab, Burma [Hartlaub, 1890, 1891; A. H. Clark, 1911, 1912, 1913] (1, H.M.). Ceylon; Doctor Sarasin [Reichensperger, 1913; A. H. Clark, 1918]. Geographical range.—From the Philippine Islands and the Macclesfield Bank southward to the Moluccas, the Great Barrier Reef, and the Kei Islands and westward to Ceylon. Bathymetrical range.—Littoral and down to 57 (?66) meters. Remarks.—Although all the specimens listed appear to be conspecific, there is considerable variation in the enlarged proximal pinnules. The segments of these pinnules are smooth and cylindrical with unmodified distal ends in the type of Car- A MONOGRAPH OF THE EXISTING CRINOIDS 201 penter’s Actinometra robustipinna from the Moluccas, in the type of my Himerometra robustipinna from Albatross station 5165, in the type of Hartlaub’s Antedon kraepelini from Akyab, Burma, and in Reichensperger’s specimens from Ceylon. In the two specimens of Bell’s Antedon inopinata from the Macclesfield Bank the distal ends of the segments of the enlarged proximal pinnules are very slightly swollen, while in Hartlaub’s three specimens of Antedon crassipinna from Amboina and in Reichen- sperger’s specimens from Little Kei the distal ends of the segments are thickened and overlap the bases of the succeeding segments. In all cases Pp has about 20 segments (17-20), except that in one of the specimens from Little Kei described by Reichensperger it has 22-24 segments. This specimen would seem to show an approach to H. magnipinna. History.—The history of this species is curiously involved. It was first described from a much mutilated specimen from the Moluccas as a species of Comasteridae, Actinometra robustipinna, by Dr. P. H. Carpenter in 1881. In the Challenger report on the comatulids published in 1888 Carpenter inserted Actinometra robustipinna in the key to the species of the Parvicirra group of Actino- metra where it was given as having [IIBr 4(3+4) series, but no IVBr series, 30 or more cirri of (by inference) 15-20 segments, and the pinnules on the division series very large and stout. In discussing Comanthus trichoptera Carpenter said that robustipinna resembles it in the presence of a large number of cirri, but is readily distinguished by the great size of its first three pinnules. In 1890 Dr. Clemens Hartlaub described Antedon kraepelini from Akyab, Burma, and Antedon crassipinna from Amboina, including with the latter as conspecific a speci- men in the Hamburg Museum from Cochinchina. He redescribed these two species in greater detail and figured them in the following year. In the key to the species which he discussed in his memoir he placed Antedon kraepelini, paired with A. martensi, under the heading IJ]Br 2, and Pp very stout. But he mentioned in the key that in kraepelini the outer IJIBr series are often 4(3+4). He placed Antedon crassipinna under the heading IIIBr series 2 and 4(3-++4), together with Antedon bipartipinna (=acuticirra). His description of Antedon kraepelini immediately follows that of A. martensi, but the description of Antedon crassipinna is many pages farther on, separated from the description of A. kraepelini by the descriptions of Antedon brockii, A. affinis, and A. nematodon and a discussion of A. ludovici. In his description of Antedon kraepelini he said that the inner IIIBr series are always 2, but the outer are often 4(3-+4). He seems to have attached no importance to this in A. kraepelini, but in his account of A. crassipinna he discussed at considerable length the curious arrangement of the II1Br series which are 2 internally and 4(3-+-4) externally. In his remarks under Antedon kraepelini Hartlaub said that the short and rapidly tapering arms are in strong contrast to the central portion of the animal, which has a large centrodorsal and relatively stout [Br series. This curious form while giving an individual comatulid a very characteristic appearance, seems to be nothing more than the result of somewhat abnormal development. Individuals of this nature are not infrequent in Heterometra reynaudi at Ceylon. In 1894 Prof. F. Jeffrey Bell described Antedon inopinata from the Macclesfield Bank. This new species he placed in Carpenter’s Granulifera group, which included 208244—40—14 202 BULLETIN 82, UNITED STATES NATIONAL MUSEUM species of Thalassometridae and Charitometridae with the ITBr series 4(3 +4) showing not the slightest resemblance to any species of Himerometra, which would fall in Car- penter’s Savignyi group. In 1895 Prof. René Koehler recorded and gave notes upon a specimen, identified as Antedon crassispina, from the Sunda Islands. In my revision of the genus Antedon published in 1907, I placed crassipinna and kraepelini under the new genus Himerometra, of which the former was made the genotype. I was unable to place Bell’s inopinata on the basis of Bell’s description. In a revision of the crinoid genus Comatula published on February 29, 1908, I placed Carpenter’s Actinometra robustipinna in the genus Comaster. In a paper published on August 25, 1908, I said of Antedon inopinata that it was described in Dr. P. H. Carpenter’s Granulifera group, but it evidently belongs to the Savignii group; I remarked further that the species is not recognizable from the description. In another revision of the species of Comasteridae published on October 30, 1908, Actinometra robustipinna was placed in the genus Phanogenia. On December 23, 1908, I briefly diagnosed a new species from Albatross station 5165 in the Philippines which I called Himerometra robustipinna. Having in 1910 examined at Leyden the type specimen of Carpenter’s Actinometra robustipinna, I said in 1911, under the heading Himerometra sp., that this specimen proves to be an example of a typical species of Himerometra, and appears to represent the species, common at Singapore, to which I had referred (in 1909) as Himerometra crassipinna (in reality H. martensi), but that the identification cannot be considered certain because of the absence of the cirri and of most of the pinnule tips. In 1912, having examined the type specimen of Hartlaub’s Antedon kraepelini at Hamburg in 1910, I said, under the heading Himerometra sp., that this appears to be a small specimen of Himerometra robustipinna (of A. H. Clark, 1908) with which it agrees in the absence of a thickening or eversion of the distal edges of the segments of the enlarged proximal pinnules, but the lack of pinnule tips and of the cirri prevent accurate determination. In my monograph on the crinoids of the Indian Ocean published in 1912, having found that Carpenter’s Actinometra robustipinna is in reality a species of Himerometra, I renamed the species described by myself in 1908 as Himerometra robustipinna, ealling it Himerometra pulcher. I gave Himerometra inopinata as a good species and included a description of it drawn up from the type specimen, which I had examined in the British Museum in 1910. Actinometra robustipinna of Carpenter I placed as a doubtful synonym under H. crassipinna, and I also discussed it under the heading Himerometra, sp. Himerometra kraepelini was given as a valid species without comment. At this time the genus Himerometra was in a chaotic condition, so a detailed revi- sion of the included forms was undertaken, which was published on November 25, 1913. In this revision Himerometra robustipinna (P. H. Carpenter) was considered as including Actinometra robustipinna P. H. Carpenter, 1881; Antedon kraepelini Hartlaub, 1890; Antedon crassipinna Hartlaub, 1890 (except the specimen from Cochinchina, which is referred to H. magnipinna); Antedon inopinata Bell, 1894; A MONOGRAPH OF THE EXISTING CRINOIDS 203 Himeromeira robustipinna A. H. Clark, 1908; and Himerometra pulcher A. H. Clark, 1912 (new name for the preceding). In my notes on the recent crinoids in the British Museum published in 1913, under the heading Himerometra robustipinna (P. H. Carpenter) I said of the type of Bell’s Antedon inopinata that the centrodorsal resembles that of the other species in the genus Himerometra. Dr. August Reichensperger in 1913 described two specimens from Ceylon, which he referred to Himerometra kraepelini. He remarked that from the fragmentary description of Himerometra robustipinna published by me in 1908 it is to be concluded that the latter is unquestionably nothing else than kraepelini, of which form the lower pinnules are exceedingly characteristic. The name pulcher, suggested by me in 1912 to replace robustipinna of 1908, he said must also be placed in the syno- nymy of kraepelini. In a key to the species of Himerometra published in the report on the unstalked erinoids collected by the Siboga in 1918 robustipinna was said, in a footnote, to include the H. kraepelini and H. crassipinna of Reichensperger. In 1926 Dr. Hubert Lyman Clark recorded two specimens from the Great Barrier Reef, Queensland, thus considerably extending the known range of the species. HIMEROMETRA MARTENSI (Hartlaub) PLATE 17, Fiaures 64, 65; Puate 18, Ficures 66, 67; Puate 19, Fiaures 72-75 [See also vol. 1, pt. 1, figs. 10, 11 (nomenclature of centrodorsal and radial pentagon), p. 65; fig. 254 (centrodorsal), p. 253; fig. 338 (cirrus), p. 285; fig. 469 (centrodorsal), p. 359; pt. 2, figs. 41, 42 (radial pentagon), p. 26; fig. 123 (arm base), p. 79; fig. 326 (proximal pinnules), p. 227; fig. 347 (distal pinnules), p. 229; figs. 442, 443 (pinnule tip), p. 261; fig. 713 (disk), p. 346.] Antedon martensi HartLuaus, Nachr. Ges. Gottingen, May 1890, p. 182 (description; Singapore); Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 18 (in key), p. 21 (detailed description and comparisons; Singapore), pl. 1, figs. 3, 6—A. H. Cuarx, Proc. U. S. Nat. Mus., vol. 43, 1912, p. 383 (identified as Himerometra crassipinna) ; Crinoids of the Indian Ocean, 1912, p. 36 (iden- tified as Heterometra martensi). Himerometra martensi A. H. Cuarx, Smithsonian Mise. Coll., vol. 50, 1907, p. 856 (listed); Proc. Biol. Soc. Washington, vol. 22, 1908, p. 7 (listed); Proc. U. S. Nat. Mus., vol. 46, 1913, p. 285 (in key), p. 286 (synonymy; range); Smithsonian Misc. Coll., vol. 61, No. 15, 1913, p. 25 (Pulau Obin, Singapore); Unstalked crinoids of the Siboga-Exped., 1918, p. 74 (in key; range) ; Smith- sonian Mise. Coll., vol. 72, No. 7, 1921, pl. 1, figs. 2, 3 (radials and centrodorsal).—GisLin, Zool. Bidrag Uppsala, vol. 9, 1924, p. 88 (articulations), p. 92 (pinnule articulations).—A. H. Cuark, Proc. Biol. Soc. Washington, vol. 47, p. 9, 19384 (Singapore) —Gistin, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 27, 28, 30, 40. Himerometra crassipinna A. H. Cuark, Smithsonian Mise. Coll., vol. 50, 1907, p. 356 (in part; listed) ; Bull. Mus. Comp. Zool., vol. 51, 1908, p. 243 (relation to H. persica); Proc. U.S. Nat. Mus., vol. 34, 1908, p. 319 (specimen erroneously labeled ‘“Japan’’); Smithsonian Mise. Coll., vol. 52, 1908, p. 212 (compared with H. bartschi), p. 213 (compared with H. robustipinna), p. 214 (compared with H. magnipinna); Proc. Biol. Soc. Washington, vol. 22, 1909, p. 7 (in part; listed) ; Vid. Medd. Nat. Foren. Kgbenhavn, 1909, p. 119 (ecology), p. 155 (Singapore; notes), p. 193 (collected at Singapore by Svend Gad); Proc. U.S. Nat. Mus., vol. 43, 1912, p.383 (Antedon martensi is this species [error]), p. 394 (British North Borneo; resembles others from Singapore; Singapore; type of martensi agrees with others at hand from Singapore); Crinoids of the Indian Ocean, 1912, p. 116 (records from Singapore and Pulau Ubin). Heterometra martensi A. H. Cuarx, Crinoids of the Indian Ocean, 1912, pp. 36, 127 (Hartlaub’s specimen from Singapore only). 204 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Craspedometra martensi Gisu6N, Kungl. Fysiogr. Sillsk. Handl., new ser., vol. 45, No. 11, 1934, p. 22. Diagnostic features.—The enlarged proximal pinnules are very stout and taper rapidly to the tip; in fully developed individuals they are composed of about 20 seg- ments the distal ends of which project abruptly beyond, but do not overlap, the bases of the segments succeeding and are armed with very numerous fine spines, which in large specimens may be confined to the outer segments; the distal ends of the proxi- mal brachials are very strongly produced and everted so that the arm bases are very rough, There are 28-43 (usually 30-34) arms, which are 70-124 (usually 90-110) mm. long. Description.—The centrodorsal is thick discoidal with the bare polar area rather small and concave, bearing 35-40 cirrus sockets in about three irregular and closely set rows about its margin, the distalmost cirrus sockets encroaching somewhat on the dorsal polar area. The cirri are XV-XXXV, 28-40 (usually about 30), 23-37 (usually about 30) mm. long, large and stout. ‘The segments are all somewhat broader than long and those in the distal half of the cirri bear small dorsal spines. The radials are only very slightly, if at all, visible. The [Br, are short and are partially united laterally. The IBr, (axillamies) are short, pentagonal, dorsally rounded, and widely free laterally. The IBr series are 4(3-++4), occasionally 2. The III Br series are commonly 2 interiorly and 4(3+-4) exteriorly, but this is by no means an invariable rule. When less than the full number of IIIBr series is developed on u postradial series it is the outer ones that are lacking; in small specimens only the two inner may be present. The elements of the division series have a peculiar swollen appearance; while this feature occurs in all the species of Himerometra it is especially noticeable in this form. The arms are 28-43 (usually 30-40) in number and are 70-124 (usually 90-110) mm. in length. The brachials are very uniform and short, discoidal, the first five or six being somewhat longer in proportion than the others, which are very short. All the brachials have overlapping and slightly everted distal edges, which give the arms a very rough appearance. Syzygies occur between brachials 3+4, again from between brachials 23-+-24 to between brachials 31-+-32, and distally at intervals of 13 to 16 muscular articulations, The pinnules on the division series and P, are enormously thickened, slightly recurved, and hornlike, with about 20 short cylindrical segments which are about as long as broad, ‘The distal ends of the segments are markedly broader than the bases of the segments following so that the profile of the pinnules shows a series of abrupt steps; they are usually abundantly armed with fine spines, though in large individuals they may be simply roughened except in the distal portion of the pinnules. P, is about 20 mm. long, and the pinnules succeeding decrease in length to P,, which is about 16 mm. long. The pinnules of the inner arms are somewhat smaller than those of the outer. The following pinnules are slender, decreasing in length to about Ps, then gradually increasing again, reaching a length of 10 mm. distally. The color in alcohol is flesh color; no markings are recorded on any specimen. Notes.—The type specimen was described by Hartlaub as follows: The centro- dorsal is a thick disk with a flat bare dorsal pole, bearing cirri in two rows. The cirri A MONOGRAPH OF THE EXISTING CRINOIDS 205 are about XX, about 25, about 18 mm. long, and stout. The distal border of the segments is somewhat produced. The distal segments bear small spines, and the antepenultimate carries a stout opposing spine. The radials are partially visible later- ally. The IBr, are entirely free laterally. The IBr, (axillaries) are pentagonal. The I[Br series are 4(3+-4), and the IIIBr series are 2. The two ossicles following the IBr axillaries are partially united interiorly. The articulations are smooth. The IBr axillaries and immediately following ossicles have no lateral processes. The arms are apparently not more than 30 in number and are all broken. The brachials are short and overlapping with, except for the first two, strongly produced distal ends. The first syzygy is between brachials 3+4, and the next is about the twenty-fourth brachial. P, is very stout and stiff, about 9 mm. in length, with 12-15 segments, of which the three basal are very large and those following decrease abruptly in size. P, is smaller and less stiff, although its basal segments are also very large. P,is much smaller, scarcely 4 mm. long, and beyond the basal segments rapidly becomes slender. The disk is 10 mm. in diameter, and is deeply incised. Sacculi are rather sparsely distributed on the pinnules. The color of the skeleton is gray-brown. Hartlaub remarked that this specimen is unfortunately much broken. As shown in his figure the arms are all broken away at the articulation between the first and second brachials except for a single stump including 15 bracbials. But, as Hartlaub says, the characters by which the species represented differs from other species are well shown. Especially characteristic are the massive lower pinnules, quite unique in form, and the strongly produced distal ends of the fourth and following brachials. He said that this species is nearest Antedon kraepelini (= Himerometra robustipinna) from Akyab with which, as also with others, it agrees in the position of the second brachial syzyey. As it is shown on the plate Hartlaub’s figure of P, (pl. 1, fig. 6) is deceptive, for the shading gives the impression that the pinnule and the portion of the arm shown are sharply flattened laterally more or less like P; and the adjacent portion of the arm base in a species of Thalassometra. But if the figure is viewed upside down it gives the proper impression of convexity. I have examined the type specimen of Hartlaub’s Antedon martensi. It is a rather small and badly broken specimen, but it agrees in every particular with the specimens in the Copenhagen Museum, with which it was compared directly. Some of the specimens collected by Svend Gad at Singapore present the follow- ing characters: June 15, 1905; one has 31 arms 70 mm. long, and the other has 36 arms 70 mm. long. June 16, 1907; the five specimens have 29-43 arms 65-80 mm. long. November 27, 1907; three specimens with 27 arms 70 mm. long, 37 arms 80 mm. long, and 38 arms 70 mm. long. These specimens are typical of the whole series. The centrodorsal has a broad, slightly concave, dorsal pole. The longest cirri have about 30 segments, sometimes rather less, but more commonly rather more. The IIBr series are always present, and are almost invariably 4(3+4). The exterior IlIBr series are 4(3-++-4) and the interior are 2; when a deficiency in the IIIBr series occurs the exterior IIIBr series are usually lacking, so that there is a tendency toward a IIIBr arrangement of 1, 2, 2, 1. The IVBr series, when present, are 2, and are 206 BULLETIN 82, UNITED STATES NATIONAL MUSEUM developed externally in reference to the I1Br series. The overlapping of the distal edges of the brachials begins at the second and is strongly developed. Py is about 13 mm. long, very stout, the component segments having prominently spinous distal ends. January 24, 1910; of the nine specimens the largest has 43 arms 100 mm. long. A very small specimen has 11 arms 35 mm. long. On the arms arising directly from a IBr axillary P, is 6 mm. long, greatly enlarged as in fully grown individuals, with 12 or 13 segments. PP, is very much smaller and more slender, and P, is smaller and slenderer still. P, and the following pinnules are very small and short. These specimens collected by Mr. Gad are all pale flesh colored or light purplish, with the perisome brown, as are all the others I have seen from the same locality. They are practically uniform in size and are all much smaller than those in the U. S. National Museum, the Museum of Comparative Zoology, the Yale University Museum, and the collection of Prof. Robert T. Jackson of Harvard University, all of which were obtained from Ward’s Natural Science Establishment. But except for the smaller size and the slightly inferior number of cirrus segments (probably cor- related with the lesser size) I am unable to detect any difference. The description of this species was drawn up from a study of 25 specimens from Singapore in the collections of the U. S. National Museum (2), the Boston Society of Natural History (19), the Museum of Comparative Zoology (1), the Yale University Museum (2), and Prof. Robert Tracy Jackson (1, now in the collection of the Museum of Comparative Zoology), all of which were secured from Ward’s Natural Science Establishment. These show the following characters: One of the specimens in the U. S. National Museum has 40 arms 100 mm. long, and XXXY stout cirri 30 mm. long. The IIBr series are all 4(3+4). Of the IIBr series 8 are 4(3-++-4) and 10 are 2, the latter most often developed interiorly. The post-radial series are rounded and well separated. P, resembles P, in character but is only about half as long. The other specimen in the U. S. National Museum has 35 arms, mostly arranged in 1, 2, 2, 1 order; the arms are 110 mm. long. The cirri are 30 mm. long and are composed of 30 segments. The inner IIIBr series are usually 2; IIIBr series are usually not present on the outer side of the I1Br series, but if present they are 4(3+-4). The specimen in the Museum of Comparative Zoology (dry) has 31 arms 110 mm. long. The cirri are XXX, 30, 30 mm. long. The IIBr series are all 4(3-+-4). The inner IIIBr series are all present, and all are 2; there is a single exterior III]Br series which is 4(3+-4). Professor Jackson’s specimen (now in the Museum of Comparative Zoology) has about 40 arms 110 mm. long. The cirri are XX XIII, 37-40, 37 mm. long. On the anterior postradial series both the IIBr series are 2, and one of the lateral post- radial series has one IIBr series 2 and the other 4(3+-4). All the other IIBr series are 4(3-++4). On the postradial series with both IJBr series 2 all four of the IJIBr series are 4(3+4). On the postradial series with one IIBr series 2 and the other 4(3+-4) the IIIBr series are 4(3+4), 2, 2,0. On the remaining postradial series the IIIBr series are as follows: 0, 2, 2, 0; 4(3+-4), 2, 2, 0; 0,2, 2,4(3+4). TheIVBr series following IIIBr 2 series are 4(3+-4), and those following III[Br 4(3-+-4) series are 2, The postradial series are well rounded and are widely free laterally. A MONOGRAPH OF THE EXISTING CRINOIDS 207 Fifteen of the specimens in the Boston Society of Natural History show the follow- ing features: 1, All the IJBr series are 4(3+-4). The three remaining IIIBr series are 2. 2. There are 34 arms 97 mm. long. The cirri have 27-30 segments and are 23 mm. long. The IIBr series are all 4(3+4). On the various postradial series the IIIBr series are as follows: 0, 2, 4(3-+4), 4(3+4); 0, 2, 2, 4(3+-4); 0, 2, 2, 0; 4(8-+4) 2, 2, 4(3+4); 0, 2, 2, 0. 3. There are 31 arms 97 mm. long. The cirri are XV, 32,37 mm. long. All the inner IIIBr series are developed, and all are 2; there is a single exterior IIIBr 4(3+-4) series. 4. The cirri have 30 segments. All the IIBr series are 4(3+4). The five remaining IIIBr series are 2. There is a single I[VBr 2 series. 5. All the IIBr series are 4(3+4). On three postradial series the IIIBr series are 0, 2, 2, 0, and on the other two they are 0, 2, 2, 4(3+4). 6. The arms are 122 mm. long, and the cirri have 30 segments. All the IIBr series are 4(3-++4). On three postradial series the IIIBr series are 0, 2, 2, 0, and on the other two they are 4(3-++4), 2, 2, 0 and 0, 4(8+-4), 4(3+-4), 4(3+-4). 7. The cirri have 30-33 segments. The IIBr series are all 4 (3+4). On three of the postradial series the IIIBr series are 0, 2, 2, 0, and on another they are 4 (3+-4), 2,2,0. There is a single IVBr 4 (3+4) series present. 8. There are 28-30 cirrus segments. The IIBr series on the left anterior and left posterior postradial series are 2, all the others being 4 (+4). On the left anterior postradial series there are no IIIBr series; on the left posterior postradial series the IIIBr series are 4 (3-+4), 0, 0, 0; on the three remaining postradial series the IJ] Br series are 4 (3-4), 0, 2, 0; 0, 4 (83+4), 2, 0; and 2, 0, 0, 2. 9. There are 31 cirrus segments. All the IIBr series are 4 (3+4). On the several postradial series the IIIBr series are: 4 (83+-4), 2, 4 (3+4), 0; 0, 2, 2, 4 (3+-4); 4 (3+4), 2, 2, 0; 0, 2, 2,0; and 4 (344), 2, 2,4 (8+4). There are two IVBr 4 (3+4) series. 10. In a small broken specimen all the II Br series are 4 (3+4). All the IIJBr series, which are present on the remaining postradial series only internally, are 2. The IVBr series are 4 (3-4). 11. The IIBr series are all 4 (8-+4). There is only one complete postradial series on which the IIIBr series are 4 (3+-4), 2, 2, 4 (+4). Two other postradial series have IIIBr 2 series remaining. 12. There are 28-30 cirrus segments. All the IIBr series are 4 (8+4). On two postradial series the IIIBr series are 0, 2, 2, 0, and on two others they are 0, 2, 2, 4 (3+4); on the fifth they are 0, 2, 2, 4 (3+4) with, on the second, an inner IVBr 4 (3-+4) series. 13. There are 30 cirrus segments, of which the outer bear dorsal spines which are rather more prominent than usual. One of the IIBr series is 2, the other nine being 4 (3+4). On two postradial series the J1IBr series are 0, 2, 2, 0; on the other three they are 0, 0, 0, 0; 0, 0, 2, 0; and 2, 2, 2, 0. 14. The arms are 124 mm. long. The IIBr series are all 4 (3+4). On two postradial series the IJIBr series are all 4 (3+4); on two more they are 0, 2, 2, 0; and on the fifth they are 4 (3+4), 2, 2, 0. 208 BULLETIN 82, UNITED STATES NATIONAL MUSEUM 15. There are 30 arms, all the inner I11Br series being present. All the IIBr series and also all the IJ] Br series are 4 (83+4). The specimen from Pulau Obin, Singapore, in the British Museum resembles those from Singapore in the U. S. National and Copenhagen Museums. ‘The speci- men from British North Borneo resembles those from Singapore, although the proxi- mal cirrus segments are a trifle shorter. Localities—Singapore, Straits Settlements; Prof. Ed. von Martens [Hartlaub, 1890, 1891; A. H. Clark, 1912] (1, Berl. M., 5373). Singapore; Svend Gad [A. H. Clark, 1909] (27, U. S. N. M., 36186, 36176, E. 1080; C. M.). Singapore [A. H. Clark, 1909] (18, B. S.). Singapore [A. H. Clark, 1909] (1, R. T. J. [mow M. C, Z.]). Singapore; from Ward’s Natural Science Establishment, through Dr. F. A. Lucas [A. H. Clark, 1909] (2, U. S. N. M., 35968). Singapore [A. H. Clark, 1934] (1, Raffles Mus.). Singapore (2, M. C. Z., 59, 289). Singapore (1, Y. M.). Pulau Obin, Singapore [A. H. Clark, 1913] (1, B. M.). British North Borneo; Herr Pagel [A. H. Clark, 1912] (1, Berl. M., 4744). Erroneous locality.—Japan [A. H. Clark, 1908] (1, B. S.). Geographical range.—Only known from Singapore and British North Borneo. Bathymetrical range.—Littoral; all the known specimens were taken in’ shore collecting. History—Dr. Clemens Hartlaub described this species, as Antedon martensi, in 1890 from a small and much mutilated individual which had been collected by Prof. Ed. von Martens at Singapore. He redescribed and figured it in 1891. It was referred to the new genus Himerometra by me when that genus was created in 1907. In 1908, under the name Himerometra crassipinna, I mentioned a specimen that I had seen labeled, probably erroneously, ‘‘Japan.” In 1909 in a revision of the family Himerometridae, martensi was retained in FHimerometra. Later in 1909 in my account of the comatulids in the Copenhagen Museum, I identified as martensi, placing the species in the genus Heterometra, a specimen from Singapore which appears really to represent H. bartschi. At the same time I recorded, as Himerometra crassipinna, and gave notes upon 16 specimens from Singapore, which are in reality H. martensi. In discussing these I mentioned other larger specimens from Singapore that I had examined in the U. S. National Museum, in the Boston Society of Natural History, and in the collection of Prof. Robert T. Jackson at Har- vard University. In 1912, in a paper on the crinoids in the Berlin Museum, under the heading Himerometra crassipinna, I gave notes upon Hartlaub’s type specimen and recorded and gave notes on another from British North Borneo. In my monograph of the crinoids of the Indian Ocean published in 1912, following my report on the crinoids of the Copenhagen Museum, I placed martensi in the genus Heterometra and gave the locality as Singapore. I included what was really martensi A MONOGRAPH OF THE EXISTING CRINOIDS 209 under Himerometra crassipinna, which is the species found at two of the localities given.—Singapore and Pulau Ubin. In 1913, in a revision of the genus Himerometra, I gave in a key to the included species the differential characters of H. martensi. A complete synonymy of the species was given, and the range was correctly noted as Singapore, Pulau Ubin, and British North Borneo. In the Siboga report published in 1918 martensi was included in the key to the species of Himerometra, and the correct range was given. In 1924 Dr. Torsten Gislén discussed the articulations of this species on the basis of material from Singapore. In 1934 I said that this species has long been known to be common at Singapore. HIMEROMETRA BARTSCHI A. H. Clark PuaTE 14, Fiaurss 52, 53; Pharr 16, Figures 57-59 Himerometra bartschi A. H. Cuarx, Smithsonian Mise. Coll., vol. 52, 1908, p. 212 (description; Albatross stations 5146; also 5147), p. 214 (comparison with H. magnipinna) ; Proc. Biol. Soc. Washington, vol. 22, 1909, p. 7 (listed); Crinoids of the Indian Ocean, 1912, p. 114 (synonymy; locality); Proc. U.S. Nat. Mus., vol. 46, 1913, p. 285 (in key), p. 288 (synonymy; range); Unstalked crinoids of the Siboga-Exped., 1918, p. 73 (in key; range), p. 75 (references; notes; station 96), p. 272 (listed).—GisLin, Kungl. Fysiogr. Saillsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 41, 48.—A. H. Cuarx, Temminckia, vol. 1, 1936, p. 302 (Amboina, 0-2 meters); John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1936, p. 103. Himerometra persica A. H. Cuark, Smithsonian Mise. Coll., vol. 52, 1908, p. 214 (Albatross station 5163; also Philippine Islands with no definite data); Proc. Biol. Soc. Washington, vol. 22, 1909, p. 7 (in part; listed); Crinoids of the Indian Ocean, 1912, p. 16 (in part; records from Philip- pine Islands). Heterometra martensi A. H. CuarK, Vid. Medd. Nat. Foren. Kgbenhavn, 1909, pp. 164, 165 (Singa- pore), p. 193 (collected at Singapore by Svend Gad); Crinoids of the Indian Ocean, 1912, p. 127 (Clark’s record from Singapore). Diagnostic features—The proximal pinnules are slender, flagellate distally, and very long, composed of 37-40 perfectly smooth segments most or all of which are longer than broad, with a few of the earlier narrowly but prominently carinate; the earlier segments of the following pinnules are very strongly carinate. The cirri have 35-43 segments, of which those in the distal half bear prominent dorsal spines. The 37-61 arms are 120-150 mm. long. Description.—The centrodorsal is thick discoidal with a rather strongly concave polar area. The cirrus sockets are arranged in two and a partial third crowded and more or less alternating rows. The cirri are XXX, 41-43, moderately stout, rather more than one-third the length of the arms, 45 mm. long. The first five segments are about twice as broad as long and those following gradually increase in length to the ninth or tenth, which is about as long as broad; after about the sixteenth they gradually decrease in length, the terminal segments being about twice as broad as long. After the seventeenth segment small but prominent dorsal spines are developed. The opposing spine is centrally situated, rather slender, in height reaching to about one-half the width of the penultimate segment. The terminal claw is considerably longer than the penulti- mate segment, and is slender and moderately curved. 210 BULLETIN 82, UNITED STATES NATIONAL MUSEUM The distal ends of the radials are approximately even with the rim of the centro- dorsal. The IBr, are short, and are laterally united for their entire length. The IIBr series are 4 (3-+4). The II1Br series are 4 (3-4) exteriorly and 2 interiorly. The IVBr series, which are developed only on the inner side of the inner IIIBr series from each IIBr series, are 4 (3+4). Each I1Br series divides once exteriorly and twice interiorly. The division series resemble those of H. martensi. The 51 arms of the type specimen are 120 mm. in length and resemble those of H. martensi. Pp and Pp are 25 mm. long, very stout basally but tapering gradually to a slender and delicate tip, with about 40 segments most of which are about as long as broad. P; is 15-17 mm. long, proportionately slenderer than the preceding pinnules, with about 32 segments, which are slightly longer than broad, the first two slightly carinate. P, is smaller and slenderer, 13 mm. long, with 30 segments of which the first is short, the next three or four are about as long as broad, and the remainder are slightly longer than broad; the second, third, and fourth segments are somewhat carinate. P; is 8 mm. long, small, and weak, with about 20 segments, of which those in the proximal half are about as long as broad and those in the distal half are slightly longer than broad; the second-sixth are rather strongly carinate. The succeeding pinnules decrease rapidly in length, the fifth and following being 5 mm. long, then slowly increase again, reaching a length of 9 mm. distally. Notes.—In the specimen from Siboga station 96 the cirri are rather stout, 30 mm. long, with 37-38 segments, of which the longest proximal are from half again to twice as broad as long; there is very little difference between the proximal and distal seg- ments; from the nineteenth or twentieth segment onward prominent dorsal tubercles or short blunt dorsal spines are developed. The 61 arms are 120 mm. long. Four of the postradial series bear 12 arms each, and the fifth bears 13 arms. Each IJBr series bears externally a I]]Br 4(3+-4) series and internally a IIIBr 2 series, which carries two IVBr 4(3+-4) series. Py is 23 mm. long, with 37 segments, very stout basally but tapering evenly to a slender and flagellate tip. The basal segments are twice as broad as long, those following gradually increasing in length and becoming about as long as broad on the twelfth and half again as long as broad in the delicate terminal portion. The pinnule is perfectly smooth, with no traces of eversion of the distal edges of the segments. Pp is 22 mm. long, with 36 segments and is only very slightly smaller than Pp. P, is similar to the two preceding pinnules but is somewhat smaller, 17 mm. long. Pz is weak and slender, 8.5 mm. long, with 23 segments, of which the second-fifth have a prominent narrow carination the distal border of which is parallel to the longitudinal axis of the segments. FP, is 6.5 mm. in length, resembling P, but very slightly smaller. P, is 5.5 mm. long, resembling P; but very slightly smaller; the second-fourth segments are carinate. From this point the amount of carination of the earlier segments of the pinnules decreases, though it is traceable to the end of the proximal fourth of the arm. This specimen differs from the type only in the slightly smaller size and in the slightly shorter proximal cirrus segments. In the specimen from Singapore the cirri are XII 26-30, 20 mm. long, and rather slender; the segments have rather prominent ends. There are 17 arms 50 mm. long. Four IIBr series are developed of which three are 4(3-+-4) and one is 2. There are A MONOGRAPH OF THE EXISTING CRINOIDS 211 three IIIBr series, all on a single postradial series, two 4(3-+4) following the IIBr 2 series, and one 2, developed internally. Pp (or P;) is 9 mm. long with 21-22 seg- ments, very stout basally but tapering evenly and becoming flagellate distally. P, is similar but is somewhat smaller. The following pinnules are small and weak. In a specimen from the Danish Expedition to the Kei Islands station 19 the cirri are large and stout, XXV, 35-36, 35 mm. long. The longest segments are about one-third again as broad as long, or on some cirri nearly or even quite as long as broad. Dorsal processes begin to appear on from the eighth to the tenth segment, and after five or six become long and conspicuous triangular dorsal spines arising from the entire dorsal surface of the segment which are either erect or directed distally; fre- quently the tip of the spine is notched. The 47 arms are 150 mm. long. Of the 10 IIBr series present 7 are 4(3+-4), one is 2, and 2 are 4, the two outer elements not being united by syzygy. On two postradial series the IJIBr series are 4(3+4) exter- nally and 2 internally. On two other postradial series three of the IIIBr series are 4(3+4), one of the two internally developed being 2. On the fifth postradial series all the IIBr series are 4(3++4). The IVBr series are all 4(3-+-4), and all are developed on the inner portion of a postradial series, next the midradial line. Pp is 22 mm. long, with 40 segments, of which the first is about twice as broad as long, the second and third are slightly broader than long, and the remainder are about as long as broad. The pinnule is moderately stout at the base but tapers in the proximal third and in the outer half is slender and flagellate. The second-fourth segments have a low well-rounded inconspicuous keel in the middle of the dorsal side. Ppis 23 mm. long, with 40 segments, in some cases resembling Py but usually somewhat stouter, and sometimes considerably stouter. The first five to eight segments have an obscure median elevation on the dorsal side and the second-eighth have the outer edge (toward the arm tip) produced into a thin narrow keel. P, is 15-17 mm. long with 31 segments, and resembles the preceding pinnules but is slenderer; the outer border of the second-eighth segments is produced into a narrow knifelike carination. P, is 11-12 mm. long, considerably slenderer than P, and is composed of 26 segments, of which the outermost are about twice as long as broad. The second-sixth have a more conspicuous carination of the outer edge than the corresponding segments of the preceding pinnules. P; is 8 mm. long, with 19 segments, and is considerably slenderer than P;. As in P2 the second-sixth segments are very strongly carinate. The following pinnules are smaller and shorter, with the third-sixth segments strongly carinate. This carination of the earlier pinnule segments remains conspicuous until about the middle of the arm and is indicated even on the second and third segments of the terminal pinnules. Of the other specimens from the Danish Expedition to the Kei Islands station 19 one has 47 arms 140 mm. long and the cirri 40 mm. long with 35 segments, and another has 37 arms 140 mm. long. In one of the specimens from Amboina there are 47 arms 140 mm. long, and the longest cirri are 50 mm. long, with 37-43 segments. The other specimen from Amboina has 42 arms 125 mm. long. The specimen collected by the Willebrord Snellius at Amboina is a fine example with 55 arms 150 mm. long. Po and Pp have 38-40 segments. 212 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Localities —Albatross station 5146; Sulu (Jolo) Archipelago, in the vicinity of Siasi; Sulade Island (E.) bearing N., 18° W., 3.4 miles distant (lat. 5° 46’ 40’ N., long. 120° 48’ 50’’ E.); 44 meters; coral sand and shells; February 16, 1908 [A. H. Clark, 1908] (1, U.S.N.M., 25438 [type]). Albatross station 5147; Sulu Archipelago, in the vicinity of Siasi; Sulade Island (E.) bearing N. 3° E., 8.4 miles distant (lat. 5° 41’ 40’ N., long. 120° 47’ 10’’ E.); 38 meters; coral sand and shells; February 16, 1908 [A. H. Clark, 1908] (1, U.S.N.M., 35205). Albatross station 5163; Sulu Archipelago, Tawi Tawi group; Observation Island bearing N. 79° W., 6.7 miles distant (lat. 4° 59’ 10’” N., long. 119° 51’ 00’ E.); 51 meters; February 24, 1908 [A. H. Clark, 1908] (1, U.S.N.M., 35096). Siboga station 96; southeastern side of the Pearl Banks, Sulu Archipelago; 15 meters; lithothamnion bottom; June 27, 1899 [A. H. Clark, 1918] (1, Amsterdam Mus.). Albatross; Philippine Islands; 1907-’08 [A. H. Clark, 1908] (1, U.S.N.M., 35198). Singapore; Svend Gad [A. H. Clark, 1909] (1, C.M.). Amboina; pier; 0-2 meters; Willebrord Snellius, May 6, 1930 (1, L.M.). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; station 19 (4). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; Amboina; breakwater; about 1 meter; February 28, 1922 (2). Geographical range.—From the Philippine Islands and Singapore south to Amboina and the Kei Islands. Bathymetrical range.—From the shoreline down to 51 meters. Remarks.—This form is probably only a large East Indian race of H. persica from the Persian Gulf. History.—This species was originally described in 1908 from a specimen dredged by the Albatross at station 5146 in the Philippines. At the same time another specimen was recorded from station 5147, near the type locality. In the same paper H. persica was recorded from Albatross station 5163, and from the Albatross Philippine collections without further data. In 1909 I recorded and described from Singapore, under the name of Heterometra martensi, a specimen that could only have been a young example of this species. I said that P, is similar to, but smaller than, Pp or P,; and that these pinnules are flagel- late distally. I remarked further that the proximal pinnules, though for the genus unique in their proportionate lengths, are of the type common to all species of Het- erometra. Pinnules resembling those of the species of Heterometra with the first the longest are found only in H. bartschi and in H. persica. In my monograph of the crinoids of the Indian Ocean published in 1912, I gave H. persica as occurring both in the Persian Gulf and in the Philippines, but in the following year in a revision of the genus Himerometra the specimens from the Philip- pines recorded in 1908 and 1912 as H. persica were redetermined as H. bartschi. In 1918 I recorded and gave notes upon a specimen from Siboga station 96, near the type locality. HIMEROMETRA PERSICA A. H. Clark Puate 19, Figures 70, 71 (See also vol. 1, pt. 1, fig. 318 (tip of cirrus), p 273; fig. 339 (cirrus), p. 285; pt. 2, fig. 272 (arm and pipnules), p. 207; fig. 444 (pinnule tip), p. 261.] A MONOGRAPH OF THE EXISTING CRINOIDS 213 Himerometra persica A. H. Cuarx, Smithsonian Mise. Coll., vol. 50, 1907, p. 356 (listed; nomen nudum); Bull. Mus. Comp. Zool., vol. 51, 1908, p. 243 (description; Persian Gulf); Proc. Biol. Soc. Washington, vol. 22, 1909, p- 7 (listed); Amer. Nat., vol. 43, 1909, p. 256 (absent from the Red Sea; Persian Gulf to Philippines (specimens from last really H. bartschi]); Crinoids of the Indian Ocean, 1912, p. 11 (occurs both east and west of Ceylon), p. 21 (range covers that of all the other species, and extends farther west), p. 116 (synonymy; range [specimens from the Philippines are H. bartschi]); Proc. U. S. Nat. Mus., vol. 46, 1913, p. 285 (in key), p. 289 (anno- tated synonymy; range); Unstalked crinoids of the Siboga-Exped., 1918, p. 73 (in key; range) .— GisLin, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 22, 23, 46, 48, 49, 52.—A. H. Cuark, John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1936, p. 99 (range), p. 103. Diagnostic features.—The proximal pinnules are slender, flagellate distally, and very long, composed of perfectly smooth segments most or all of which are longer than broad, with a few of the earlier narrowly but prominently carinate; the earlier seg- ments of the following pinnules are very strongly carinate. The cirri have 35 segments of which those in the outer half have prominent dorsal spines. The 20-25 arms are about 150 mm. long. Description.—The centrodorsal is low hemispherical with a large bare polar area. The cirri are about XXV, 35, 27 mm. long. Most of the cirrus segments are slightly longer than broad, the distal about as long as broad; the last 16-18 bear sharp dorsal spines. The radials are just visible beyond the rim of the centrodorsal. The IBr, are trapezoidal, about three times as broad as long. The IBr, (axillaries) are pentagonal, about half again as broad as long, with a sharp distal angle. The IIBr series are 4(3+4) or 2. The division series are strongly rounded dorsally and widely free laterally. The 20-25 arms are 150 mm. in length. The first eight brachials are roughly oblong, those following becoming wedge-shaped, much broader than long, and oblong toward the end of the arms. Syzygies occur between brachials 3+-4, again in the vicinity of the seventeenth brachial, and distally at intervals of 6 to 13 (usually about 8) muscular articulations. Pp is 13 mm. long, very slender and flagellate, with 36 segments, all of which are somewhat longer than broad, but not much so; the first four segments are the broadest and are slightly carmate. P, is similar but longer, 16 mm. in length, and stouter basally; the five or six basal segments are sharply carinate, and from this point the pinnule tapers gradually to the long and delicate flagellate tip. P, resembles P, and is of the same length. The next few pinnules decrease rapidly in length, and those following increase somewhat distally, but do not become very long. The carination of the basal pinnule segments becomes less and less merked distally and is not notice- able after P,; it is at its maximum on P, and P). The color in alcohol is dull brown, the skeleton somewhat lighter. Locality.—Persian Gulf; F. W. Townsend [A. H. Clark, 1908] (2, M.C.Z., 291). Remarks.—This species is as yet known only from the two specimens described in 1908. It is very distinct from all the others in the genus with the exception of H. bartscht, which will probably prove to be simply its Malayan variety. 214 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Genus CRASPEDOMETRA A. H. Clark Antedon (part) P. H. Carpenter, Journ. Linn. Soc. (Zool.), vol. 16, 1882, p. 509, and following authors. Himerometra (part) A. H. Cuarx, Smithsonian Mise. Coll., vol. 50, 1907, p. 356. Craspedometra A. H. Cuarx, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 8 (diagnosis; genotype Antedon acuticirra P. H. Carpenter, 1882), p. 175 (referred to the Himerometrinae); Mem. Australian Mus., vol. 4, 1911, pp. 731,*732 (in keys), p. 735 (1 species in Australia), p. 763 (original reference; characters; range); Crinoids of the Indian Ocean, 1912, p. 10 (does not extend beyond Hongkong; absent from‘Japan; reason), p. 11 (occurs both east and west of Ceylon), p. 12¥(represented"in Red’ Sea“region and in southeast African region), p. 21 (distri- bution in detail), p. 56 (in key), p. 117 (original reference; type) —F. W. CLarKz and WHEELER, U. S. Geol. Surv. Prof. Pap. 90-L, 1915, p. 195 (inorganic constituents of the skeleton); Prof. Pap. 102, 1917, pp. 23ff. | (same).—A. H. Crarx, Unstalked crinoids of the Siboga-Exped., 1918, p. 72 (in key; range).—F. W. Cranky and Wueeter, U. S. Geol. Surv. Prof. Pap. 124, 1922, p. 20 (inorganic constituents of the skeleton).—Gistin, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, p. 44. Diagnosis —A genus of Himerometridae in which the cirri are long, with 40-60 segments, nearly straight or slightly curved, tapering gradually from a stout proximal portion to a sharply pointed tip, with the distal segments much longer than broad and without dorsal processes; there is no opposing spine, and the terminal claw is nearly straight. P is shorter and more slender than P,, which in turn is shorter and more slender than P,. The IIIBr series are usually developed only on the inner side of the IIBr series, and are 2; if they are developed both interiorly and exteriorly they are commonly 2 interiorly and 4(3+4) exteriorly. Geographical range—From the Gulf of Martaban and the Andaman Islands to Amboina, and northward to the Philippine Islands. Bathymetrical range.—Littoral and sublittoral; the only recorded depth is 33 meters. Remarks.—While the single species included in this genus is, because of the long, nearly straight, gradually tapering, sharp pointed and smooth cirri, strikingly different in its general appearance from the species of Heterometra it scarcely deserves to be separated from them. It agrees with them in all its other characters. Moreover, slight dorsal tubercles are present in one of the known specimens on the outer cirrus segments, while in some of the species of Heterometra these are so slight as to be barely noticeable. History.—The single species now included in this genus was originally described in 1882 in the genus Antedon under four different names. With my first revision of the old genus Antedon in 1907 this species was placed in the new genus Himerometra. In 1909 the genus Craspedometra was established with Antedon acuticirra P. H. Carpenter, 1882, as the genotype, and the nominal species listed as referable to it were Craspedometra acuticirra, C. australis, C. bipartipinna, and C. ludovici, all of which represent in reality the same form. In 1911 I described two new species under the names Craspedometra ater and C. madagascarensis, and in 1912 in my monograph of the crinoids of the Indian Ocean I listed these, gave the name Oraspedometra amboinae to the specimens from Amboina referred by Hartlaub in} 1891 to Antedon ludovici, and referred Carpenter’s Antedon anceps to the genus Craspedometra. In 1913 Dr. A. Reichensperger used the names Craspedometra amboinae and C. anceps and suggested that the specimen he referred A MONOGRAPH OF THE EXISTING CRINOIDS 215 to C. anceps might represent a new species, in which event it could be called C. aruensis. Recognizing the rather obvious fact that Craspedometra, including acuticirra, ater, madagascarensis, and amboinae, was quite impossible of definition as distinct, corti Heterometra, in 1918 I included it in a key to the genera of the Himerometridae with structural Gharactens and a geographical range applying only to acuticirra, although it was not specifically stated that this was the only species now included. CRASPEDOMETRA ACUTICIRRA (P. H. Carpenter) Puate 13, Ficurzs 50, 51; Puars 20, Ficures 76-81 [See also vol. 1, pt. 1, fig. 85 (side view), p. 189; fig. 255 (centrodorsal), p. 253; fig. 471 (centrodorsal), p. 361; pt. 2, figs. 43, 44 (radial pentagon), p. 26; figs. 141, 142 (arm base), p. 83; figs. 449, 450 (pinnule tip), p. 261; figs. 716, 717 (disks), p. 346.] Antedon acuticirra P. H. Carpenter, Journ. Linn. Soc. (Zool.), vol. 16, 1882, p. 509 (description; locality unknown); Proc. Zool. Soc. London, 1882 (1883), p. 746 (specific formula) ; Challenger Reports, Zoology, vol. 26, pt. 60, 1888, pp. 55, 116, 253, 366, 380.—Harriaus, Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 29 (synonym of A. ludovici), p. 32 (characters of the type specimen).—Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1582 (listed)—A. H. Cuarx, Proc. U. S. Nat. Mus., vol. 34, 1908, p. 455 (listed); Smithsonian Misc. Coll., vol. 60, No. 10, 1912, p. 2 (identity); Crinoids of the Indian Ocean, 1912, p. 35 (same). Antedon ludovict P. H. Carpmnrer, Journ. Linn. Soe. (Zool.), vol. 16, 1882, p. 510 (description; Hongkong); Proe. Zool. Soe. London, 1882 (1883), p. 746 (specific formula) ; Challenger Reports, Zoology, vol. 26, pt. 60, 1888, pp. 55, 253, 366, 380.—HartLaus, Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 19 (in key), p. 29 (includes A. acuticirra and ?A. bipartipinna [specimens described from Amboina are Heterometra amboinae]), p. 30 (notes on the type specimen).— Déperie, Denkschr. med.-nat. Ges. Jena, vol. 8, pt. 5, 1898, p. 477 (Amboina).—Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1582 (listed) —A. H. Cuarx, Proc. U. S. Nat. Mus., vol. 34, 1908, p. 481 (listed); Mem. Australian Mus., vol. 4, 1911, p. 720 (of Déderlein, 1898 = Craspedometra acuticirra); Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p. 2 (of P. H. Carpenter, 1882 = Craspedometra acuticirra); Crinoids of the Indian Ocean, 1912, p. 35 (of P. H. Carpenter, 1888 = C. acuticirra + C.amboinae [error]), p. 37 (of Hartlaub, 1891 = ? [C. amboinae should have been given here]), p. 39 (of Déderlein, 1898 = C. acuticirra). Antedon australis P. H. Carppnter, Journ. Linn. Soc. (Zool.), vol. 16, 1882, p. 510 (in text under heading of A. acuticirra; Sydney, New South Wales; notes)—A. H. Cuarx, Proc. U. S. Nat. Mus., vol. 34, 1908, p. 459 (listed); Vid. Medd. Nat. Foren. Kgbenhayn, 1909, p. 159 (synonym of acuticirra; description of the type); Mem. Australian Mus., vol. 4, 1911, p. 713 (history; identity). Antedon bipartipinna P. H. Carpenter, Journ. Linn. Soe. (Zool.), vol. 16, 1882, p. 512 (description; Hongkong); Proc. Zool. Soc. London, 1882 (1883), p. 747 (specific formula); Challenger Reports, Zoology, vol. 26, pt. 60, 1888, pp. 55, 253, 366, 380.—Harrtaus, Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 19 (in key), p. 29 (?synonym of A. ludovici), p. 31 (notes on the type specimen).—Mrnckert, Archiv fiir Naturg., Jahrg. 71, 1905, vol. 1, Heft 1, p. 228 (regenera- tion).—Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1582 (listed) —A. H. Cuark, Proc. U.S. Nat. Mus., vol. 34, 1908, p. 462 (listed) ; Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p. 2 (= C. acuticirra); Crinoids of the Indian Ocean, 1912, p. 35 (same). Himerometra ludovict A. H. Cuarx, Smithsonian Mise. Coll., vol. 50, 1907, p. 356 (listed). Crospedometra acuticirra A. H. Cuarxk, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 9 (listed); Vid. Medd. Nat. Foren. Kgbenhavn, 1909, p. 159 (includes A. australis; description of the type specimen of australis), p. 161 (Singapore; description of a small specimen), p. 193 (collected at Singapore by Svend Gad); Proc. U. S. Nat. Mus., vol. 37, 1909, p. 31 (color of cirri compared 216 BULLETIN 82, UNITED STATES NATIONAL MUSEUM with that of C. aliena, sp. nov.); A. H. Cuarx, Bull. Mus. Hist. Nat., Paris, 1911, No. 4, p. 251 (Singapore; description of a specimen); Die Fauna Siidwest Australiens, vol. 3, Lief. 13, 1911, p. 440 (East Indian species occurring south to Sydney), p. 443 (range on the east coast) ; Mem. Australian Mus., vol. 4, 1911, p. 718 (= A. australis), p. 722 (ranges south to Port Jackson or Sydney), p. 763 (synonymy; characters; Australian record; range); Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p. 2 (acuticirra, ludovici, australis, and bipartipinna all the same species), p. 19 (discussion and descriptions); Crinoids of the Indian Ocean, 1912, pp. 35, 39 (synonymy), p. 39 (synonymy; new records; descriptions), fig. 8, a, b, p. 118; Proc. Biol. Soc. Washington, vol. 26, 1913, p. 179 (range in eastern Asia).—REICHENSPERGHR, Abh. Senck. naturf. Ges., vol. 35, Heft 1, 1913, p. 99 (comparison with C. amboinae), p. 100 (comparison with C. anceps [aruensis ?sp. n.]).—A. H. Cuarg, Journ. Washington Acad. Sci., vol. 5, 1915, p. 214 (Malayan species; range and its significance); Internat. Rev. gesamt. Hydrobiol. und Hydrogr. 1915, pp. 223 ff. (detailed account of distribution in Australia); Rec. Indian Mus., vol. 34, pt. 4, 1932, p. 551 (listed), p. 556 (Trotter Island); Treubia, vol. 14, livr. 2, 1933, p. 207 (55 specimens in collection), p. 208 (not included in Siboga collection), p. 211 (off Cape Jabung; near Deli; Malacca Strait; notes) —T#. Mortensen, Hong Kong Nat., Suppl., No. 3, 1934, p. 5 (Hongkong).—Guistin, Kungl. Fysiogr. Siillsk. Handl., new ser., vol. 45, No. 11, 1934, p. 27. atcaioneia australis A. H. Cuarx, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 9 (listed). Craspedometra bipartipinna A. H. Cuark, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 9 (listed). Craspedometra ludovici A. H. Cuark, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 9 (listed). Diagnostic features.—The cirri are long, stout basally and gradually tapering to a sharp point distally, nearly straight or slightly curved, composed of 40-60 segments of which the distal are much longer than broad and are entirely smooth dorsally; there is no opposing spine and the terminal claw is nearly straight. Description of a specimen labeled Sydney, N. S. W.—The centrodorsal is large, thick discoidal, with a broad flat polar area 5 mm. in diameter. The cirrus sockets are arranged in a single marginal row. The cirri are XVI, 42-46, 45 mm. long, large and stout basally but gradually tapering, rather more rapidly in the distal portion, to a very slender sharp pointed tip. The first four or five segments are about two and one-half times as broad as long, and those following gradually increase in length, becoming about as long as broad at the tenth, then slightly longer than broad, and in the terminal 10 or 12 about twice as long as broad. The cirri are rounded dorsally with no dorsal spines or carination. The terminal claw is long, about as long as the penultimate segment, and nearly straight. The anterolateral corners of the radials are barely visible in the interradial angles of the calyx. The IBr, are very short, six or eight times as broad as long, rather longer laterally than in the midradial line, almost entirely united laterally. The IBr, are low triangular, about three times as broad as the median length, the lateral borders about as long as those of the IBr,, well separated laterally. There are 6 I1Br series present, all 4(3+4). Each IIBr series bears two IJIBr 2 series, there being in all 12 of these. One IIIBr series bears a IVBr series of 2 on the inner side of the IIBr axillary and outer side of the following IIIBr axillary. The division series are well separated, well rounded dorsally, the sides smooth without lateral projections or carination. The ossicles immediately following each axillary are united interiorly. The 29 arms are 130 mm. long. The first eight brachials are approximately oblong, about three times as broad as long, those following becoming obliquely wedge-shaped, A MONOGRAPH OF THE EXISTING CRINOIDS ZIV almost triangular, about three times as broad as long, then less obliquely wedge- shaped, and almost short oblong or discoidal in the outer half of the arm, though never regularly short discoidal as in Himerometra. The brachials are slightly over- lapping. On an arm arising from a [Br axillary P, is 9 mm. long, moderately stout basally but tapering rapidly and becoming slender and almost flagellate in its distal half. It is composed of 28 segments of which the first is large and broad, about twice as broad as long, the second is smaller but similarly proportioned, the third is slightly longer, and those following are similar to the third, but become short in the terminal 10 or 12. The larger proximal segments are moderately carinate, the carination being somewhat roughened and finely tubercular. P, is 13 mm. long, stout basally and tapering to a slender tip, with 31 segments of which those in the proximal half are not so long as broad and rather strongly carinate, the following are about as long as broad with diminishing carination, and the terminal are short. P; is 21 mm. long, stouter than P,, tapering evenly to a slender tip, with 38 segments most of which are about as long as broad, rather shorter basally and rather longer distally. The segments in the proximal half have somewhat tubercular carinate processes which, however, are never very large, and there is a prominent ridge running along the entire pinnule distal to the median outer line. P, is 25 mm. long, about as stout basally as P, and similar to it, with 38 segments. P; is 14 mm. long, about as stout basally as P, but tapering more evenly and not so slender distally, with 25 segments, most of which are not quite so long as broad. Pp is like P, as described, though some- what smaller. On arms springing from a IIIBr axillary P, is like P, as described, P, is 17 mm. long, like P, as described, with 35 segments, P; is 28 mm. long, with 38 segments, resembling P, as described, and P, is 14 mm. long resembling P; as de- scribed. The carination of the proximal pinnule segments dies away after about Py,. The distal pinnules are small and very delicate, 7 mm. long. The color in alcohol is uniform dark brown. Notes.—In the specimen collected by the Investigator in the Andaman Islands the cirri are X, 43+, 50 mm. long. There are 22 arms 120 mm. Jong. All the I[Br series are present. The two IIIBr series are 2, developed internally. The color is whitish, the articulations purple, the arms crossed by regular broad deep purple bands; the cirri are white, each segment with a saddle of purple ventrally, becoming rusty brown at the tip. The specimen from Yé, Burma, is an especially fine example of the species. The cirri are XII, 42+,43-+,and44-+. There are 36 arms90mm.long. The IIBr series are all 4(3+4), and the IIIBr series are all 2; the 10 internal IIIBr series are all present, and in addition there are six external [1[Brseries. The synarthrial tubercles are prominent. ; In the specimen from the Mergui Archipelago the cirri are XIII, 43-44, 40 mm. long. There are 31 arms 120 mm. long. The IIBr series are all present, and all are 4(3+4). The 10 internal II[Br series are all present, and there is an additional ex- ternal one. Two of the IIIBr series are 4(3+-4), the other nine being 2. 208244—40——-15, 218 BULLETIN 82, UNITED STATES NATIONAL MUSEUM In color the two last are yellowish white, the pinnules with occasional small ir- regular blotches of light purple; the cirri are white, becoming rusty brown at the tip, each segment with a saddle of dull purplish. The specimen from Singapore collected by M. Maindron is small, with 13 arms. The two I[Br series are 4(3-+4). The single II1Br series, which is internally de- veloped, is2. The cirri have 28-34 segments. This specimen resembles the following. The specimen collected by Svend Gad at Singapore has the centrodorsal moderate in size, discoidal, with the bare polar area flat. The cirri are XI, 36-40, 35 mm. long, very stout basally but tapering, with especial rapidity in the distal third, to a very sharp and slender tip. The first segment is very short and those following gradually inerease in length to the seventh, which is about as long as broad, after the middle of the cirrus becoming very gradually longer than broad, and about half again as long as broad in the terminal portion. The terminal claw is long and slender, about as long as the penultimate segment, and is slightly curved. There are 19 arms. Four IIBr series are present, one on each of four postradial series; all of them are 4(3+4). Two of the IIBr series bear two IIIBr series each, and the two other IIBr series bear a single IIIBr series each, developed internally. All the IJIBr series are 2. P, is about 5 mm. long and is composed of 18 segments all of which are about as long as broad, and the second-fifth are rather strongly carinate. P, is 11 mm. long, much stouter than P, but tapering evenly to a slender and delicate tip, and is composed of 28 segments, which at first are about twice as broad as long, becoming about as long as broad on about the sixth, then slowly increasing in length and being about twice as long as broad dis- tally, but shorter again terminally. P;is 20mm. long, stouter than P, but, like it, taper- ing evenly to a slender and delicate tip. The fourth-sixth segments are about as long as broad, and those following gradually increase in length, being about twice as long as broad distally. The second-fifth or -sixth segments are, as in Ps, rather strongly carinate. P, is shorter than P, and is less stout and stiff; like the pinnules following it is composed of mostly squarish segments. One of the specimens from off Cape Jabung has 28 arms, and another has 22 arms. Nine of the specimens from near Deli have 22 (one), 23 (two), 24 (three), 26 (one), 28 (one), and 29 (one) arms. In the specimen with 28 arms the arms are about 140 mm. long. In one of the specimens from Malacca Strait near the mouth of the Deli River there are 35 arms; all the IJIBr series are 2. Prof. Ludwig Déderlein recorded two specimens from Amboina without comment. He gave a reference to Hartlaub, and it is quite probable that he determined the speci- mens from Hartlaub’s description and figures. If this is true they represent Hetero- metra amboinae and not this species. The specimen from Hongkong described as Antedon ludovici by Carpenter pre- sented, according to him, the following characters: The centrodorsal is a thick disk with a flattened dorsal surface and a single or partially double row of cirri. The cirri are XXV, 40-50. The cirrus segments are tolerably equal, all of them, except for those at the extreme end, being broader than long. Even in these terminal seg- ments the length is but little greater than the width; they bear a slight tubercle in the middle of the dorsal surface which is most marked on the penultimate segment. The radials are partially visible. The IBr, are short, widely oblong, almost completely A MONOGRAPH OF THE EXISTING CRINOIDS 219 united laterally, and slightly raised in the middle of their junction with the axillaries, The IBr, (axillaries) are half again as long as the IBr,, widely pentagonal, with open distal angles, and also slightly raised in the middle of the proximal edge. The 30 arms are 100 mm. long and are composed of 150+ brachials. The postradial series may divide three times. The IIBrseries are 4(3-+-4), and the IIIBr series are 2. The first ossicles following each axillary are rhomboidal and are closely united laterally, with the distal edge very slightly incised to receive the convex hinder edge of the following ossicle. The first syzygial pair, composed of brachials 3-+-4, is short and oblong. The next four or five brachials are also short with somewhat oblique terminal faces and slight backward projections alternately on the inner and outer sides. The fourth and sixth brachials are thus longer on their inner sides, and the fifth and seventh on their outer ones. The following brachials are short, bluntly wedge-shaped, and overlapping, nearly twice as broad as long, becoming more oblong about the middle of thearm. The first syzygy is between brachials 3-+-4, and the next is from between brachials 94-10 to between brachials 17+-18, usually between brachials 13+-14; there is then an interval of 8-15 (usually 10 or 11) muscular articulations between the syzygies. Pp is quite small, little more than half as long as, and far more slender than, those on the second and third brachials, which are about equal. The next pair are still larger, and the third pair (on the sixth and seventh brachials) still more so, consisting of about 35 stout segments and reaching nearly 35 mm. in length. That on the seventh brachial is rather the smaller, and the next pair is considerably smaller, only about equal to the first pair. The diminution continues to about the seventh pair, which are not especially small, being scarcely less than 10 mm. long. The re- maining pinnules are of about the same length, but gradually become slenderer. The basal segments of the lower pinnules from Pp onward have rather sharp dorsal keels. These may be less marked on the stout segments of the pinnules of the second and third pairs, but reappear on the second and four following segments in the fourth pair, and then gradually decrease in distinctness, disappearing altogether by about the twenty-fifth segment. The disk is naked and considerably incised and is 15 mm. in diameter. Sacculi are closely set along the pinnule ambulacra. The color in alcohol is light brown. The other specimen from Hongkong was described by Carpenter as Antedon bipartipinna in the following terms. The centrodorsal is a thick slightly convex disk bearing a single row of cirri. The cirri are XIV, with nearly 60 segments and may reach almost 60 mm. in length. They are long and stout. The basal segments are very broad, nearly 2 mm. in width, the fifteenth and following are about as long as broad, and the terminal are slightly longer than broad, quite smooth dorsally. The terminal claw is very imperfectly formed and there is no trace of an opposing spine. The radials are partially visible at the angles of the calyx. The IBr, are shorter in the middle line than laterally, where they are closely united to their fellows. The IBr, and the short, almost triangular, axillaries rise to a slight tubercular eleva- tion in the middle line of their junction. The IIBr series are 4(8+4). The [IBr series are usually 4(3-++4) but sometimes 2. The first ossicles after each axillary are rhomboidal and closely united laterally, and the second are more wedge-shaped. The middle of the articulation between the first two ossicles is tubercular, as in the case 220 BULLETIN 82, UNITED STATES NATIONAL MUSEUM of the articulation between the elements of the IBr series. The division series are in close lateral contact. The 35 arms are nearly 125 mm. long and consist of 200+ bra- chials. The first syzygial pair (composed of brachials 3+-4) and the next four or five brachials are oblong, and the following brachials are short, sharply wedge-shaped, and very slightly overlapping, twice as broad as long. About the middle of the arm the brachials become more equal sided, and they are nearly oblong in the terminal por- tion. Syzygies occur between brachials 3-4-4, again from between brachials 12+-13 to between brachials 16+-17, and distally at intervals of from 7 to 13 (usually 9 or 10) muscular articulations. Pp is quite short. Pp», P;, and P, are rather longer. They all have somewhat the appearance of being in two parts, as if they had been broken and had regenerated. The lower half consists of wide and thick segments with dorsal keels, while the outer half is composed of quite small segments and grows, as it were, out of the middle of the wide lower portion. This is least marked on P2, which is nearly twice as long as P, and is stouter and more uniformly tapering. P; is still longer, reaching 25 mm. in length and consisting of about 50 broad segments, of which the lower are keeled. P, is nearly as long but is less stout. On the inner side of the arm the seventh brachial bears a large pinnule (P,) like the preceding one (P3). That on the fifth brachial (P,) is much smaller, and that on the ninth (Py) is variable, sometimes small and sometimes nearly as large as its fellow of the fourth pair. In some arms the fourth or eighth segments may bear the largest pinnule (that is, P, or Py may be the largest). Beyond the fourth pair (P, and Py) the length decreases, rapidly at first but after- ward more gradually until about the twentieth (twenty-second counting syzygial pairs as two brachials) brachial beyond which the pinnules are tolerably uniform in size, decreasing again toward the arm tips. The carination of the basal segments of the lower pinnules dies away gradually and is lost after the tenth pair. The disk is naked and much incised, 15 mm. in diameter. Sacculi are closely set along the pin- nule ambulacra. The color is deep purple, almost black. Hartlaub wrote that the type specimen of Antedon bipartipynna, which he ex- amined at Hamburg, at first gives quite the same impression as one of the specimens from Amboina which he identified as Antedon ludovici (=Heterometra amboinae; see page 297). The most striking features are afforded by the cirri, which are long and stand in a single row and have no opposing spine or terminal claw. But Hart- laub said that one cannot attribute great weight to these features, since certain aspects of the specimen point to an abnormal development. First of all, there is the irregular position of the first brachial syzygy. On many arms it lies between brachials 3+4, on others between brachials 2+3, and on still others there is a syzygy between brach- ials 3+-4 and another between brachials 54-6; in individual cases the first brachial syzygy is between brachials 12413 or 13+14. Concerning the peculiarities of the lower pinnules Hartlaub said that it is a fact that they have been broken off and regenerated. Their structure, however, corresponds completely to that characteristic of the lower pinnules of A. ludovici. He said further that both species further agree in the possession of both III[Br 2 and I1IBr 4(3+4) series. Carpenter’s statement that the IIIBr series are usually 4(3-+-4) but sometimes 2 is not correct. The speci- men has many more IIIBr 2 than IIIBr 4(3+4) series, and indeed the inner IIIBr series are for the most part 2. A MONOGRAPH OF THE EXISTING CRINOIDS 221 I examined both these specimens from Hongkong at the Hamburg Museum in 1910. They undoubtedly represent the same species as the one without locality that Carpenter described as Antedon acuticirra. 'The second specimen from Hongkong, described as Antedon bipartipinna by Carpenter, has slightly smoother arms than the type of A. acuticirra. The specimen without locality was described as Antedon acuticirra by Carpenter in the following terms. The centrodorsal is a thick disk with a flat dorsal surface and the cirri arranged in a single marginal row. The cirri are XV, 55, and may reach 50 mm. in length. They are long and tapering and consist of smooth segments. The basal segments are broad, and the tenth is about as long as broad, while those following diminish in width and thickness, though the length changes but little. The later ones are longer than broad but are not laterally compressed. The cirri taper gradually to a sharp point. There is no trace of an opposing spine. The terminal claw is small and is but slightly curved. Some of the radials are partially visible, and some of the short IBr, are partly concealed. The IBr, are closely united laterally. The IBry (axillaries) are nearly twice as long as the IBrj, almost triangular, with open distal angles. The IIBr series are 4(3+4), and the IIIBr series are usually 2, but sometimes 4(3+4). The first ossicles following each axillary are closely united interiorly, and there are slight synarthrial tubercles. The 26 arms are about 125 mm. long and are composed of 200+ brachials. The first brachials are rhomboidal, short and broad. The second brachials are more wedge-shaped. The first syzygial pair (composed of brachials 3+4) and the next four or five brachials are short and oblong, the following brachials are bluntly wedge-shaped, twice as broad as their longer side and slightly overlapping, and the middle and later brachials are more oblong and overlap rather less. Syzygies occur between brachials 3+-4, again from between brachials 10+11 to between brachials 17+18, and distally at intervals of 6 to 33 (usually 11 to 16) muscular articulations. Pp is comparatively small, with keeled and expanded basal segments. P, is considerably larger on the outer arms but remains small on the inner ones. P, and P; increase in size, the latter reaching 25 mm. in length and consisting of nearly 40 segments. P, is small like Pp, and P, and P, are much larger, P, approaching the size of P;. P, is somewhat smaller again, though it is still long, and the next pair are a good deal shorter than their immediate predecessors, though somewhat larger and stouter than P;. Where IIBr series are not present but the arms arise from a IBr axillary, the fourth pair of pinnules are large like their immediate predecessors. The size of the pinnules decreases to about the seventeenth brachial and then increases very slowly again, the outer pinnules reaching only about one-third the length of the largest lower pinnules, which have broad and strongly keeled basal segments. On the smaller pinnules after the fifth pair this carination is less marked, but it is traceable for some little way out on to the arms. The disk is naked and considerably incised, 15 mm. in diameter. Sacculi are very close along the pinnule ambulacra. The color is nearly white, with traces of a deep violet remaining. I examined this specimen at the Hamburg Museum in 1910. The synarthrial tubercles are rather prominent, and the distal edges of the brachials are rather strongly overlapping. It is just like others from the eastern side of the Bay of Bengal. 222 BULLETIN 82, UNITED STATES NATIONAL MUSEUM In the remarks following his description of Antedon acuticirra Carpenter men- tioned a specimen from Sydney, New South Wales, which he had seen in the Copen- hagen Museum bearing the manuscript name Antedon australis Liitken. This is the specimen from which the description given above (page 216) was drawnup. Carpenter said that it has some resemblance to Antedon acuticirra. It has the same arrangement of the division series, a few long and many-segmented cirri, and a large third pair of pinnules. But it has relatively shorter axillaries and a shorter intersyzygial interval, usually 7-9 instead of 11-16 muscular articulations. He said that without a more detailed examination of the Copenhagen specimen than he had up to that time been able to make it was difficult to be certain as to its identity with the one which he described as Antedon acuticirra. Localities —Investigator; Andaman Islands [A. H. Clark, 1912] (1, U. S. N. M., 36177, 36198). Investigator; Yé, Burma (on the eastern shore of the Gulf of Martaban, at about lat. 15°20’ N.) [A. H. Clark, 1912] (1, I. M.). Investigator; Gregory Island, Mergui archipelago [A. H. Clark, 1912] (1, U.S.N.M., 35199). East side of Trotter Island, Mergui archipelago; November 30, 1913 [A. H. Clark, 1932] (1, I. M.). Singapore; M. Maidron, 1884 [A. H. Clark, 1911] (1, P. M.). Singapore; Svend Gad [A. H. Clark, 1909, 1912] (1, C. M.). Off Cape Jabung, northern coast of southern Sumatra (lat. 1°03’ S., long. 104°35’ E.); July 3, 1908 [A. H. Clark, 1933] (3, Buitenzorg Mus.). Near Deli, northern Sumatra (lat. 3°40’ N., long. 99°10’ E.); 16-18 meters; June 25, 1908 [A. H. Clark, 1933] (48, U.S. N. M., E. 3247; Buitenzorg Mus.). Malacca Strait, near the mouth of the Deli river, Northeastern Sumatra (lat. 3°53’ N., long. 98°46’ E.); 10 meters; June 22, 1908 [A. H. Clark, 1933] (4, Buiten- zorg Mus.). ?Amboina; Prof. Richard Semon [Déderlein, 1898; A. H. Clark, 1912]. Albatross station 5157; Tawi Tawi group, Sulu (Jolo) archipelago; Tinakta Island (N.) bearing S. 80° W., 3.3 miles distant (lat. 5°12’30’” N., long. 119°55’50’’ E.); 33 meters; fine sand; February 21, 1928 (1, U. S. N. M., 35186). Hongkong [P. H. Carpenter, 1882, 1883, 1888; Hartlaub, 1891; A. H. Clark, 1909, 1912] (1, H. M.). Hongkong [P. H. Carpenter, 1882, 1883, 1888; Hartlaub, 1891; A. H. Clark, 1907, 1909, 1911, 1912] (1, H. M.). No locality [P. H. Carpenter, 1882, 1883, 1888; Hartlaub, 1891; A. H. Clark, 1909, 1912] (1, H. M.). Erroneous locality—Sydney, New South Wales [P. H. Carpenter, 1882; A. H. Clark, 1909, 1911, 1912] (1, C. M.). Geographical range.—From the Gulf of Martaban and the Andaman Islands to Amboina and northward to the Philippine Islands and Hongkong. Bathymetrical range.—Littoral and sublittoral; from the shore line down to 33 meters. History.—In a paper on the comatulids of the Hamburg Museum published in 1882, Dr. P. H. Carpenter described a specimen without locality under the name of A MONOGRAPH OF THE EXISTING CRINOIDS 223 Antedon acuticirra, and in discussing this mentioned another specimen from Sydney in the Copenhagen Museum bearing the manuscript name Antedon australis given it by Liitken which had some resemblance to it. He also described two other species, Antedon ludovici and A. bipartipinna, both of which were based upon single specimens from Hongkong. In the key to the species discussed in the paper he included Antedon acuticirra and A. ludovici under the headings “Rays divide three times. First division of three joints, the axillary with a syzygy; second of two joints, the axillary without a syzygy [that is, I]Br series 4 (3-4) and JIIBr series 2] and also “Fifty or more cirrus-joints. Lower pinnules similar on all the arms.’’ The characters by which he distinguished acuticirra from ludovici were that in acuticirra the terminal cirrus seg- ments are smooth, while in /udovici the distal cirrus segments have dorsal tubercles. As given in the key the chief characters of A. bipartipinna are “Rays divide three times, each division of three joints, the axillary with a syzygy [that is, the I[Br and IJIBr series are both 4 (3+-4)],” while the cirri are long and stout with an imperfect terminal claw and smooth terminal segments. In the following year Carpenter gave specific formulas for Antedon acuticirra, A. ludovici, and A. bipartipinna. In the Challenger report on the comatulids published in 1888, Carpenter gave emended specific formulas for Antedon acuticirra, A. ludovici, and A. bipartipinna, and mentioned (in discussing Antedon [Thalassometra] bispinosa) that in A. acuticirra the “radial axillaries’’ (the IBr.) come into contact above the depressed lateral portions of the second radials (the IBr;). As none of these forms had been secured by the Challenger Carpenter did not discuss them in detail, merely including them in the key to the species of the “‘Savignyi group.” In this key he gave acuticirra as having the cirrus segments smooth and longer than broad, with no opposing spine, while in ludovici the cirrus segments are broader than long, the outer have faint tubercles, and the penultimate has an opposing spine. In the key Antedon bipartipinna was paired with A. philiberti, these two forms being, according to Carpenter, distinguished from all the other species in the ‘“‘Savignyi group” by the possession of IIIBr 4 (3+-4) series. Of A. bipartipinna he said that there are nearly 60 cirrus segments of which the outer are longer than broad and quite smooth, while the terminal segments of the lower pinnules are much smaller than the basal ones. He gave A. [Heterometra] philiberti as having 45 cirrus segments of which the later are short and spiny. In studying the collection of comatulids made by Prof. J. Brock at Amboina, Dr. Clemens Hartlaub found 12 specimens that he at first considered as representing a new species. He accordingly prepared a detailed description of them, but later he decided that they belonged to the form which had been described by P. H. Carpenter as Antedon ludovici. In 1891 he published the description he had drawn up under that name, in which he included Antedon acuticirra as a synonym and A. bipartipinna as a probable synonym. He said that he had sent a specimen from Brock’s Amboina collection to Carpenter, who had pointed out that it was apparently identical with his Antedon ludovici from Hongkong and that he convinced himself of the correctness of this conclusion by an examination of the type specimen of ludovict at Hamburg. He gave a detailed comparison between his Amboina specimens (which really represent Heterometra amboinae; see page 297) and the type specimen of Antedon ludovici. He regarded the Chinese form (ludovici) to which he assigned also the type specimens of 224 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Antedon bipartipinna and A. acuticirra as very variable, but believed it to differ constantly from the Amboina form in the greater number of arms. He said that the principal reason that led him to consider the Chinese and Amboina forms as identical was the agreement in the form of the cirri, arms, brachials, and pinnules. He regarded as also characteristic the synarthrial tubercules, the shortness of the first syzygial pair, the uneven dorsal surface of the first 7 or 8 brachials, and the carination of the pinnules in the proximal portion of the arms. He wrote that if one places weight on these apparently constant characters, and if one also takes into consideration the obvious inclination of A. ludovici toward variation, it may well be asked whether A. bipartipinna, coming also from Hongkong, is not really the same species. In 1895 Prof. René Koehler recorded this species from Amboina, and in 1898 Prof. Ludwig Déderlein again recorded specimens from Amboina. Koehler’s record was based on specimens of Heterometra amboinae (see page 300), and it is probable that Déderlein’s record was based on the same species. In 1907 I referred ludovici, considered as including acuticirra, australis, and bipartipinna, to the new genus Himerometra. In 1909 on the establishment of the new genus Craspedometra I listed as referable to this genus Craspedometra acuticirra, C. australis, C. bipartipinna, and C. ludovici. Thanks to the kindness of Dr. Th. Mortensen I had been able to examine the specimen supposedly from Sydney in the Copenhagen Musem that bore Liitken’s manuscript name Antedon australis. Later in 1909 I published a description of this specimen, referring it to Craspedometra acuticirra. At the same time I recorded and gave notes upon a specimen from Singapore. I remarked that this specimen is so different from the one from Sydney that I at first considered it as a representative of a different species, which was the reason for describing it so minutely. Subsequently I had had the opportunity of studying the series of specimens belonging to the Indian Museum at Calcutta, and my examination of these specimens proved to my satisfac- tion that the example from Sydney and the one from Singapore, different in general appearance as they are, really belong to the same specific type. In 1911 I recorded another specimen from Singapore that I had seen in the Paris Museum in 1910, and in 1912 I gave notes upon the type specimens of Antedon acuti- eirra, A. ludovici, and A. bipartipinna, which I had examined at the Hamburg Museum. All three forms, together with A. australis, I referred to the same species, which I called Craspedometra acuticirra. In my monograph on the crinoids of the Indian Ocean, which was published in 1912, Lincluded Antedon ludovici, A. australis, and A. bipartipinna in the synonymy of Craspedometra acuticirra, and recorded three specimens, one from the Andaman Islands, one from the Mergui Archipelago, and one from the Burmese coast, giving notes on all three. In 1932 I recorded from the collection of the Indian Museum a broken specimen from Trotter Island, and in 1933 I recorded and gave notes upon 55 specimens from three localities off the coast of Sumatra in the collection of the Buitenzorg Museum in Java. Dr. Th. Mortensen in 1934 listed this species as one of the crinoids known from Hongkong. A MONOGRAPH OF THE EXISTING CRINOIDS 225 Genus HETEROMETRA A. H. Clark Comatula (part) Aupoutn, in Savigny, Description de l’igypte, 1817 (1826), p. 205, and following authors. Alecto (part) J. Miittmr, Monatsber. preuss. Akad. Wiss., 1841, p. 185. Comatula (Alecto) (part) J. Miitumr, Monatsber. preuss. Akad. Wiss., 1846, p. 178. Antedon (part) P. H. Carpenter, Trans. Linn. Soc. (Zool.), ser. 2, vol. 2, 1879, p. 29, and following authors. Himeromeira (pert) A. H. Ciarx, Smithsonian Mise. Coll., vol. 50, 1907, p. 356. Amphimetra (part) A. H. Cuarx, Proc. Biol. Soe. Washington, vol. 22, 1909, p. 7. Craspedometra (part) A. H. Cuarx, Proc. U. S. Nat. Mus., vol. 37, 1909, p. 31. Heterometra A. H. Cuarg, Proc. Biol. Soe. Washington, vol. 22, 1909, p. 11 (diagnosis; genotype Antedon quinduplicava P. H. Carpenter, 1888); Amer. Nat., vol. 48, 1909, p. 254 (represented in the Red Sea); Zool. Anz., vol. 34, 1909, p. 367 (arm division similar to that of Antedon fluctuans [Zygometra comata]); Proc. Biol. Soc. Washington, vol. 22, 1909, p. 175 (referred to the Himero- metrinae); A. H. Cuarx, Mem. Australian Mus., vol. 4, 1911, pp. 731, 732 (in keys), p. 734 (key to the Australian species), p. 767 (original reference; characters; range); Crinoids of the Indian Ocean, 1912, p. 9 (possesses additional species in Australia), p. 10 (absent from Japan; reason), p. 11 (represented in the Ceylon region; this the western limit of the large and highly multibrachiate species), p. 12 (represented in Red Sea region and in southeast African region), p. 21 (range in detail), p. 56 (in key), p. 120 (original reference; type); Die Fauna Siidwest- Australiens, vol. 4, Lief. 6, 1913, p. 311 (discussion of the small species and comparison with the small species of Zygometra); Unstalked crinoids of the Siboga-Exped., 1918, p. 72 (in key; range), p. 75 (key to the included species).—GisL&n, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 45, 46, 48.—A. H. Cuarx, Temminckia, vol. 1, 1936, p. 305.—Gistin, Kungl. Fysiogr. Sallsk. Lund Forh., vol. 7, No. 1, 1936, p. 17. Diagnosis—A genus of Himerometridae in which the cirri are of moderate length with the distal segments always bearing dorsal tubercles or spines and the terminal claw always strongly curved; the distal cirrus segments are usually shorter than the proximal, or both distal and proximal cirrus segments are short; P, is shorter and slenderer than P,, and P, is usually shorter and more slender than P3; the enlarged proximal pinnules are much elongated, slender or rather stout, smooth, basally cari- nate, or with the outer segments armed at the distal end with spines or flangelike projections; there are more than 10 arms. In a few species 10-armed individuals occur, but these are always distinguishable from the species of Amphimetra by the characteristic proximal pinnules which are strongly keeled basally or have their outer segments provided distally with spines or flangelike projections. Geographical range-—From southern Japan southward to Port Curtis, Queens- land, and Baudin Island, northwestern Australia, and westward to Madagascar and the eastern coast of Africa from the Red Sea south to Zanzibar and Bagamoyo. Bathymetrical range—From the low-tide mark down to 111 meters. The species of Heterometra are especially characteristic of the zone from the low-tide mark down to about 50 meters. Of the 25 species only 7 are known from a greater depth than 50 meters, and all but 4 have been taken in shore collecting. Remarks.—The genus Heterometra includes a rather heterogeneous assemblage of 25 species. In the genus as a whole all the characters of the cirri, pinnules, and arms are exceedingly variable and more or less overlap those of the other genera of the family Himerometridae. The species of the genus Heterometra are therefore recognizable not so much by a special and distinctive generic combination of charac- 226 BULLETIN 82, UNITED STATES NATIONAL MUSEUM ters as by the fact that they do not fall in the genera Craspedometra, Himerometra, Amphimetra, or Homalometra. The genus Heterometra is divisible into five fairly well marked groups, which might possibly be considered as of subgeneric value. But with our present meager knowledge of most of the species it is quite inadvisable to subdivide it. In the first group the enlarged lower pinnules are comparatively slender and are entirely smooth, with no carination of the basal segments and no modification of the distal ends of the outer segments. The two species in this group, H. savignii and H. nematodon, appear not to be very closely related to each other. Another group in which the cirri are smooth, the distal segments being without dorsal processes although they may be more or less strongly carinate, and the en- larged lower pinnules are usually smooth though in large individuals the basal segments may be slightly carinate, is represented by H. quinduplicava only. Though in the character of the enlarged lower pinnules H. quinduplicava is intermediate between the species with these pinnules smooth and the species with these pinnules carinate basally, it is quite unique in the character of its cirri. In a third group the enlarged lower pinnules have the outer segments with their distal ends armed with spines or with rounded triangular extensions of the distal portions of the prismatic ridges. In H. variipinna the enlarged lower pinnules are composed of cylindrical segments with spinous distal ends, while in H. crenulata the lower pinnules are very strongly prismatic with the distal ends of the prismatic ridges on each segment produced into conspicuous rounded triangular lateral exten- sions so that the profiles of these pinnules are very strongly serrate, the teeth being broadly rounded. Heterometra crenulata is exceedingly variable in every feature. Individuals are not infrequent in which there are only 10 arms, though in others there may be as many as 30 arms. Individuals vary from extraordinarily robust with stiff and spinelike enlarged lower pinnules to very slender with slender lower pinnules which become almost flagellate distally. The processes on the segments of these pinnules may be broad and involve basally a large part of the segment, or they may be reduced to scarcely more than an overlapping spine. The cirri may be prominently spiny or practically smooth, stout, or slender and tapering. Some forms of H. cren- ulata approach H. producta—which may possibly turn out to be merely an extreme variation of it—while H. producta and H. propinqua probably will be found to inter- grade. Although H. crenulata and H. variipinna seem to be so very different, the enlarged lower pinnules in the latter may be more or less prismatic, at least basally, suggesting that the two species may be rather closely allied. In all the other species of Heterometra the enlarged lower pinnules are entirely smooth, but their earlier segments bear more or less strongly developed and usually conspicuous carinate processes on the side toward the arm tip. They are divisible into two quite distinct groups. The first of these, or the fourth group of Heterometra as a whole, includes 11 species—astyanaz, affinis, reynaudi, amboinae, ater, joubini, pulchra, gravieri, compta, singularis, and madagascarensis. In these species the brachials, though actually short, are relatively long, and are always more or less wedge-shaped with oblique ends, the earlier brachials, beyond the first few, being markedly wedge-shaped and in some cases almost triangular. This group is divisible into three subgroups—one A MONOGRAPH OF THE EXISTING CRINOIDS 227 in which the cirri are long and tapering with the outer segments longer than broad and with only slight dorsal tubercles, including astyanaz only; one in which the outer cirrus segments are broader than long but some of the earlier are as long as, or longer than, broad, including affinis, pulchra, gravieri, singularis, compta, and madagascar- ensis; and one in which all the cirrus segments are markedly broader than long, in- cluding, reynaudi, amboinae, ater, and joubini. Some species are more or less on the border line between the last two subgroups. Through H. astyanax the first section of this group approaches very closely the genus Craspedometra, which possibly should not be separated from it, and in another direction it also approaches the slender and delicate Homalometra. Through H. compta, in which P, and P, are similar and of approximately the same length, the second section approaches the genus Himerometra as represented by H. persica and H. bartschi, though the brachial structure of the species of Himerometra is most nearly like that of the next group. In the fifth and last group of the species of Heterometra, including philiberti, sarae, parilis, africana, schlegelii, bengalensis, and flora, the brachials are exceedingly short with parallel ends which become only slightly oblique on the earlier brachials. Within this group there are two subgroups, one, including philiberti only, in which IIIBr series occur and are wholly or mostly 4(3-+4) instead of 2 as would be expected, and another, including sarae, parilis, africana, schlegelii, bengalensis, and flora, in which, so far as is known, IIIBr series never occur. The species of this group are evidently closely allied to the species of Amphimetra, and this relationship is empha- sized by the occasional occurrence in some of them of fully grown individuals with only 10 arms. They approach the species of Amphimetra through H. sarae in which the earlier segments of the proximal pinnules have the carination reduced to a mere sharpening of the side toward the arm tip. Their brachial structure, and especially the occurrence of I1IBr 4(3+-4) series in philiberti, indicates that they also approach the species of Himerometra, through H. persica and H. bartschi, more closely than do the species of any of the other groups. Both of the species in the first group have a very limited range, H. savignii occurring only from the Red Sea to Muscat, and H. nematodon being found only in Queensland and the Aru Islands. The single species of the second group ranges from the Andaman Islands to the Philippines. The species of the third group occur from the Maldive Archipelago to the Philippines and southward to northern Australia. The species of the fourth group range from east Africa and the Red Sea to the Philip- pines and southward to the Torres Strait region. The species of the fifth group are found from east Africa to southern Japan, but are not represented in Australia. The species of Heterometra are all inhabitants of very shallow water and are most abundant in a narrow band of a few meters’ width just below the low tide mark. One species has been recorded from a depth of 111 meters, and others from a depth of 88 meters, but most of the records are for much shallower water. Indeed, of the 25 species only 4 (astyanaz, pulchra, compta, and propinqua) have not been taken in ordinary shore collecting. All the species are of medium or rather large size, but the largest of them are not so large as the larger species of Himerometra, or as the species of Craspedometra. The species of Heterometra are among the most difficult of all comatulids to identify. Most of those of which any considerable number of individuals are known 228 BULLETIN 82, UNITED STATES NATIONAL MUSEUM are exceedingly variable, generally or locally, in all their structural characters and also in their size. History—It is possible that the Alecto horrida described by Dr. W. E. Leach in 1815, which is the type species of the genus Alecto, represents the species described by Prof. Johannes Miiller in 1846 as Comatula (Alecto) reynaudi. But Leach’s de- scription is very vague, his figure is not with certainty identifiable, and his type specimen can not be found, so that the identity of horrida with reynaudi rests on quite inconclusive circumstantial evidence. The first indubitable species of this genus to be recorded was assigned by Savigny to the genus Comatula; it was figured in his account of the zoology of Egypt published in 1826 under the editorship of Audouin. In 1841 Prof. Johannes Miiller referred the same species to Alecto, and in 1846 he described a second species under the generic designation Comatula (Alecto). In 1879 Dr. P. H. Carpenter assigned savignii, philiberti, and reynaudi to the genus Antedon as understood by him. In 1882 he referred to the species of this type as belonging to the Savignyii group. In 1888 he defined the Savignyi group and included in it reynaudi, savignyi, and philiberti, the recently described variipinna, acuticirra, ludovici, and bipartipinna (the last three falling in the genus Craspedo- metra as now understood), and the new species anceps, quinduplicava, and angustiradia (the last a species of Adelometra). As one of the three specimens had only 10 arms, Carpenter also included Antedon anceps in his Milberti group. Antedon clemens he included in his Palmata group which consisted of species with the IIBr series 2. Hartlaub in 1891 followed Carpenter in placing clemens in the Palmata group, and included in the Savignyi group angustiradia, reynaudi, savignyi, anceps, variipinna, quinduplicava, ludovici, bipartipinna, and philiberti, and his new species bengalensis, martensi, kraepelini, brockii, affinis, nematodon, and crassipinna. The Savignyi group of Carpenter and Hartlaub was therefore the equivalent of the very closely allied genera Heterometra, Himerometra, and Craspedometra combined. In my first revision of the old genus Antedon, which was published in 1907, the species of Heterometra were all referred to the new genus Himerometra. In a revision of the family Himerometridae published in 1909, the genus Himero- metra was redefined in such a way as to include only those species in which Pp is larger and longer than P;, which in turn is larger and longer than P;. The species with I[Br 4(3+-4) series and therefore more than 10 arms, and Pp smaller than P,, were distributed among three genera. The species in which the middle and distal brachials are exceedingly short and discoidal and the elements of the division series and lower brachials are swollen were assigned, together with a number of 10-armed species, to the new genus Amphimetra. These species were anceps (=quinduplicava), producta, schlegelii, and variipinna (in the sense used by Carpenter, that is, including both variipinna and crenulata). The species with the middle and distal brachials not especially short, more or less obliquely wedge-shaped, and the elements of the division series and earlier brachials not swollen, were assigned to the new genus Heterometra. These species were affinis, bengalensis, brockii (=variipinna), quinduplicava, reynaudi, and savignii. The species with the cirri tapering distally, the distal cirrus segments twice as long as broad, and no opposing spine, were assigned to the new genus Cras- pedometra. ‘The only species at that time included in Craspedometra which is of present A MONOGRAPH OF THE EXISTING CRINOIDS 229 interest was ludovici which the author accepted on the basis of Hartlaub’s description of specimens from Amboina (=H. amboinae) published in 1891. Miiller’s philiberti was assigned to the genus Himerometra with a query. In the same year (1909) I described, under the name Craspedometra aliena, a supposed new species from the Philippines. The type specimen of this supposed new species is an example of H. crenulata, while other specimens in the type series represent H. quinduplicava. In another paper also published in 1909 I described Amphimetra mortenseni. In 1911 a specimen was recorded from the Philippines under the name Craspe- dometra anceps (=quinduplicava), and Antedon clemens and Craspedometra aliena were placed in the synonymy of anceps. In a paper on the crinoids of the coasts of Africa published in the same year, Amphimetra africana, Craspedometra ater, and Craspedometra madagascarensis, all referable to Heterometra as herein understood, were described, and in a paper on the crinoids of the Paris Museum, which appeared a few months later, the type specimen of Miiller’s philiberti was described under the name of Amphimetra philiberti, of which the recently described Amphimetra mortenseni was said to be a synonym. In my memoir on the crinoids of the Indian Ocean published in 1912, the genus Heterometra is given as including nematodon, reynaudi, savignii, compta, affinis, brockit (=variipinna), martensi (=Himerometra martensi), joubini, gravieri, aspera (=bengalensis), singularis, bengalensis, quinduplicava, and, in an appendix, pulchra. In the genus Amphimetra are placed philiberti (Gncluding mortensent), variipinna (including crenulata), producta, schlegelii, africana, and a number of 10-armed species. To Craspedometra are assigned ater, madagascarensis, amboinae (u new name for the species described and figured as ludovici by Hartlaub), and anceps. In a paper on the crinoids of the Hamburg Museum published in 1912 philiberti, schlegelii (from which the 10-armed individuals are excluded and considered as repre- senting laevipinna), nematodon, crenulata, and varvipinna are assigned to Amphimetra. In a paper on the crinoids of the Berlin Museum published in the same year africana is assigned to Amphimetra, ater and madagascarensis to Craspedometra, and savignit to Heterometra. In a paper on the crinoids of the British Museum, which appeared in 1913, Amphimetra is given as including crenulata, nematodon, anceps, producta, flora (sp. nov.), and africana, as well as several 10-armed species; Hartlaub’s martensi is cor- rectly assigned to Himerometra; and quinduplicava, reynaudii, and savignii are given under Heterometra. Dr. August Reichensperger in his memoir on Merton’s crinoids from the Aru and Kei Islands published in 1913 recorded and discussed Amphimetra variipinna, Cras- pedometra amboinae, and Craspedometra anceps, and Dr. Torsten Gislén in his account of Mjoéberg’s crinoids from northwestern Australia published in 1919 (but written be- fore the appearance of the Siboga report in 1918) discussed Amphimetra variipinna. For some time it had been increasingly evident that the genera Heterometra, Amphimetra, and Craspedometra were quite artificial and could not be maintained on the basis of the original diagnoses. So a thorough revision of the family Himerome- tridae was undertaken, and the results of this study were incorporated in the report upon the unstalked crinoids of the Siboga expedition, published in 1918. In this re- 230 BULLETIN 82, UNITED STATES NATIONAL MUSEUM port Heterometra was given as including qguinduplicava (including clemens and anceps), savignii, nematodon, variipinna (including brockii), producta, propinqua, crenulata (including Comatula dubia von Graff, 1877, Antedon decipiens and A. irregularis Bell, 1884, Antedon bidentata von Graff, 1884, Antedon dubia P. H. Carpenter, 1888, and Oraspedometra aliena A. H. Clark, 1909, with the exception of the 10-armed specimens mentioned in the description of the last named), philiberti (including mortenseni), flora, africana, schlegelii, singularis, pulchra, amboinae, joubini, gravieri, ater, madagascar- ensis, reynaudii, compta, aspera (=bengalensis), bengalensis, and affinis. This inter- pretation of the genus Heterometra has been accepted herein, but the key to the species given in the Siboga report proving rather unsatisfactory, entirely new keys have been prepared. KEY TO THE SPECIES IN THE GENUS HETEROMETRA a', Outer cirrus segments smooth dorsally, or with a slight median keel, never with dorsal spines or tubercles; 25-36 (usually about 30) cirrus segments; 16-26 arms 60-145 mm. long; division series rather broad with slightly produced borders, though not in contact beyond the IBr axillary (from Philippines to Andaman Islands; 0-80 meters) -- ------ quinduplicava (p. 245) a?. Outer cirrus segments bearing spines or tubercles dorsally. b'. Basal segments of the enlarged proximal pinnules without carinate processes on the side toward the arm tip. c!, Enlarged proximal pinnules smooth and slender, their component segments without lateral processes or spinous distal ends. d', Usually 20 arms, up to 125 mm. in length; enlarged lower pinnules very long and slender, composed of about 20 segments; cirri composed of 23-33 (usually 25-30) segments of which the outer bear long dorsal spines (Red Sea and eastward to Muscat; 0-18 [?22] meters) 22. seelst oe oe bee ok wasn ben cee oe See ee eee savignii (p. 235) d@. Arms 38-48 in number, 105-115 mm. long; enlarged lower pinnules composed of 26-30 segments; cirri composed of 40-50 segments of which the outer bear prominent, though not especially long, dorsal spines (Queensland and Aru Islands; 0-22 [736] meters) nematodon (p. 241) c?, Segments of the enlarged proximal pinnules with variously modified distal ends so that the profile of these pinnules is serrate. d', Distal ends of the segments of the enlarged proximal pinnules with rounded processes, but no spines; enlarged proximal pinnules sharply prismatic and usually much stiffened, with the distal portion of the prismatic ridges on each segment produced into a rounded- triangular process so that the profiles of these pinnules are very strongly serrate, with conspicuous rounded teeth; division series and arm bases very rugged, the component ossicles swollen dorsally, with prominent synarthrial and articular tubercles; 10-30 (usually between 11 and 22) arms 70-150 mm. long; cirri very variable, with 25-46 (usually between 30 and 45) segments (from Philippines to Andamans [possibly Mal- dives], and southward to northern Australia; 0-111 meters) ______- crenulata (p. 253) @. Distal ends of the segments of the enlarged proximal pinnules with spines, either all around or segregated at the prismatic angles. el, Enlarged proximal pinnules stout, composed of cylindrical or only very slightly pris- matic segments which have everted and spinous distal ends, and usually larger on the outer than on the inner arms of each ray; disk more or less extensively plated; 11-28 arms 50-140 mm. long; cirri with 16-38 (usually 30-35) segments (Amboina and Ard Islands; 0-20" meters) 22. 5 et) SORE ans Oe Ae variipinna (p. 278) e?. Enlarged proximal pinnules slender, composed of strongly prismatic segments, which bear one or (usually) several spines at the distal prismatic angles, and the same size on all the arms; disk unplated; slender, with not more than 13 arms 50-100 mm. long. j'. Cirri with 23-26 segments which from the tenth or eleventh onward are about as long as broad (Singapore and Maldive Islands; littoral)___..______ producta (p. 275) A MONOGRAPH OF THE EXISTING CRINOIDS 231 f?. Cirri with 24-36 (usually 30-33) segments all of which are about twice as long as broad at the ends (south of eastern Borneo; 82-88 meters)____propinqua (p. 277) b?. Basal segments of the enlarged proximal pinnules with prominent thin carinate extensions or keels on the side toward the arm tip. ct. IIIBr series always present, all or most of them 4(3+-4); brachials exceedingly short, mostly oblong; cirri with 26-45 (usually 30-35) segments, the longest never so long as broad; median dorsal tubercles or small spines are developed from the twelfth-fourteenth onward; 18-27 arms 85-150 mm. long (Andaman Islands to Kwala Cassan and Java; littoral) philiberti (p. 313) c?, III Br series 2, or absent. d', The outer cirrus segments bear long, slender, sharp, and prominent dorsal spines which begin abruptly. el, Cirri stout, with all the component segments broader than long. f!. Cirri very stout with all the segments much broader than long; 10-13 arms 170-180 mm. long; cirri XV, 29-30 (Maldive Islands; littoral)____._______ flora (p. 333) f?. Cirri less stout, the longest segments from one-third to one-half again as broad as long; 19-39 (most commonly about 30) arms, 75-85 mm. long; cirri XXI-X XXIII, 25-34 (usually about 30) (Delagoa Bay)_-_-----_---------- delagoae (p. 334) e?. Cirri slender or moderately stout, with the longest segments as long as, or longer than, broad. fi. Larger, the 11-23 arms 100-150 mm. long; brachials in the middle and outer portion of the arms more or less wedge-shaped; P2; and P; composed of 25-33 segments. g'. Usually 35-40 cirrus segments; P; and P; with 27-33 segments; brachials in the middle and outer part of the arms distinctly wedge-shaped, with produced distal edges; distal border of radials usually with a row of small bead-like tubercles; 11-20 arms 100-150 mm. long (Malacca Strait; 0-73 meters) --singularis (p. 290) g’. Usually 27-80 cirrus segments; P, and Ps with 25 segments; brachials in the outer half of the arms short and almost oblong, with the distal edges not produced; radials concealed; 19-23 arms 120-125 mm. long (Bagamoyo and Zanzibar to Waxin and from Kurrachi to Persian Gulf; 0-88 [?89] meters) __ africana (p. 325) f?. Small, the 12-13 arms 70-85 mm. long; brachials in the middle and outer portions of the arms exceedingly short and discoidal; P2 and P; composed of 18-25 segments; cirri with 26-35 (usually 30-35) segments, 20-23 mm. long (southern Japan) schlegelii (p. 329) d?. The outer cirrus segments bear dorsal tubercles or short, broad, and usually blunt, spines which develop very gradually. el, Cirri tapering conspicuously distally so that the terminal third or fourth is only about half as broad in lateral view as the proximal portion; cirri almost smooth, with only slight and inconspicuous subterminal dorsal tubercles on the last 10-15 segments; distal cirrus segments always longer than broad, varying from slightly longer than broad to twice as long as broad; cirri long, 30-33 mm. (rather more than one-fourth the arm length); 19 arms 120 mm. long; brachials short; P; with 27 segments (off western vavas 24 meters) oo. esos eee a ne eee ee a astyanax (p. 284) e. Cirri tapering not at all, or only very slightly, distally, the distal segments being never longer than broad and usually markedly broader than long, and bearing well-developed dorsal processes. fi. Cirri with not more than 31 segments. g'. Brachials exceedingly short and discoidal, with more or less strongly produced ends; 20-31 cirrus segments; 11-18 arms (Ceylon and Bay of Bengal to Philip- pines 310-44 [784] imieters)/ 222 2= Sees See esos n ae bengalensis (p. 321) g?. Brachials longer, wedge-shaped, the distal ends not produced; 15-18 arms 65-120 mm. long (Philippines to Singapore and Amboina; 0-36 meters) __affinis (p. 285) f?. Cirri with 30-44 segments. g'. Cirri with all the segments broader than long. hi. Ps; as large as, or larger than, P3. 232 BULLETIN 82, UNITED STATES NATIONAL MUSEUM i!. Longest cirri with more than 40 segments, the outer with sharp dorsal spines; smaller, the arms up to 110 mm. in length (Maldive Islands and Ceylon and western and northern coasts of Bay of Bengal as far as Akyab, Burma; 0-62)[760] meters) 22> 2. 252 ee eee a eee eee reynaudi (p. 302) i, Longest cirri with less than 40 segments, the outer with dorsal tubercles or broad and usually blunt spines; larger, the arms 145-180 mm. long. j'. Brachials not excessively short, beyond the proximal fourth of the arms 3-4 times as broad as long; cirri stouter, the segments subequal, all markedly broader than long; dorsal surface of the ossicles of the division series and the brachials not conspicuously swollen (Kei Islands, Amboina and Singapore to Ceylon; 0-50 meters)_------.----- amboinae (p. 297) j. Brachials exceedingly short and discoidal, beyond the proximal fourth of the arms 6-8 times as broad as long with almost or quite parallel ends; cirri more slender, with 30-33 segments, the earlier segments nearly as long as broad. k'. Dorsal surface of the ossicles of the division series, and of the brachials to near the arm tips, swollen; profile of the proximal portion of the arms strongly and roundedly serrate, the serrations pointing distally; carination of the basal segments of the earlier pinnules very slight, almost vestigial; 19 arms 145 mm. long (Amboina; about 2 Meters) hise0 ae sees A eh ES EE Sa! sarae (p. 318) k?, Dorsal surface of the ossicles of the division series and brachials not swollen; profile of the proximal portion of the arms angularly serrate, the serrations pointing proximally; dorsal profile of arms beyond the proximal third smooth; carination of the basal segments of the earlier pinnules conspicuous, though not exaggerated; 10-16 arms 100-150 mm. long (Philippines; 38-44 meters)_____.__------- parilis (p. 319) }?, P; much smaller than P». i. Cirri with 32-36 segments; 14 arms about 160 mm. long; P2 with 30 segments (Red) Sea) ect stbe 3 sce ett al ee EB ee ater (p. 311) #,. Cirri with 39-43 segments; 20 arms 80 mm. long; P; with 21 segments (Zanzi- Paryss Sets ear ee as Be ye joubini (p. 312) g*. Cirri with the longest segments longer than, or at least as long as, broad. ht. Longest cirri with 36-39 segments. uw. Outer cirrus segments only slightly broader than long; P2 only very slightly longer than P, and resembling it, but with.a few less (25) segments which become more elongate distally; P3; the largest and longest pinnule, slightly stouter throughout than P2:, with 22 segments; 24 arms about 85 mm. long (Burma eastward to Philippines; 27-73 [?89] meters) _____ pulchra (p. 287) @, Outer cirrus segments about twice as broad as long; P; half again as long as P,, slender and flagellate distally, with 24 segments; Ps slightly smaller than P,, but much larger than P;; 19 arms 80 mm. long (Zanzibar) _ gravieri (p. 289) }?, Longest cirri with 31-36 segments. a', Outer cirrus segments from half again to twice as broad as long; P; and P; similar and of the same length, 13 mm. long with 26 segments; P; shorter, 9 mm. long with 19 segments; 16-25 arms 110 mm. long (Pedro shoal, off the' west coast (of India) 22 2223— 52 p ee ee compta (p. 293) #. Outer cirrus segments scarcely broader than long; P; 10 mm. long with 26-28 segments; P, longer, 12 mm. long with 27 segments, and somewhat stouter; P, similar to P; but very slightly larger and longer; 13-18 arms 65-130 mm. long (Madagascar).._._.....-------- madagascarensis (p. 295) A MONOGRAPH OF THE EXISTING CRINOIDS 233 ALTERNATIVE KEY TO THE SPECIES IN THE GENUS HETEROMETRA a’. Brachials distinctly wedge-shaped, the ends never quite parallel, those in the proximal portion of the arms (except for the first few) very obliquely wedge-shaped and sometimes triangular. b1. Outer cirrus segments smooth dorsally, with or without a slight median carination, but never with dorsal spines or tubercles; 25-36 (usually about 30) cirrus segments; 16-26 arms 60-145 mm. long (from Philippines to Andaman Islands; 0-80 meters) quinduplicava (p. 245) b%. Outer cirrus segments bearing spines or tubercles dorsally. cl. Cirri tapering conspicuously distally so that the terminal third or fourth is only about half as broad in lateral view as the proximal portion; cirri almost smooth, the last 10-15 segments with only slight and inconspicuous dorsal tubercles; distal cirrus segments always longer than broad, varying from slightly longer than broad to twice as long as broad. d', Enlarged lower pinnules smooth, the segments with unmodified distal ends; P3; the largest and longest pinnule, 19 mm. long with 27 segments; cirri 30-33 mm. long with 37-39 segments; 19 arms 120 mm. long (off western Java; 24 meters) -astyanax (p. 284) d?, Enlarged lower pinnules with the distal angles of the component segments slightly swollen and produced; P; the longest pinnule, 10.5 mm. long with 17 segments; cirri usually about 30 mm. long with usually 30-33 segments; 10-13 (usually 11-13) arms 90-120 mm. long (south of eastern Borneo; 82-88 meters)_propinqua (p. 277) c?. Cirri tapering not at all, or only very slightly, distally, the distal segments being never longer than broad and usually markedly broader than long, and bearing well developed dorsal processes. d', Enlarged proximal pinnules slender and smooth, their basal segments without carinate processes on the side toward the arm tip, and their outer segments with smooth and unmodified distal ends. el. Usually 20 arms, up to 150 mm, in length; enlarged lower pinnules very long and slender, composed of about 20 segments; cirri with 23-33 (usually 25-30) segments of which the outer bear long dorsal spines (Red Sea and eastward to Muscat; 0-18 (?222]bimeters) 22 e_ ee 8 eye ne ee ese ey eer savignii (p. 235) e?, Arms 38-48 in number, 105-115 mm. long; enlarged lower pinnules composed of 26-30 segments; cirri composed of 40-50 segments of which the outer bear promi- nent, though not especially long, dorsal spines (Queensland and Aru Islands; 0-22 [?36]}meters) i ro _ bab eof) ee Oe ee ears Oe nematodon (p. 241) d@. Enlarged proximal pinnules with the basal segments carinate on the side toward the arm tip, or with the outer segments bearing fine spines or other processes on the distal ends. el. Enlarged proximal pinnules smooth, a few of the basal segments bearing carinate processes on the side toward the arm tip. f'. Longest cirri with not more than 30 segments; 15-18 arms 65-120 mm. long (Phil- ippines to Singapore and Amboina; 0-86 meters) ------------- affinis (p. 285) f?. Longest cirri with 30-40 segments. g!. Cirri with all the segments broader than long. ht, P; as large as, or larger than, Po. i. Longest cirri with more than 40 segments, the outer with sharp dorsal spines; arms up to 110 mm. in length (Maldive Islands and Ceylon and western and northern coasts of Bay of Bengal as far as Akyab, Burma; 0-62 [?69] TNE FETS eee eee eee ae a Se Ropes ae reynaudi (p. 302) 72, Longest cirri with less than 40 segments, the outer with dorsal tubercles or broad and usually blunt spines; larger, the arms 145-180 mm. long (Kei Islands, Amboina and Singapore to Ceylon; 0-50 meters) amboinae (p. 297) h?. Ps much smaller than P, i, Cirri with 32-36 segments; 14 arms about 160 mm. long; P, with 30 segments (Red!'Sea) i) Sok ee a 22a te 2 oe 2 SEES Ia ESBS SSeS ater (p. 311) 208244—40——_16 234 BULLETIN 82, UNITED STATES NATIONAL MUSEUM ®. Cirri with 39-43 segments; 20 arms 80 mm. long; P; with 21 segments (Zan- sibat)sivte2 ae as eee a es ot de een ine joubini (p. 312) g?. Cirri with the longest segments longer than, or at least as long as, broad. i‘, Longest cirri with 36-39 segments. i, Outer cirrus segments only slightly broader than long. j'. P, only very slightly longer than P, and resembling it but with a few less (25) segments which become more elongate distally; Ps the largest and longest pinnule, slightly stouter throughout than P,, with 22 segments; 24 arms about 85 mm. long (from Burma eastward to the Philippines; 27—75 4 (780) umeters) a2 —-=- - 2 ee Shee ee pulchra (p. 287) jt. Ps much longer than P; with about the same number (33-34) of segments; Ps much shorter than Py, equaling P; in length, with 27 segments; 11-20 arms 100-150 mm. long (Malacca Strait; 0-73 meters) -singularis (p. 290) @, Outer cirrus segments about twice as broad as long; P2 half again as long as P,, slender and flagellate distally, with 24 segments; P, slightly smaller than P., but much larger than P;; 19 arms 80 mm, long (Zanzibar) _gravieri (p. 289) }?. Longest cirri with 31-36 segments. @, Outer cirrus segments from half again to twice as broad as long; P; and P, similar and of the same length, 13 mm. long with 26 segments; P; shorter, 9 mm. long with 19 segments; 16-25 arms 110 mm. long (Pedro shoal, off Wweat) coast / of india) 3255 -sa22 2b oe 35. tp ee Se eee compta (p. 293) #, Outer cirrus segments scarcely broader than long; P; 10 mm. long with 26-28 segments; P; longer, 12 mm. long with 27 segments, and somewhat stouter; P; similar to P, but very slightly larger and longer; 138-18 arms 65-130 mm. long (Madagascar) -_----.--------------- madagascarensis (p. 295) e?. Enlarged proximal pinnules with a serrate profile, the distal ends of the component segments armed with spines or bearing high keel-like processes on the outer ends of the prismatic ridges. J’. Enlarged proximal pinnules stout, composed of cylindrical or only very slightly prismatic segments which have everted and spinuos distal ends; disk more or less extensively plated; 11-28 arms 50-140 mm. long; cirri with 16-38 (usually 30-35) segments (Amboina and the Aru Islands; 0-20 meters)__--variipinna (p. 278) J*. Enlarged proximal pinnules sharply prismatic, with the distal portion of the prismatic ridges on each segment extended outward in the form of high thin rounded-triangular processes or broad blunt teeth. g'. Segments of the enlarged proximal pinnules with the distal portion of the prismatic ridges produced into a high thin rounded-triangular process, so that the profiles of these pinnules are very strongly serrate with conspicuous rounded teeth; division series and arm bases very rugged, the component ossicles swollen dor- sally, with prominent though rather broadly rounded synarthrial and articular tubercles; 10-30 (usually between 11 and 22) arms 70-150 mm. long; cirri very variable, with 25-46 (usually between 30 and 45) segments (from Philippines to Andamans [possibly Maldives] and southward to northern Australia; 0-111 meters) -..__- we cstasase tig an Saha le teh ne 8 crenulata (p. 253) g@. Enlarged lower pinnules rather slender, becoming flagellate distally, the basal seg- ments sharply carinate and the segments from the third outward with the entire distal edge somewhat produced and overlapping, the outer portion of the pris- matie ridges being produced into a blunt spine, which is directed obliquely for- ward; cirri slender, 15 mm. long with 23-26 segments; 10-13 arms 50 mm. long (Singapore to Maldive Islands; littoral)..._.........___- producta (p. 275) a*, Brachials beyond the proximal fourth of the arms exceedingly short, discoidal, with the ends parallel, the brachials within the proximal fourth of the arms also exceedingly short, with parallel or slightly oblique ends. b'. I11Br series present, all or mostly 4 (3-++-4); 18-27 arms (Andaman Islands to Kwala Cassan BUC aves CLONE) se ee nee ee philiberti (p. 313) A MONOGRAPH OF THE EXISTING CRINOIDS 235 3. III Br series absent, or if present all or mostly 2. cl. Longest proximal cirrus segments about as long as broad—sometimes slightly broader than long or slightly longer than broad. d', Longest cirri, in fully grown individuals, with more than 30 segments. e!. Larger, with 10-19 arms 100-150 mm. long. f!. Outer cirrus segments with small carinate dorsal spines which deyelop gradually; proximal portion of the arms with a distinctly serrate profile. g'. Profile of the proximal portion of the arms strongly and roundedly serrate, the ser- rations pointing distally; carination of the basal segments of the earlier pinnules very slight, almost vestigial; 19 arms 145 mm. long (Amboina; about 2 Meters). 745E weoealn MY. pace ee Age Bebe lia bgt hry en dh J sarae (p. 318) g?. Profile of the proximal portion of the arms angularly serrate, the serrations pointing proximally; dorsal profile of arms beyond the proximal third smooth ; carination of the basal segments of the earlier pinnules conspicuous, though not exaggerated; 10-16 arms 100-150 mm. long (Philippines; 38-44 meters) ____parilis (p. 319) f?. Outer cirrus segments with long and prominent dorsal spines, which begin abruptly; elements of the division series and earlier brachials with the dorsal surface unmodi- fied (Bagamoyo and Zanzibar to Waxin and from Kurrachi to Persian Gulf ; 0-88 [289] omieters) x sare |e bee hy aac IR wa lh gy Ran ev kt a africana (p. 325) e?, Small and delicate, with 12-13 arms 70-85 mm. long (southern Japan) -schlegelii (p. 329) d. Longest cirri with never more than 31, and usually much less than 30 segments; distal or proximal ends of the earlier brachials produced (Ceylon and Bay of Bengal to Philip- pines; 0-44 '[284]' meters) _ {3° Sept eee wey tes Yeas bengalensis (p. 321) ce’. All the cirrus segments much broader than long, the longest proximal segments being one- third to one-half again as broad as long. d'. Cirri very stout and strongly curved, with 29-30 segments; 10-13 arms 170-180 mm. longy (Maldive: Islands) -20~=-ss5eei45 20 PEAS Bl, pine AE Lig flora (p. 333) d. Cirri less stout and less strongly curved, with usually about 30 segments; 19-39 (most commonly about 30) arms 75-85 mm. long (Delagoa Bay)___________ delagoae (p. 334) HETEROMETRA SAVIGNII (J. Miiller) PuatE 21, Figures 82-84; Pirate 22, Ficures 85-88 [See also vol. 1, pt. 2, figs. 440, 441 (pinnule tips), p. 261.] Comatula multiradiata Aupoutn, in Savigny, Description de Egypte, 1817 (1826), p. 205, pl. 1, fig. I, 1-6 (Red Sea).—Leucxanr, Isis, vol. 5, 1839, p. 612 (myzostomes).—LeEucKart, Bruch- stiike III, Helminthol. Beitrage, Programm zur Eréffnung der Wintervorlesungen in Freiburg, 1842, pt. 1, p. 8 (myzostomes).—von Grarr, Das Genus Myzostoma, 1877, pp. 2, 22 (myzos- tomes); Challenger Reports, Zoology, vol. 10, pt. 27, 1884, p. 32 (myzostomes).—Bratn, Centralbl. fiir Bakteriol. und Parasitenk., vol. 3, 1888, p. 183 (myzostomes).—A. H. CLARK, Amer. Nat., vol. 43, 1909, p. 265 (recorded by Audouin from Red Sea in 1817; identified by J. Miller as [Heterometra] savignii); Proc. U. S. Nat. Mus., vol. 40, 1911, p. 1 (of Audouin, in Savigny; history of confusion by de Blainville with Comatula [Oligometrides] adeonae), p. 2 (identified as savignii by J. Miiller), p. 2 (of Leuckart; identified as savignii by P. H. Carpenter [in von Graff]); Crinoids of the Indian Ocean, 1912, p. 2 (identity). Comatula sp. Leuckanrt, Zeitschr. fiir organ. Physik, vol. 3, Heft 4, 1833, p. 387 (from Audouin, in Savigny). Comatula adeonae DE BuaINvILLE, Manuel d’actinologie, 1834 (1836), pl. 26, figs. 1-5 [but not p. 249] (after Audouin).—[ANonymovs], Penny encyclopedia, vol. 7, 1837, p. 391 (figure, but not description; from de Blainville).—{Knicut], Natural history, or Second division of the English encyclopedia, vol. 2, 1867, p. 99 (same). Alecto savignit J. Miitter, Monatsber. preuss. Akad. Wiss., 1841, p. 185 (description; Red Sea).— J. Mturmr, Archiv fiir Naturg., 1841, vol. 1, p. 144 (same); Abh. preuss. Akad. Wiss., 1841 (1843), p. 181.—A. H. Crank, Proc. U. 8. Nat. Mus., vol. 34, 1908, p. 453 (commonly emended to “savigny?”’); Amer. Nat., vol. 43, 1909, p. 254 (history); Proce. U. S. Nat. Mus., vol. 40, 1911, p. 2 (history; synonymy); vol. 43, 1912, p. 383 (identity). 236 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Comatula (Alecto) savignii J. Mituuer, Abh. preuss. Akad. Wiss., 1847 (1849), p. 257 (detailed description; Red Sea).—A. H. Cuarx, Crinoids of the Indian Ocean, 1912, p. 30 (identity). Comatula savignyi Dusarvin and Hup#, Histoire naturelle des zoophytes, Bchinodérmes, 1862, p. 203 (synonymy; description; coasts of Egypt).—von Marrens, Von der Decken’s Reise in Ost- Africa, vol. 3, 1869, p. 129 (Red Sea). Antedon savignii P. H. Carpenter, Trans. Linn. Soc. (Zool.), ser. 2, vol. 2, 1879, p. 23 (arm struc- ture), p. 29 (listed as an Antedon).—Bp tt, Proc. Zool. Soc. London, 1882, p. 533 (listed), p. 534 (specific formula) —P. H. Carpsnrer, Proc. Zool. Soc. London, 1882 (1883), pp. 740 ff. (discussion of Bell’s method of formulation and corrected formula). Antedon savignyi Lupwic, in Kossman, Reise nach dem Rothen Meer, vol. 5, Echinoderma, 1880, p. 4 (Red Sea; no definite locality).—von Grarr, Challenger Reports, Zoology, vol. 10, pt. 27, 1884, p. 13 (myzostomes).—P. H. Carprnrer, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, pp. 55, 252, 253, 255, 263, 366, 380 (discussion; Red Sea; Muscat [Kurrachee refers to H. africana]).—Hartiavus, Nova Acta Acad. German, vol. 58, No. 1, 1891, p. 17 (in key), p. 19 (notes on one of the cotypes), p. 47 (compared with A. klunzingeri [Lamprometra palmata}), pl. 2, fig. 20 (drawn from a cotype), p. 113 (in Géttingen Mus.).—Hamann, Bronns Klassen und Ord- nungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1582 (listed) —Cuapwicx, Journ. Linn. Soe. (Zool.), vol. 31, 1908, p. 47 (Suez Bay, 4 fathoms; Ul Shubuk, 9 fathoms; Khor Shinab, 10-12 fathoms; Salaka; characters; color in life; commensal ophiurans).—A. H. Cuark, Proce. U. S. Nat. Mus., vol. 40, 1911, p. 5 (recorded by Chadwick from the Red Sea; identity) ; Crinoids of the Indian Ocean, 1912, pp. 35, 41 (identity).—Boutenamr, Proc. Zool. Soc. London, 19138, pp. 88, 102 (Ul Shubuk, 9 fathoms; myzostomes). Himerometra savignii A. H. CuarK, Smithsonian Misc. Coll., vol. 50, 1907, p. 356 (listed). Heterometra savignit A, H. Cuarx, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 11 (listed); Amer. Nat., vol. 43, 1909, p. 254 (history; range), p. 255 (reported by Chadwick from Red Sea); Vid. Medd. Naturh. Foren. Kgbenhavn, 1909, p. 164 (compared with H. singularis); Proc. U.S. Nat. Mus., vol. 40, 1911, pp. 2, 5, 9, 14, 24 (history; synonymy; localities [record from Kurrachi is H. africana); Bull. Mus. Hist.Nat., Paris, 1911, No. 4, p. 251(Muscat) ; Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p. 26 (division series compared with those of Petasometra helianthoides).— A, H. Cuark, Proc. U.S. Nat. Mus., vol. 43, 1912, p. 381 (ecotype in U. S. N. M.), p. 383 (iden- tity; original specimens include 1 of Craspedometra [Heterometra] ater), p. 394 (localities; notes) ; Crinoids of the Indian Ocean, 1912, pp. 2, 30, 35, 41 (identity), p. 124 (synonymy; localities {specimens from Kurrachi and Straits of Ormuz, and the notes, refer to H. africana]); Smith- sonian Mise. Coll., vol. 61, No. 15, 1913, p. 27 (published reference to specimens in B. M.; localities; notes; specimens from Kurrachi are H. africana).—A. H. Cuarx, Unstalked crinoids of the Siboga-Exped., 1918, p. 76 (in key; range [except Persian Gulf and Kurrachi, which refer to H. africana]).—Gist6n, Kungl. Fysiogr. Siillsk. Handl., new ser., vol. 45, No. 11, 1934, p. 48.—A. H. Crark, John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1986, p. 99 (range), p. 104. Heterometra savignyii GisL£N, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, p. 22. Diagnostic features —The enlarged proximal pinnules are very long, slender, and perfectly smooth, without carinate processes on the earlier segments. There are usually 20 arms up to 150 mm. in length, which are composed of wedge-shaped bra- chials. The cirri have usually 25-30 segments of which the outer bear long and prominent dorsal spines. Description.—The centrodorsal is discoidal, moderately thick, with a flat dorsal pole 3 mm. in diameter. The cirrus sockets are arranged in two and a partial third alternating marginal rows. The cirri are XV-XXX (usually about XX), 23-35 (usually 26-33), 15-22 (usually about 20) mm. long. Dorsal spines, usually long, slender, and prominent, are developed from about the twelfth segment onward. There are no basal rays. A MONOGRAPH OF THE EXISTING CRINOIDS 237 The radials are usually even with the rim of the centrodorsal in the midradial line and visible as low triangles in the interradial angles. The IBr, are short, slightly over three times as broad as long, with the lateral borders nearly straight and parallel. They are united in the proximal half but are separated in the distal half by a moder- ately broad U-shaped gap. The [Bry (axillaries) are broadly pentagonal, twice as broad as long. There are usually 10 IIBr 4(3-+4) series present; exceptionally IIBr 2 series occur. III[Br 2 series are sometimes present. The division series are well rounded dorsally and well separated laterally, the sides perfectly smooth with no trace of lateral production. The 13-21 (usually 20) arms are 90-150 mm. long in fully developed individuals. The brachials are smooth. Syzygies occur between brachials 3+-4, again from between brachials 9+10 to between brachials 21-+22 (usually about brachials 19+20), and distally at intervals of 4-14 muscular articulations. Pp is stout basally, but tapers rapidly, becoming very slender distally. P, is larger and stouter than Pp and tapers less rapidly. P, is similar to P;. P, is some- times large and long, like the pinnules preceding. The lower pinnules are very long but slender and perfectly smooth. Notes.—Miiller’s original description, which was based on the specimens from the Red Sea collected by Hemprich and Ehrenberg, is as follows: The cirri are XXX, 24-29; the outermost 12-14 segments bear a process on the dorsal side. The radials are visible. The IIBr series are 4(3+4). In the next division the axillary has no syzygy. The 20 arms are up to 150 mm. in length. The second ossicle beyond the radial (IBr) axillary can be rocked from side to side on the first (which means that there is a synarthry between the first two ossicles beyond the IBr axillary). The brachials are smooth. The intersyzygial interval is 4 to 9 muscular articulations. The first pinnule is on the second segment after either the first or the second arm division. The second and third pinnules are the largest, and sometimes the fourth is also large. The first pinnule is thick basally but tapers rapidly and is not so long as the second. The skin of the disk is soft. The color in alcohol is yellowish brown. Miller gave the size as up to one foot, which would mean an arm length of 150 mm. Carpenter examined Miiller’s specimens in the Berlin Museum, and in his key to the species in the Savignyi group published in the Challenger report (1888) he added some features to those given by Miller. He noted that IIIBr 2 series may or may not be present. There are no lateral processes on the segments of the lower pinnules. The cirri are XXV—XXX, between 25 and 35, and the later cirrus segments have strong spines. The second syzygy is between about brachials 19+20. Pp is smaller than P,. Hartlaub gave an excellent figure of one of Miiller’s original specimens, which the museum at Géttingen received in exchange from the Berlin Museum. He noted that this species does not always have as many as 20 arms, and that a second specimen from Miiller’s material at Géttingen has not more than 13 arms. In one of Miiller’s specimens in the Berlin Museum that I examined the centro- dorsal is discoidal, moderately thick, with a flat dorsal pole 3 mm. in diameter. The cirrus sockets are arranged in two and a partial third alternating marginal rows. The 238 BULLETIN 82, UNITED STATES NATIONAL MUSEUM cirri are about 20 mm. long and composed of 23-25 segments; but the best-developed cirri are all broken off. Long and prominent dorsal spines are developed from the twelfth segment onward. No basal rays are visible. The radials are even with the rim of the centrodorsal in the midradial line, and are slightly visible as low triangles in the interradial angles. The IBr, are slightly over three times as broad as long with the lateral edges nearly straight and parallel, united in the proximal half but separated in the distal half by a moderately broad U-shaped gap. The IBr, are broadly pen- tagonal and twice as broad aslong. The division series are well rounded dorsally and well separated laterally, the sides being perfectly smooth with no trace of lateral production. There are 20 arms 125 mm. long. Ten I/Br4(3~+-4) series are present. The lower pinnules are very long but are slender and perfectly smooth. Two other specimens in the Berlin Museum examined by Miiller have 20 arms each, 10 IIBr 4(3+-4) series being present in both cases. They resemble the one just described. Carpenter’s note that II] Br 2 series may or may not be present was not based upon any specimens seen by Miiller, for all the specimens known to Miiller with the exception of the one mentioned by Hartlaub with 13 arms, the one figured by Hartlaub, which has 18 arms, and the type of H. ater, which Miller did not distinguish from H. savignii (see page 311) with 14 arms, had exactly 20 arms, and none of them had any II[Br series at all. The specimens possessing III Br series known to Carpenter were those in the British Museum from Kurrachi, which are representatives not of 7. savignii but of the quite different H. africana. Dujardin, in Dujardin and Hupé, in his account of Comatula savignyi did not follow his usual procedure of simply translating Miiller’s description but wrote one of his own. He said the expanse is 90 mm. There are 20 branches or tertiary arms carrying numerous pinnules of almost equal length, giving the arms a feathery ap- pearance. The cirri are XXX, 29-30; the last segment is clawed, and the 7 or 9 pre- ceding bear a small spine. The habitat is given as the coasts of Egypt. This de- scription was not drawn up from any specimens but from the original figures published by Savigny in 1817 and reproduced by de Blainville in 1836. While in the heading the species is credited to J. Miiller, no reference is given to his descriptions. The only references given are to Savigny’s plate, to Audouin’s explanation of this plate, and to de Blainyille’s description and plate of Comatula adeonae. Dujardin’s short description accords perfectly with these figures in the size, number of arms, number of cirri, and number of cirrus segments given, while the figures afford no indication of the presence of elongated proximal pinnules, nor of the number of the elements in the division series. The locality given is to be inferred from the inclusion of the species in Savigny’s work, and the identity of the species delineated by Savigny was made clear by Miiller in 1841 and again in 1849. Both the of specimens from the Gulf of Suez in the British Museum are small. The two specimens recorded by Chadwick from Suez Bay in 7 meters have 13 arms. In both,syzygies occur between brachials 3-+4 and 9+-10 and distally at irregu- lar intervals of 5 to 14 muscular articulations. The color in alcohol is purple with purplish white cirri and almost white pinnules. The single specimen recorded from Ul Shubuk by Chadwick had 20 arms, 10 I1Br 4(3+4) series being present. The position of the second syzygy varies from A MONOGRAPH OF THE EXISTING CRINOIDS 239 between brachials 17-++-18 to between brachials 21+22, and the distal intersyzygial interval is 11 to 13 muscular articulations. This specimen was serving as host to no less than 15 ophiurans (see Part 2, p. 645). The single specimen from Salaka has 20 arms. Like one of those from Khor Shinab its color when living was brown and white. Of the two specimens from Khor Shinab one has 14 and the other 18 arms. The second syzygy is in some cases between brachials 15+16, but usually it varies in po- sition from between brachials 18-+-19 to between brachials 20-++-21, and others follow at intervals of 8 to 10 muscular articulations. When living one was brown and white and the other was purple and white. In the specimen in the Paris Museum from Muscat the dorsal spines on the outer cirrus segments are very sharp. This suggests that this specimen may repre- sent H. africana instead of H. savignii. When I examined it I did not question Carpenter’s earlier determination. Remarks.—The chief characteristics of this species are the strongly rounded and well-separated postradial series and division series, which do not have produced margins; the very long and slender, perfectly smooth, and evenly tapering P, and P3; and the long and slender spines on the outer cirrus segments. Nearly all individuals have exactly 20 arms, 10 IIBr 4(3+-4) series being present. Localities.—Red Sea; Egyptian coast [Audouin, in Savigny, 1826; Leuckart, 1833, 1839, 1842; de Blainville, 1836; Anonymous, 1837; J. Miiller, 1841, 1849; Dujardin and Hupé, 1862; Knight, 1867; Ludwig, 1880; A. H. Clark, 1907, 1908, 1909, 1911, 1912, 1918]. Red Sea; Hemprich and Ehrenberg [Leuckart, 1839, 1842; J. Miiller, 1841, 1843, 1849; von Graff, 1877, 1884; Dujardin and Hupé, 1862; P. H. Carpenter, 1879, 1883, 1884, 1888; Ludwig, 1880; Bell, 1882; Braun, 1888; Hartlaub, 1891; A. H. Clark, 1907, 1908, 1909, 1911, 1912, 1918] (3, U.S.N.M., 35213; Berl. Mus., 1054, 1056). Red Sea; Count von der Decken [von Martens, 1869]. Gulf of Suez [A. H. Clark, 1912, 1913] (2, B.M.). Suez Bay; 7 meters [Chadwick, 1908; A. H. Clark, 1911]. Tor, Red Sea [A. H. Clark, 1911, 1912] (1, Mus. f. Meereskunde, Berlin). Ul Shubuk; 16 meters; muddy bottom [Chadwick, 1908; A. H. Clark, 1911, 1912; Boulenger, 1913]. Salaka; at the anchorage, among corals [Chadwick, 1908; A. H. Clark, 1911, 1912]. Khor Shinab; 18-22 meters [Chadwick, 1908; A. H. Clark, 1911, 1912] (1, U.S. N.M., 35081). Muscat; M. de Rousseau, 1841 [P. H. Carpenter, 1888; A. H. Clark, 1911, 1912] (1, P.M.). No locality [A. H. Clark, 1912] (1, Berl. M., 5371). Geographical range.—Red Sea and eastward to Muscat. Bathymetrical range.—From the shoreline down to 18 (?22) meters. Most of the records are from the shoreline or from very shallow water. History.—This species was first made known through the excellent figures pub- lished by M. Savigny in 1826 and identified in the explanation of the plate as Comatula multiradiata. 'The species represented was referred to in 1833 by Professor Leuckart 240 BULLETIN 82, UNITED STATES NATIONAL MUSEUM as Comatula sp.; but later, in 1839 and 1842, he accepted Audouin’s determination and called it Comatula multiradiata. In 1836 M. de Blainville reproduced Savigny’s figures but in some unexplained way used them as illustrating Lamarck’s Comatula adeonae (= Oligometrides adeonae), a 10-armed species. His reproduction of these figures, together with his identification of them, was incorporated in the Penny Encyclopedia in 1837, and in the Natural History, or Second Division of the English Encyclopedia, in 1867. Prof. Johannes Miiller was unable to identify the species figured by Savigny with any form known to him, so in 1841 he proposed the new name Alecto savignii for the animal represented by Savigny in figure 1 on plate 1. Miller gave no description, merely stating that it possessed 20 arms. Hemprich and Ehrenberg had brought back from the Red Sea a number of specimens of this species, and from them Professor Leuckart took some myzostomes, which he recorded in 1839 and 1842, calling the host Comatula multiradiata. His record was republished by von Graff in 1877 and 1844, and by Braun in 1888. Miiller studied the specimens that had been brought to Berlin by Hemprich and Ehrenberg and in 1849 published a description of them under the name of Comatula (Alecto) savignii, again referring Savigny’s plate 1, figure 1. Departing from his usual custom of simply translating Miiller’s descriptions, Dujardin, in Dujardin and Hupé, 1862, drew up an original description of Comatula savignyi J. Miiller, which was prepared not from any specimens but from Savigny’s figures as reproduced by de Blainville. Although he credited the species to Miller, the only references he gave were to Savigny’s plate, to Audouin’s explanation of the figures on the plate, and to de Blainville’s text and plate. He saw the curious error that de Blainville had made and placed his Comatula adeonae without comment in the synonymy of Comatula savignyy. Dujardin gave as the habitat of Comatula savignyi the coasts of Egypt, which was a natural inference from the inclusion of the species in Savigny’s work. In 1869 this species was mentioned as found in the Red Sea by Professor von Martens in his account of the animals collected on Count von der Decken’s journey in East Africa. Dr. P. H. Carpenter in 1879 included Antedon savignyi in the list of species that he was able definitely to place in the genus Antedon as he understood it. It was again mentioned as occurring in the Red Sea by Prof. Hubert Ludwig in 1880. In October 1882, Prof. F. Jeffrey Bell proposed a specific formula for this species, which was emended by Dr. P. H. Carpenter in April of the following year. In his discussion of Antedon adeonae in the Alert report published in 1884 Prof. F. Jeffrey Bell wrote: There is a curious error in connexion with this species which does not seem to have been noticed. Lamarck described it as “C. radiis pinnatis denis &e.:’’ de Blainville, while quoting Lamarck, refers also to his own figures in his ‘Atlas’ (pl. xxvi); in this reference he is followed by J. Miiller and by the editors of the second edition of Lamarck. The figures, however, when referred to are seen to be those of a species with twenty arms and with cirri nearer thirty than twenty. It is not perhaps necessary at this distance of time to waste time in inquiring what species it is that de Blainville has there figured. A MONOGRAPH OF THE EXISTING CRINOIDS 241 In the Challenger report published in 1888 Dr. P. H. Carpenter, who had examined in Berlin the specimens collected by Hemprich and Ehrenberg and described by Miller, and also other specimens in the British and Paris Museums, gaye some additional information regarding this species in the form of notes incorporated in the key to the species of the Savignyi group of Antedon. He said that in addition to being found in the Red Seait also occurs at Muscat and Kurrachi but is not known to extend farther eastward. The first locality was represented by a specimen in the Paris Museum and the second by specimens in the British Museum, although he did not mention this. The specimens from Kurrachi represent H. africana, and there is a possibility that the specimen from Muscat also represents that species. Dr. Clemens Hartlaub in 1891 gave an excellent figure of one of Miiller’s original specimens which had been received in exchange by the Gottingen Museum, and mentioned another of Miiller’s original specimens which was already in that Museum. In my first revision of the old genus Antedon published in 1907, this species was referred to the new genus Himerometra. Herbert C. Chadwick in 1908 recorded this species from Suez Bay, Ul Shubuk, and Khor Shinab, giving the essential characters of his specimens and the color in life, and mentioning the commensal ophiurans. In a review of the family Himerometridae published in 1909 I transferred this species to the new genus Heterometra. In a paper on the crinoids of the Paris Museum published in 1911 I recorded a specimen from Muscat and gave a partial description of it. This is the specimen upon which P. H. Carpenter based his mention of Muscat as a locality for the species in 1888. Ina paper on the crinoids of the coasts of Africa published in the same year Tor was added to the known localities for this species, on the basis of a specimen in the Museum fiir Meereskunde in Berlin. In my memoir on the crinoids of the Indian Ocean published in 1912 I recorded and described one specimen from Kurrachi and seven from the Straits of Ormuz and also gave a summary of all the known localities. Reexamination of these eight specimens disclosed the fact that the basal segments of the proximal pinnules bear high carinate processes, and further study has shown that they are representatives of H. africana. In a paper on the crinoids of the Berlin Museum published in the same year I described at considerable length three of Hemprich and Ehrenberg’s specimens from the Red Sea, one from Tor in the Museum fiir Meereskunde and one without locality. In 1913 in a paper on the crinoids of the British Museum I recorded two small specimens from the Gulf of Suez and described seven from Kurrachi. These last, which represent not H. savignii but H. africana, had previously been examined by P. H. Carpenter, and the mention of Kurrachi as a locality for this species in the Challenger report is based upon them. In the same year Boulenger described the myzostomes from the specimen from Ul Shubuk recorded in 1908 by Chadwick. HETEROMETRA NEMATODON (Hartlaub) Puate 28, Ficure 117 Antedon milberti (part) Brit, Report Zool. Coll. H. M.S. Alert, 1884, p. 156 (Port Molle, 12—20 fathoms).—A. H. CrarK, Smithsonian Misc. Coll., vol. 61, No. 15, 1913, p. 23 (identity). 242 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Antedon nematodon Harriaus, Nachr. Ges. Géttingen, May 1890, p. 185 (description; Bowen, Queensland).—Harriaus, Nova Acta Acad. German, vol. 58, No. 1, 1891, p. 19 (in key), p. 27 (detailed description and comparisons; Bowen), pl. 1, fig. 9—Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1582 (listed), pl. 12, fig. 6 (from Hart- laub).—A. H. Ciark, Vid. Medd. Nat. Foren. Kgbenhavn, 1909, p. 117 (identity); Mem. Aus- tralian Mus., vol. 4, 1911, p. 718 (identity); Smithsonian Misc. Coll., vol. 60, No. 10, 1912, p. 2 (identity); Crinoids of the Indian Ocean, 1912, p. 37 (identity). Himerometra nematodon A. H. Cuarx, Smithsonian Misc. Coll., vol. 50, 1907, p. 356 (listed). Heterometra nematodon A. H. Cuark, Die Fauna Siidwest-Australiens, vol. 3, Lief. 13, 1911, p. 441 (Australian tropical species occurring south to Port Molle), p. 443 (range on the east coast); Mem. Australian Mus., vol. 4, 1911, p. 718 (identity), p. 721 (occurs south to Bowen), p. 734 (in key), p. 768 (synonymy; characters; Australian record; Port Molle, Alert); Crinoids of the Indian Ocean, 1912, p. 37 (identity), p. 120 (synonymy; localities)—Rr1cHENSPERGER, Abh. Senck. naturf. Ges., vol. 35, Heft 1, 1913, p. 82 (Aru Islands), p. 101 (localities; notes).—A. H. Crarx, Unstalked crinoids of the Siboga-Exped., 1918, p. 76 (in key; range).—H. L. Cuarx, The echinoderm fauna of Torres Strait, 1921, p. 192 (distribution) —Gistin, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 27, 30. Amphimetra nematodon A. H. Cuark, Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p. 2 (identity), p. 15 (detailed description of the type; Bowen), p. 16 (compared with A. [Heterometra] phili- berti); vol. 61, No. 15, 1913, p. 23 (published reference to specimen in B. M.; Port Molle, 12—20 fathoms) ; Internat. Revue gesamt. Hydrobiol. und Hydrogr., 1915, pp. 224 ff. (detailed account of distribution in Australia). Diagnostic features—The enlarged proximal pinnules are slender and smooth, without carinate processes on the basal segments, and are composed of 26-30 segments. P;, Ps, and P; are of about the same size, or P, is the longest and P; the shortest, but P, is always larger than Pp. The cirri are 25-30 mm. long and are composed of 40-50 segments, which are subequal, the outer bearing prominent, though not especially long, dorsal spines. The 27-48 arms are 105-115 mm. long. Description.—The centrodorsal is a thick convex disk with the cirri arranged in two and a partial third irregular rows. The cirri are about XXX, 40-50, about 25 mm. long. The segments are approxi- mately similar, none of them being longer than broad. In the distal two-thirds, or at the least in the distal half, of the cirri the segments bear rather strong dorsal spines. The cirri are rather strongly compressed laterally at their ends. The radials are not visible. The IBr, are very short, and are entirely free laterally. The IBr, (axillaries) are almost triangular and are also very short. The IIBr series are 4 (3+-4), exceptionally 2. The IIIBr series are 2, but 4 (3+4) when following I1Br 2 series. The IVBr series are 2, but only a few are present. The elements of the I1Br and subsequent series are, when compared with the elements of the IBr series, rather long. The first ossicles following an axillary are almost entirely united interiorly. The 38 arms are more than 80 mm. long. They are composed of short discoidal brachials the distal ends of which are rather strongly produced, and which in the proximal third of the arms overlap somewhat laterally. Only the basal brachials have smooth ends. The arms are somewhat compressed with a narrow dorsal surface. The first syzygy occurs between brachials 3-+-4, and the second from brachials 23-424 onward, often about brachials 31+32, but also further on, about brachials 41+42. The distal intersyzygial interval is 13 to 21 muscular articulations. A MONOGRAPH OF THE EXISTING CRINOIDS 243 In general the proximal pinnules are of delicate structure and are composed of smooth cylindrical segments. P, is about 12 mm. long and is composed of about 25 segments of which the basal are moderately thick, those in the distal half are very slender and become toward the end of the pinnule somewhat longer than broad. The segments are dark in the middle with light-colored ends. P; is similar in appear- ance, about 14 mm. long, and is composed of about 30 segments. It tapers somewhat more gradually than P,. P, is very small. The pinnules following P, decrease rather abruptly in length to about the thirteenth brachial (P;) and then gradually become longer again, reaching a length of about 10 mm. The disk is about 15 mm. in diameter and is deeply incised. In the general dark color of the animal the sacculi are not conspicuous. The color is blackish brown. Notes.—The preceding description is that given by Hartlaub. My notes on the type specimen are as follows: Two cirri still attached to the centrodorsal have 42 and 43 segments of which the twelfth and following bear dorsal spines. The longest cirrus segments are not quite so long as broad. In the outer two-thirds, or at least one-half, of the cirri the segments are twice as broad as long. The dorsal spines are rather small and slender, though sharp and prominent. The division series are narrow and rather widely separated. There are 38 arms about 105 mm. long. The brachials are extremely short and discoidal, with produced distal edges. The enlarged proximal pinnules are comparatively slender, and are smooth, not carinate. I remarked that this species appears to be most closely related to H. philiberti but that it is very easily distinguished from that form by the absence of carination on the proximal pinnules and by having the IIIBr series 2 instead of 4 (3+-4). The specimen dredged by the Alert at Port Molle is small. Reichensperger says that the specimen from Dr. H. Merton’s station 11 in the Aru Islands agrees almost completely with Hartlaub’s description, but it is somewhat larger than the type. The cirri are XX XIII, 40, up to 30 mm. long. All the cirrus segments are longer than broad. Toward the end the cirri become laterally compressed, and from the twelfth to the eighteenth onward the segments bear sharp and prominent dorsal spines. The radials are concealed. The IBr, are free laterally. In the division series the strong preponderance of series with two elements is characteristic. He gives two examples of arm division. On one of these there are two IIBr 4 (8+4) series; one of the latter bears two II]Br 2 series and a single IVBr 2 series, while the other bears a IIIBr 2 series externally and a IIIBr 4 (3+4) series internally, the latter carrying two IVBr 2 series. In the other example there are two IIBr 4 (8+-4) series and four IlIBr 2 series, beyond which there are four IVBr series, the two outer 4 (3+4) and the two inner 2. The 48 arms are 115 mm. long. The first ossicles following each axillary are almost entirely united laterally the result being that the arms are rather close together and often are almost parallel to each other. The dorsal arm sculpture is as figured by Hartlaub. P, is 13 to 14 mm. long and composed of 26 segments of which the first eight are rather thick and markedly broader than long. From the ninth segment onward the pinnule is more flagellate, and finally very slender with segments that are longer than broad. P,, P2, and P; are of similar structure but are stouter in the distal portion and not so flagellate. P, and P; decrease rapidly in length and in number of segments. P,is only 5mm.long. The disk is 18 mm. in diameter. 244 BULLETIN 82, UNITED STATES NATIONAL MUSEUM The specimen from Dr. Merton’s station 10 is small. The cirri are XXIII, up to 42, relatively very long, 19 to 23 mm. in length. The first two segments are broader than long, the third and fourth are about as long as broad, those following about as far as the twelfth are very markedly longer than broad, and from that point onward the segments are broader than long. The spines on the distal segments are sharp and prominent, and the cirri are strongly compressed laterally in the distal half. The 27 arms are only about 38 mm. long. All the division series following the IIBr series except for two are2. Py is 11 mm. long and is composed of 22 segments of which the proximal are relatively less broad than in the specimen from station 11. P, is about 12 mm. long and composed of about 28 segments. P, is markedly shorter, while P, is scarcely half as long as P, and is composed of about 12 segments. The disk is 10 mm. in diameter. The calyx and arms are dorsally light flesh color (in alcohol grayish white), the pinnules centrally being more or less violet; the cirri are dark violet. Remarks.—Reichensperger said that because only two specimens of this fine species were previously known he had given a rather detailed account of some of the features of his two specimens with particular reference to the great difference in size between them. Of the smaller specimen he said that there is no possibility of doubt regarding the correctness of the specific determination because of the general habitus, the division series, the structure of the arms and other features. He noted that this individual shows very convincingly how little weight may at times be attributed to the relative length of the segments of the cirri and lower pinnules. He remarked that he was under the impression that in the crinoids increase in length often out- stripped, as it were, increase in width, and that as a result during development the relative proportions of the segments, so far as regards length and breadth, reversed themselves as we not rarely find to be the case with other features. Naturally, there- fore, in the crinoids the systematic worth of characters of this sort, which heretofore have often been very highly regarded, must be discounted, and they can only be taken into consideration in a relative way. We know far too little about the growth changes in the comatulids, as apart from variations due to other causes, to be able to say anything definite one way or another at the present time. But Dr. Reichensperger has raised an interesting point and one worthy of intensive study. The species of Heterometra are among the most puzzling of all the comatulids, and features that in other genera we believe may be confidently relied upon, such as the character of the cirri and the relative proportions of their segments and the rela- tive proportions of the segments in the proximal pinnules, in some species of Hetero- metra, at least, are wholly unreliable. This is well illustrated by H. crenulata, which seems to behave quite as described by Dr. Reichensperger for H. nematodon, and furthermore seems to be most unstable in the same region—the Aru Islands. Localities—Bowen, Queensland [Hartlaub, 1890, 1891; A. H. Clark, 1907, 1909, 1911, 1912, 1915, 1918; Reichensperger, 1913; H. L. Clark, 1921] (1, H. M.). Alert; Port Molle, Queensland; 22-36 meters [Bell, 1883; A. H. Clark, 1913] (1, B. M.). Aru Islands; Dr. H. Merton’s station 10; north of Penambulai; 8 meters; stony bottom; April 2, 1908 [Reichensperger, 1913]. A MONOGRAPH OF THE EXISTING CRINOIDS 245 Aru Islands; Dr. H. Merton’s station 11; off Pulu Bambu; 10 meters; rocky bottom, with sand and coral; April 3, 1908 [Reichensperger, 1913]. Geographical range.—Aru Islands and Queensland south to Port Molle. Bathymetrical range.—From the shoreline down to 22 (?36) meters. History.—The first known specimen of this species was a small individual col- lected by the Alert. At the time the Alert collections were studied it was not dis- tinguished from the specimens of Amphimetra discoidea dredged at the same time and was recorded with these under the name of Antedon milberti. Dr. Clemens Hartlaub in 1890 described this species as Antedon nematodon from a specimen from Bowen, Queensland, in the Hamburg Museum, redescribing and figur- ing it in the following year. When the genus Antedon was first revised by me in 1907 nematodon was assigned to the new genus Himerometra. In my memoir on the recent crinoids of Australia and also in my report on the crinoids collected by the Hamburg Southwest Australian Expedition, both of which were published in 1911, nematodon was transferred to the genus Heterometra. In both of these memoirs Port Molle was given as a locality for the species, and in the former this record was said to be based upon a small specimen which had been col- lected by the Alert. In my monograph on the crinoids of the Indian Ocean published in 1912, I gave Heterometra nematodon as occurring at Bowen and Port Molle and said that at the British Museum I had found, in a large jar full of specimens of Amphi- metra discoidea (labeled Antedon milberti), a small example of this species dredged by the Alert at Port Molle in 12 to 20 fathoms. In a paper on the crinoids of the Hamburg Museum published in 1912 notes were given on the type specimen, which the author had examined in 1910, and in 1913 in a paper on the crinoids in the British Museum the specimen from Port Molle was formally recorded. In both of these papers the species was assigned to the genus Amphimetra under the heading Amphimetra nematodon. In 1913 Dr. August Reichensperger recorded, as Heterometra nematodon, two speci- mens collected by Dr. H. Merton in the Aru Islands in 1908. He gave extensive notes on these and also discussed certain peculiarities of the species as illustrated by them. In the report on the unstalked crinoids of the Siboga expedition published in 1918, nematodon was included in the key to the species of Heterometra but was not further mentioned as it was not collected by the Siboga. HETEROMETRA QUINDUPLICAVA (P. H. Carpenter) Puiate 26, Figure 107; Puare 34, Figures 149-151 [See also vol. 1, pt. 1, figs. 185, 186 (centrodorsal), p. 235; fig. 253 (centrodorsal), p. 253; fig. 470 (centrodorsal), p. 359.] Antedon sp. P. H. CarPENTER, Quart. Journ. Geol. Soc., Feb. 1880, p. 41, footnote, pp. 42, 43 (Chal- lenger station 212); Challenger Reports, Zoology, vol. 11, pt. 32, 1884, p. 51, 9th line. Antedon clemens P. H. Carrenter, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 229 (descrip- tion; Challenger station 212), pl. 39, fig. 5—Harrtaus, Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 37 (in key). Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1581 (listed).—A. H. Cuarx, Proc. U. S. Nat. Mus., vol. 34, 1908, p. 273 (ab- normal ray of type in its relation to arm structure of Hudiocrinus); vol. 39, 1911, p. 541 (same as A. anceps P. H. Carpenter and as Craspedometra aliena A. H. Clark [referring to specimens 246 BULLETIN 82, UNITED STATES NATIONAL MUSEUM other than the type, which is Heterometra crenulata]) ; Crinoids of the Indian Ocean, 1912, p. 34 (identity); Unstalked crinoids of the Siboga-Exped., 1918, p. 76, footnote 1 (identity). Antedon anceps P. H. Carpenter, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 254 (de- scription; Challenger station 212), pl. 35, figs. 1-3—Hanrrtavus, Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 17 (in key).——Kopuier, Mém. Soe. Zool. France, vol. 8, 1895, p. 420 (Sunda Islands; Biliton; notes) —Munexert, Archiv fiir Naturg., Jahrg. 71, 1905, vol. 1, Heft 1, p. 212 (syzygies; regeneration).—HamaAnn, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, pp. 1579, 1582 (listed). —A. H. Cuark, Proc. U. 8. Nat. Mus., vol. 39, 1911, p. 541 (same as A. clemens P. H. Carpenter and as Craspedometra aliena A. H. Clark [referring to specimens other than the type, which is Heterometra crenulata}) ; Crinoids of the Indian Ocean, 1912, pp. 33, 35 (of P. H. Carpenter, 1888=Craspedometra anceps), p. 40 (of Chadwick, 1904= Heterometra bengalensis), p. 74 (B. M., MS.=Capillaster multiradiata); Smithsonian Mise. Coll., vol. 61, No. 15, 1913, p. 80 (B. M., MS.=Capillaster multiradiata) ; Unstalked crinoids of the Siboga-Exped., 1918, p. 76, footnote 1 (synonym of quinduplicava). Antedon quinduplicava P. H. CArrenTER, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 262 (description; Challenger Station 212), pl. 4, figs. 1, a-d, pl. 47, figs. 4,5 —Harriavus, Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 18 (in key) Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1582 (listed) —A. H. Cuarx, Crinoids of the Indian Ocean, 1912, p. 35 (identity).—Guistin, Arkiv fér Zool., vol. 19, No. 32, 1928, p. 4, No. 17 (Challenger station 212; notes). Himerometra anceps A. H. Cuarxk, Smithsonian Mise. Coll., vol. 50, 1907, p. 356 (listed); Smith- sonian Mise. Coll., vol. 52, 1908, p. 214 (Albatross station 5147; also Philippine Islands). Himerometra clemens A. H. Cuarx, Smithsonian Misc. Coll., vol. 50, pt. 3, 1907, p. 356 (listed). Himerometra quinduplicava A. H. Cuarx, Smithsonian Mise. Coll., vol. 50, 1907, p. 356 (listed); vol. 52, 1908, p. 214 (Albatross station 5139; also Philippine Islands). Amphimetra anceps A. H. Cuark, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 7 (listed); Smith- sonian Mise. Coll., vol. 61, No. 15, 1913, p. 23 (published references to specimens in B. M.; Challenger station 212; notes).—GuistiN, Arkiv fdr Zool., vol. 19, No. 32, February 20, 1928, p. 4, No. 17 (Challenger station 212; notes); Kungl. Fysiogr. Sillsk. Handl., new ser., vol. 45, No. 11, 1934, p. 16. Craspedometra aliena (part) A. H. Ciark, Proc. U. 8S. Nat. Mus., vol. 37, 1909, p. 31 (Albatross station 5157; specimens other than the type, which is Heterometra crenulata). Heterometra quinduplicava A. H. Cuark, Proc. Biol Soc. Washington, vol. 22, 1909, p. 11 (listed); Vid. Medd. Nat. Foren. Kgbenhayn, 1909, p. 162 (Singapore; description of a specimen), p. 163 (compared with H. aspera), p. 193 (collected at Singapore by Svend Gad); Crinoids of the Indian Ocean, 1912, p, 35 (identity), p. 180 (synonymy; range); Smithsonian Mise. Coll., vol. 61, No. 15, 1913, p. 27 (published reference to specimen in B. M.; Challenger station 212); Unstalked crinoids of the Siboga-Exped., 1918, p. 76 (in key; range), p. 81 (synonymy; notes; station 99), p. 272 (listed) —GuisLin, Arkiv for Zool., vol. 19, No. 32, 1928, p. 4 (notes); A. H. Crarx, Treubia, vol. 14, livr. 2, 1933, p. 212 (north of central Java; notes).—GisLEN, Kungl. Fysiogr. Siillsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 22, 27. Craspedometra anceps A. H. Cuark, Proc. U. S. Nat. Mus., vol. 39, 1911, p. 541 (Albatross station 5355; clemens a synonym; aliena also a synonym [but not type specimen, which is Heterometra crenulata]); Crinoids of the Indian Ocean, 1912, pp. 33, 35 (=anceps), p. 34 (=clemens), p. 120 (synonymy; includes aliena; discussion of specimens upon which names have been based), p. 316 (Andaman Islands).—F. W. CLarke and WHEELER, U.S. Geol. Surv. Prof. Pap. 90-D, 1914, pp. 34 ff. (inorganic constituents of the skeleton); Prof. Pap. 102, 1917, pp. 20 ff. (same); Prof. Pap. 124, 1922, p. 17 (same). Diagnostic features.—The cirri have 25-36 (usually about 30) segments, which are subequal, usually about as long as broad; the outer are more or less carinate dorsally but never bear dorsal spines or tubercles. The enlarged proximal pinnules are slender and smooth, with the basal segments more or less abruptly rounded, sharpened, or obscurely carinate on the side toward the arm tip. The 10-26 A MONOGRAPH OF THE EXISTING CRINOIDS 247 (usually 15-20) arms, which are 60-145 mm. long, are composed of very short brachials. Description.—The centrodorsal is discoidal, moderately thick, with the bare dorsal pole flat, 4 mm. in diameter. The cirrus sockets are arranged in one and a partial second irregular and closely crowded marginal rows. The cirri are XXI, 32-36, 30 mm. long, moderately slender. The first segment is very short, and those following gradually increase in length to from the eighth to the thirteenth, which, with those following, is about as long as broad. The cirri decrease almost imperceptibly in width from the base outward and more rapidly on the last six or eight segments. On the terminal four to six segments the dorsal surface is slightly swollen so that the dorsal profile is slightly convex, the highest point in the convexity of the antepenultimate segment being beyond the middle. The middorsal line of the last five segments is more or less sharply carinate. The opposing spine is small, terminal, with the dorsal profile usually rising in a straight line from the proximal end of the segment; but occasionally the apex is slightly produced. The radials are concealed by the centrodorsal except for their distal angles, which are visible interradially as low triangles with the produced apices reaching about halfway to the midradial line. The IBr, are very short, six to eight times as broad as long, slightly incised in the midradial portion of the distal edge. The IBr. (axil- laries) are short, two to three times as broad as long, triangular with the lateral angles truncated, forming sides that are slightly shorter than the lateral borders of the [Br,. The distal borders are nearly straight and the distal angle is somewhat greater than a right angle. The lateral borders of the elements of the division series are slightly produced and are in lateral contact. There are present eight I[Br 4 (3+4) series and eight I][Br 2 series. The 26 arms are 145 mm. long. The brachials beyond the proximal fourth of the arm are exceedingly short, four or five times as broad as long, and practically oblong. P, is about 12 mm. long, with 25 segments. It tapers from the third to the tenth or eleventh segment and is very slender and delicate from that point onward. The basal segments are broader than long, the eighth segment is about as long as broad, and the outer are about half again as long as broad. The second-seventh segments have a prominent, though narrow, carinate process the crest of which is straight and parallel with the axis of the segment. This is highest on the fourth segment and slowly decreases in height distally. The outer surface of the earlier segments is swollen, this swelling gradually narrowing and in the outer half of the pinnule becoming a sharp gablelike ridge. On arms arising from a IIIBr series P; is 15 mm. long, with 33 segments, and tapers much more gradually, thus more nearly resembling P;. P»: is about 20 mm. long, with 37 segments, tapering gradually and regularly to a delicate and flagellate tip. The basal segments are broader than long, the eighth is about as long as broad, and those following slowly increase in length so that the terminal are about twice as long as broad. The second-seventh segments have narrow carinate processes with a straight crest which are higbest on the fourth and fifth; these give the pinnule the appearance of being broadest at this point and slowly tapering both toward the base 248 BULLETIN 82, UNITED STATES NATIONAL MUSEUM and distally. A low crest runs along the outer surface of the pinnule which is at first broad and well rounded but becomes sharp after the eighth segment. P; is about 25 mm. long, with 35 or more segments, resembling P, and slightly, if at all, stouter basally but tapering more gradually so that it is considerably stouter in the outer portion. P, is 19 mm. long, with 26 segments, resembling P, but slightly less slender and with longer segments distally. P; is 13 mm. long, with 26 segments, and is less stout basally than the preceding pinnules. The preceding description is based upon a magnificent specimen from the Danish Expedition to the Kei Islands station 72, which was originally described as a new species under the name of Heterometra apollo but which does not appear to differ in any essential feature from the specimens of H. quinduplicava previously known. Notes.—The specimen from Albatross station 5355 is small with 18 arms about 80 mm. long and cirri from 20 to 22 mm. long. Carpenter described Antedon quinduplicava from Challenger station 212 as follows: The centrodorsal is a thin disk with sloping sides. The cirri are about XVIII, 30. The segments are tolerably equal; the last few are rather compressed and faintly carinate, and the penultimate bears a slight opposing spine. The radials are just visible. The IBr, are rather closely united, forming a median prominence (synarthrial tubercle) with the pentagonal axillaries, and there is a similar but less marked prominence on the first two segments beyond the axillary. The I[Br series are 4(3+-4), and the IIIBr series, when present, are 2. The 16 arms are about 80 mm. long. They are composed of 150 or more smooth brachials of which all but the terminal are broader than long. The lower brachials are subtriangular and the later ones are quadrate or almost oblong. Syzygies occur between brachials 3-+-4, again from between brachials 9+-10 to between brachials 15+16, and distally at intervals of 5 to 11 (usually 8 or 9) muscular articu- lations. P, is about equal to P;. P, is smaller again, while P, and P, are both longer and stouter, reaching 11 mm., and are composed of about 18 smooth segments most of which are longer than broad, and the lower are carinate. The pinnules of the next pair are generally smaller again. But in arms arising from a [Br axillary P; and occasion- ally P, may be large like those immediately preceding. The disk is 7 mm. in diameter, naked and much incised. Sacculi are abundant. The color in alcohol is brownish white with the perisome darker. Carpenter said that he had some doubts as to the propriety of separating this form from Antedon anceps, which was found at the same station. The general characters of the cirri, calyx, and large lower pinnules are the same in both types. One individual of Antedon anceps has only 10 arms, but another has three and a third four IJBr series. The outer parts of the arms are rather serrate, and P, is distinctly smaller than P,. On the other hand, the two individuals he referred to quinduplicava each have I11Br series, nearly smooth arms with longer intersyzygial intervals, and P, of about the same size as P,. Carpenter remarked that, considering the remarkable series of variations in Heterometra crenulata, he thought it quite possible that we are here dealing with another case of the same kind, but in the absence of the necessary intermediate links he preferred to keep quinduplicava separate from anceps for the time being. He said that the only species which approaches quinduplicava and anceps at all closely is A MONOGRAPH OF THE EXISTING CRINOIDS 249 H. savignii in which IIIBr series may or may not be present. But its more numerous and spiny cirri readily distinguish it from them both. One of the two individuals included in Antedon quinduplicava by Carpenter was a mere fragment, which had lost its cirri, disk, and most of its arms. As it was prac- tically useless in this condition, Carpenter made a preparation of its calyx (see beyond, page 250). I examined the type specimen of Carpenter’s Antedon quinduplicava in the British Museum and found it to resemble those collected by the Albatross in the Philippines. The outer cirrus segments are slightly carinate dorsally. Carpenter’s description of Antedon anceps from Challenger station 212 is as follows: The centrodorsal is a low convex disk. The cirri are about XX, 25-35. The segments are tolerably uniform; a few of them are longer than broad, and the later are slightly cearinate. The radials are partially visible. The IBr, are short and partly united, and there is a more or less distinct tubercle (a synarthrial tubercle) in the middle of their junction with the widely pentagonal IBr (axillaries). The I[Br series, when present, are 4(3-+4). The 10—14 arms are 85 mm. long. They consist of about 150 brachials of which the earlier are triangular and much broader than long, their successors be- coming more quadrate and finally almost oblong with a slight tendency to overlap. Syzygies occur between brachials 3+4, from between brachials 9+-10 to between brachials 13-14, and distally at intervals of from 3 to 10 (usually from 5 to 8) muscular articulations. The lowest pinnule, whether P, or P,, is considerably smaller than its successor on the same side. In arms that spring direct from the [Br axillary the largest pinnules are P,, P;, and P., which may reach 12 mm. in length and consist of 20 smooth segments most of which are longer than broad, the later ones being carinate. On the inner arm borne by a IIBr axillary the largest pinnules are P, and P,, though on the outer arms these are little, if at all, larger than P;. But P, is always small. The disk is naked and much incised. Sacculi are very abundant on the disk, arms, and pinnules. The color in alcohol is white with patches or bands of a faded purple, and the perisome darker. Gislén examined the type specimen of anceps in the British Museum in 1925. He said that it is a very curious individual. Two of the postradial series are either unbranched or forked after the IBr,._ He remarked that it is very interesting to notice that P, in both cases occurs to the left, just as in Hudiocrinus. Antedon clemens from Challenger station 212 was described by Carpenter as fol- lows: The centrodorsal is hemispherical. The cirri are about XXV, about 30. The segments are tolerably uniform and smooth, the penultimate with a small spine. The radials are not visible. The IBr, are slightly united laterally and the IBr, (axillaries) are pentagonal. In the single specimen included by Carpenter under this name one postradial series does not divide at all, three possess IBr series, and one has a I]Br 2 series. The 11 arms are about 75 mm. long. The arms are smooth, composed of tri- angular brachials, which are much broader than long, the brachials later becoming quadrate. Syzygies occur between brachials 3-++-4, again from between brachials 9+ 10 to between brachials 13-++14, and distally at intervals of 2-10 (usually 5-7) muscular articulations. The first pinnule is 5 mm. long with about 20 segments, which diminish greatly in size after the first five or six. The next pair (P2 and P,) are much longer with a 208244—40-—17 250 BULLETIN 82, UNITED STATES NATIONAL MUSEUM smaller number of stouter segments, several of which are considerably longer than broad. The third pair (Ps and P,) are smaller again. The more distal pinnules of each pair (that is, those on the inner side of the arm) are smaller than the more proxi- mal ones (on the outer side of the arm). The pinnules of the eighth and following brachials have slight dorsal keels on their lower segments. The disk is 7 mm. in diameter, naked and rather incised. Sacculi are abundant on the pinnules. Carpenter said that on the undivided postradial series there is a pinnule on the second ossicle, but none on the next. It is clear therefore that a [Br series is present, but the IBr, is not axillary, as in the arms of the species of Hudiocrinus (see page 143). One postradial series does not divide until the fourth ossicle beyond the radial, the axillary being a “syzygial joint,’ and the second and third segments bear pinnules. One of the [Br series is thus 5(4+5). In Carpenter’s description the division of the postradial series as described would give 10 arms, but the number of arms is said to be 11. The mention in the discussion of the species of this aberrant [Br series cor- rects this discrepancy. I examined the type specimen of Antedon clemens in the British Museum in 1910 and found that it most certainly represents the same species as the type specimen of A. anceps. The specimen from Siboga station 99 has 26 arms 80 mm. long. Two of the arms arise directly from IBr axillaries. There are eight I[Br 4(3+-4) series each of which bears internally a IlIBr 2 series. The cirri are XVIII, 27-30, 22 mm. long. This example exactly resembles one from Albatross station 5139. The specimen from north of central Java is small. It has 16 arms and is under- going adolescent autotomy. Of the two specimens recorded by Professor Koehler from Biliton, which were in very bad condition, one has 12 and the other 13 arms. The specimen from the Danish Expedition to the Kei Islands station 103 is smaller than that upon which the description of the species as given above is based (from station 72); it has 23 arms 110mm.long. Of the nine IIBr series present, eight are 2 and one is 4(3+4). Three IIIBr series are present, of which two, both on the same IIBr series, are of a single ossicle only, and oneis2. There is a single [VBr 2 series, developed internally on a II]Br 1 series. The cirri are XX, 29-31, from 20 to 25 mm. long. The two specimens from Singapore have the arms about 60 mm. long. One has 21 arms, resulting from the presence of six I/Br 4(3-++4) series and five IIIBr 2 series, the latter all developed internally in 1, 2, 2, 1 order. The other specimen is similar. These specimens agree in every way with others at hand from the Philippine Islands. P,, though larger than usual in this genus, is nevertheless smaller than P,. The specimen from the Andaman Islands is immature with 14 arms about 100 mim. long. Radials.—Carpenter found, by dissecting one of the two specimens that he referred to Antedon quinduplicava, that each of the radial areas on the ventral surface of the centrodorsal is marked at its proximal end by a large bilobate pit, so that every two pits are separated by an interradial ridge. These pits seem to be nothing but an unusual development of the radial pits that occur around the lip of the centrodorsal in so many comatulids and receive the lower ends of the axial radial canals, and so A MONOGRAPH OF THE EXISTING CRINOIDS 251 in fact they are. But their capacity is increased by the presence of corresponding pits on the under surface of the radial pentagon into which the axial canals, contained between the inner faces of the radials and the spouts of the rosette, open directly. Carpenter said that the only comatulid in which he had found any large cavity of a similar kind within the calyx is Comatula solaris. In Comatula solaris the radial axial canal, though it terminates blindly at the top of the centrodorsal, communicates with a large cavity in the lowest part of the radial, while in H. quinduplicava this cavity is outside and below the radial, between it and the centrodorsal. Localities.— Albatross station 5355; North Balabac Strait; Balabac Light bearing S. 61° W., 16.6 miles distant (lat. 8°08’10’’ N., long. 117°19’15’” E.); 80 meters; coral and sand; January 5, 1909 [A. H. Clark, 1911, 1912, 1918] (1, U. S. N. M., 35104). Challenger station 212; Philippine Islands, off Zamboanga (lat. 6°54’ N., long. 122°18’ E.); 18 meters; sand; January 30, 1875 [P. H. Carpenter, 1880, 1884, 1888; Hartlaub, 1891; A. H. Clark, 1907, 1908, 1909, 1911, 1912, 1918; Reichensperger, 1913] (5, B. M.). Siboga station 99; anchorage off North Ubian (lat. 6°07’30’’ N., long. 120°26/00’’ E.); 16-23 meters; lithothamnion bottom; June 28-30, 1899 [A. H. Clark, 1918] (1, Amsterdam Mus.). Albatross station 5139; in the vicinity of Jolo (Sulu); Jolo Light bearing S. 51° W., 3.6 miles distant (lat. 6°06’00’’ N., long. 121°02’30’’ E.); 36 meters; coral sand; February 14, 1908 [A. H. Clark, 1908, 1912, 1918] (1, U. S. N. M., 35183). Albatross station 5147; Jolo (Sulu) archipelago; in the vicinity of Siasi; Sulade Island (E.) bearing N. 3° E., 8.4 miles distant (lat. 5°41’40’” N., long. 120°47’10” E.); 38 meters; coral sand and shells; February 16, 1908 [A. H. Clark, 1908, 1912, 1918]. Albatross; Philippine Islands; 1907-08 [A. H. Clark, 1908] (8, U.S.N.M., 35206). North of central Java (lat. 5°41’ S., long. 109°21’ E.); November 21, 1907 [A. H. Clark, 1933] (1, Buitenzorg Mus.). Biliton; M. Korotnev, 1885 [Koehler, 1895; A. H. Clark, 1912]. Danish Expedition to the Kei Islands station 72; Dr. Th. Mortensen; Java Sea (lat. 5°41’ S., long. 105°57’ E.); 35 meters; stones and sponges; July 28, 1922 (1, the type specimen of H. apollo; C.M.). Danish Expedition to the Kei Islands station 103; Dr. Th. Mortensen; Sunda Straits (lat. 6°05’ S., long. 105°42’ E.); 52 meters; shell bottom; August 4, 1922 (1). Singapore; Svend Gad [A. H. Clark, 1909, 1918] (2, C.M.) Investigator; Andaman Islands [A. H. Clark, 1912] (1, I.M.). Geographical range.—From the Philippines to Biliton, Java, Singapore, and the Andaman Islands. Bathymetrical range-—From the shoreline down to 80 meters. The average of eight records is 33 meters. History.—This species was first mentioned by Dr. P. H. Carpenter simply as ‘‘Antedon” in an article on fossil comatulids published in 1880. He said that in one of the Challenger Antedons only four rays divide out of the five, the fifth remaining simple and undivided as is the case with all the rays of Ophioerinus (= Eudiocrinus). This refers to the specimen that he later (1888) described under the name of Antedon clemens. In the same paper he said that in a Challenger species of Antedon from 252 BULLETIN 82, UNITED STATES NATIONAL MUSEUM station 212 the ridges bordering the interarticular ligament fossae above run almost vertically upward for some little distance and then curve outward, so that the size of the muscular fossae above them is greatly reduced. Farther on he said that the only Antedon he knew that in the relation of the interarticular ligament and the muscular fossae presents features at all resembling those of Actinometra (that is, the Coma- steridae) is the anomalous Challenger species from station 212. In this the arrange- ment of the ridges is essentially the same as it is in Actinometra, though the shape of the distal faces of the radials is somewhat different. The specimen which he dissected was the fragment identified (in 1888) as Antedon quinduplicava which had lost its cirri, disk, and most of its arms. As it was practically useless in this condition, he made a preparation of its calyx. In the Challenger report on the stalked crinoids published in 1884 Carpenter mentioned an Antedon in which one of the rays consists of five joints, the axillary being a syzygy. This was the specimen he later (1888) described under the name of Antedon clemens. In the Challenger report on the comatulids published in 1888 Carpenter described Antedon clemens in the Palmata group and Antedon anceps and A. quinduplicava in the Savignyi group. Antedon anceps he also placed in his key to the species of the Milberti group and he noted it among the species of that group collected by the Challenger for the reason that one of the specimens had only 10 arms. He said that the single specimen of Antedon clemens is unique in every way, but he had some doubts as to the propriety of separating A. quinduplicava from A. anceps. The three specimens identified as Antedon anceps, the single specimen and fragment of another determined as A. quinduplicava, and the single specimen described as A. clemens all came from Challenger station 212. Dr. Clemens Hartlaub in 1891, following Carpenter, included clemens in his key to the species of the Palmata group, and anceps and quinduplicava in his key to the species of the Savignyi group. Prof. René Koehler in 1895 recorded and gave notes upon two specimens identified as Antedon anceps from Biliton. In my revision of the old genus Antedon published in 1907, anceps, clemens, and quinduplicava were referred to the new genus Himerometra. In 1908 I recorded Himerometra quinduplicava from Albatross station 5139, and also from the Albatross Philippine collection with no definite locality, and Himerometra anceps from Albatross station 5147 and from the Philippines with no definite locality. In my revision of the family Himerometridae published in 1909, I referred anceps to my new genus Amphimetra, while quinduplicava was assigned to the new genus Heterometra; clemens was not mentioned. In a paper on the crinoids in the Zoological Museum at Copenhagen published in 1909, I recorded and gave notes upon two specimens of Heterometra quinduplicava from Singapore. In a paper including descriptions of five new comatulids published on August 23, 1909, I established the new species Craspedometra aliena. The type specimen of this supposed new form was an example of Heterometra crenulata, but other examples included under this name in reality represented quinduplicava. A MONOGRAPH OF THE EXISTING CRINOIDS 253 In a third paper on the crinoids collected by the Albatross in the Philippines, which was published in 1911, I recorded a small specimen of Craspedometra anceps from station 5355 and added a note to the effect that an examination of the Challenger types in London had shown that Carpenter’s Antedon clemens is the same thing as his Antedon anceps and as my Craspedometra aliena. In my memoir on the crinoids of the Indian Ocean published in 1912, I included Craspedometra anceps, with Antedon clemens and Craspedometra aliena as synonyms. I said that an examination of the Challenger material in London had shown me that my Craspedometra aliena is merely a very large form of the species that Carpenter called anceps and, furthermore, that Carpenter’s clemens is also the same thing as his anceps. I remarked that the earlier pinnules of this species sometimes have the curious production of the distal corners of the segments which Carpenter described in Amphimetra variipinna (= Heterometra crenulata) and in Oligometra serripinna. This statement was the result of the confusion arising from the inclusion of specimens of both Heterometra crenulata and H. quinduplicava under Craspedometra aliena. Farther on I gave Heterometra quinduplicava, with its synonymy and range. Jn an appendix I recorded a specimen of Craspedometra anceps from the Andaman Islands. Tn a paper on the comatulids in the British Museum published in 1913, I included Amphimetra anceps, with Antedon clemens as a synonym, and also Heterometra quin- duplicava. In my memoir on the unstalked crinoids collected by the Siboga, published in 1918, I placed Antedon clemens and A. anceps in the synonymy of Heterometra quin- duplicava, of which I recorded a specimen from station 99. On the basis of the type specimen Oraspedometra aliena was placed in the synonymy of Heterometra crenulata. In 1933 I recorded a specimen in the Buitenzorg Museum from north of central i HETEROMETRA CRENULATA (P. H. Carpenter) Puate 23, Fiacures 89-92; Puate 24, Fiaures 93-96; PLatre 25, Ficure 101; Puare 26, Figures 102, 103 [See also vol. 1, pt. 2, figs. 445, 446 (pinnule tip), p. 261; fig. 718 (disk), p. 346.] ?Comatula dubia von Grarr, Das Genus Myzostoma, 1877, pp. 15, 22, 72, 79 (Semper, MS.; near Bohol, Philippines; myzostomes; nomen nudum).—Bravun, Centralbl. fiir Bakteriol. und Parasitenk., vol. 3, 1888, p. 185 (myzostomes; after von Graff).—A. H. Crarx, Unstalked crinoids of the Siboga-Exped., 1918, p. 77, footnote ® (identity). Antedon crenulata P. H. Carpenter, Journ. Linn. Soc. (Zool.), vol. 16, 1882, p. 507 (description; neighborhood of Borneo; discussion); Proc. Zool. Soc. London, 1882 (1883), p. 746 (specific formula).—A. H. Cuarx, Smithsonian Misc. Coll., vol. 60, No. 10, 1912, p. 2 (identity).— Gisiin, Kungl. Svenska Vet.-Akad. Handl., vol. 59, No. 4, 1919, p. 25. Antedon decipiens BELL, Proc. Zool. Soc. London, 1882, p. 534 (specific formula).—P. H. CarPEenTeER, Proc. Zool. Soc. London, 1882 (1883), pp. 740 ff. (discussion of Bell’s method of formulation and corrected formulae).—Brtt, Report Zool. Coll. H. M. 8. Alert, 1884, p. 155 (specific for- mula), p. 159 (description; Arafura Sea, 32-36 fms.; Dundas Strait; Prince of Wales Channel; station 144 [probably Thursday Island]), p. 162 (comparison with A. irregularis), pl. 11, figs. B, Ba.—A. H. Cuarx, Mem. Australian Mus., vol. 4, 1911, p. 714 (identity); Crinoids of the Indian Ocean, 1912, p. 31 (identity).—Gristin, Kungl. Svenska Vet.-Akad. Handl., vol. 59, No. 4, 1919, p. 25.—H. L. Cuan, The echinoderm fauna of Torres Strait, 1921, p. 6 (identity). Antedon irregularis Butt, Proc. Zool. Soc. London, 1882, p. 534 (specific formula).—P. H. Car- PENTER, Proc. Zool. Soc. London, 1882 (1883), pp. 740 ff. (discussion of Bell’s method of formu- lation and corrected formula).—BEtt, Report Zool. Coll. H. M. 8. Alert, 1884, p. 155 (specific 254 BULLETIN 82, UNITED STATES NATIONAL MUSEUM formula), p. 161 (description; Prince of Wales Channel; Torres Strait), pl. 18, figs. A, a-c.— P. H. Carrenrer, Quart. Journ. Mier. Sci., vol. 27, 1887, p. 386 (sacculi fairly abundant on pinnules).—A. H. Crarx, Mem. Australian Mus., vol. 4, 1911, p. 714 (identity); Crinoids of the Indian Ocean, 1912, p. 31 (identity); Smithsonian Misc. Coll., vol. 61, No. 15, 1918, p. 83 (identity).—Gisuin, Kungl. Svenska Vet.-Akad. Handl., vol. 59, No. 4, 1919, p. 25.—H. L. Crark, The echinoderm fauna of Torres Strait, 1921, p. 6 (identity). Antedon dubia von Grarr, Challenger Reports, Zoology, vol. 10, pt. 27, 1884, pp. 15, 18, 47 (Bohol, Philippines; myzostomes; nomen nudum).—P. H. Carpenter, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, pp. 110, 197, 258-261 (Arrou Islands), pl. 36, figs. 1-6.— Mincxmrt, Archiv fir Naturg., Jahrg. 71, 1905, vol. 1, Heft 1, p. 212 (syzygies; regeneration).—Gist£Nn, Kungl. Svenska Vet.-Akad. Handl., vol. 59, No. 4, 1919, p. 25. Antedon bidentata von Grarr, Challenger Reports, Zoology, vol. 10, pt. 27, 1884, pp. 15, 16, 17, 49, 57, 59 (Challenger station 186; myzostomes); Narrative, vol. 1, pt. 1, 1885, p. 316 (same); Zoology, vol. 20, pt. 61, 1887, p. 7 (same).—P. H. Carpenter, Quart. Journ. Micr. Sci., vol. 27, 1887, p. 386 (sacculi scanty on pinnules); Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 97 (Challenger station 187; naked disk), p. 262 (identity).—Bravn, Centralbl. fir Bakteriol. und Parasitenk., vol. 3, 1888, p. 186 (myzostomes; after von Graff). Antedon variipinna P. H. Carpenter, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 198 (Arrou Islands; notes), p. 256 (synonymy; description; Chatlenger station 186; Arrou Islands; other records; discussion), pl. 36, figs. 1-6, pl. 48, figs. 3-5, pl. 49, figs. 1, 2—Harriaus, Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 17 (in key).—Bertu, Proc. Zool. Soc. London, 1894, p. 394 (northwest Australia, 9-38 fathoms).—KonuiErR, Mém. Soc. Zool. France, vol. 8, 1895, p. 419 (Sunda Is.; Biliton; notes) —D’Arcy THompson, Proc. Roy. Soc. Edinburgh, vol. 22, 1900, p. 322 (range).—Bntt, in Gardiner, Fauna and geography of the Maldive and Lacca- dive Archipelagoes, vol. 1, pt. 3, 1902, p. 224 (South Nilandu, 19 fathoms).— Mincxert, Archiv fir Naturg., Jahrg. 71, 1905, vol. 1, Heft 1, p. 212 (syzygies; regeneration).—Hamann, Bronns KJassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, pp. 1580, 1582 (listed).—A. H. Criark, Mem. Australian Mus., vol. 4, 1911, pp. 716, 719 (identity of records by P. H. Car- penter, 1888, and Bell, 1894); Crinoids of the Indian Ocean, 1912, pp. 34, 35, 38, 40 (identity of previous records); Smithsonian Misc. Coll., vol. 61, No. 15, 1913, p. 88 (identity of speci- mens in B. M.).—-H. L. Cuarx, The echinoderm fauna of Torres Strait, 1921, p. 5 (identity).— GisL6Nn, Kungl. Svenska Vet.-Akad. Handl., vol. 59, No. 4, 1919, p. 25. Himerometra variipinna A. H. Cuarx, Smithsonian Misc. Coll., vol. 50, 1907, p. 356 (listed; in part); Proc. Biol. Soc. Washington, vol. 21, 1908, p. 225 (comparison with H. producta); Smithsonian Mise. Coll., vol. 52, 1908, p. 216 (Albatross Station 5157). Amphimetra variipinna A. H. Cuark, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 7 (listed) ; Proc. U. S. Nat. Mus., vol. 37, 1909, p. 32 (proximal pinnules compared with those of Craspedometra aliena); Vid. Medd. Nat. Foren. Kgbenhayn, 1909, p. 156 (Singapoze; description of specimens), p. 157 (compared with A. producta), p. 193 (collected at Singapore by Svend Gad); Proc. U. S. Nat. Mus., vol. 39, 1911, p. 540 (Albatross stations 5358, 5481); Die Fauna Siidwest-Australiens, vol. 3, Lief. 13, 1911, p. 486 (Alert records; northwest Australia), p. 440 (East Indian species occurring south to Baudin Island and Port Curtis), p. 443 (range on east coast), p. 444 (range on west coast), p. 446 (summary of west Australian records); Mem. Australian Mus., vol. 4, 1911, p. 722 (occurs south to Port Curtis), p. 723 (northwest Australia; Baudin Island), p. 734 (in key), p. 764 (synonymy; characters; localities; notes; range), p. 765, footnote (Chadwick’s record from Ceylon is Heterometra reynaudi); Smithsonian Misc. Coll., vol. 60, No. 10, 1912, p. 16 (Antedon variipinna of Carpenter not this species); Crinoids of the Indian Ocean, 1912, p. 31 (includes decipiens and irregularis Bell, 1884), pp. 34, 35 (=variipinna P. H. Carpenter, 1888), p. 36 (=variipinna Bell, 1894), p. 110 (synonymy; Pocock Island, 20 fathoms; description and comparisons; summary of previous records); Die Fauna Siidwest-Australiens, vol. 4, Lief. 6, 1913, p. 314 (correction of identity).—Gist4n, Kungl. Svenska Vet.-Akad. Handl., vol. 59, No. 4, 1919, p. 3 (listed), p. 5 (stations 7, 11), p. 23 (detailed descriptions), p. 26. Craspedometra aliena A. H. Cuarx, Proc. U. S. Nat. Mus., vol. 37, 1909, p. 31 (Albatross station 5157; description); vol. 39, 1911, p. 541 (considered a synonym of anceps [quinduplicava]).— REICHENSPERGER, Abh. Senck. naturf. Ges., vol. 35, Heft 1, 1913, p- 100 (comparisons) .— A MONOGRAPH OF THE EXISTING CRINOIDS 255 A. H. Cuarx, Unstalked crinoids of the Siboga Exped., 1918, p. 77, footnote 1 (synonym of crenulata). Amphimetra crenulata A. H. CLarxk, Smithsonian Mise. Coll., vol. 60, No. 10, 1912, pp. 2, 16 (identity) ; Die Fauna Siidwest-Australiens, vol. 4, Lief. 6, 1913, p. 314 (correction of previous records) ; Smithsonian Mise. Coll., vol. 61, No. 15, 1913, p. 22 (published references to specimens in B. M.; localities; notes); Internat. Revue gesamt. Hydrobiol. und Hydrogr., 1915, pp. 223 ff. (detailed account of distribution in Australia)—H. L. Cuarx, Biol. Results Fishing Exper. F. I. S. Endeavour, 1909-14, vol. 4, pt. 1, 1916, p. 21 (new locality; notes).—Gisuin, Kungl. Svenska Vet.-Akad. Handl., vol. 59, No. 4, 1919, p. 3 (discussion); Zool. Bidrag Uppsala, vol. 9, 1924, p. 39, footnote (disk reaches to IIBr,), p. 44 (reversion), p. 47 (obliquity of joint faces), pp. 77, 78 (syzygies), p. 285 (Cape Jaubert, 21 m.; details), fig. 33, p. 52, fig. 69, p. 75. Antedon philiberti Brit. Mus., MS. A. H. Cuarx, Crinoids of the Indian Ocean, 1912, p. 110; Smith- sonian Mise. Coll. vol. 61, No. 15, 1913, p. 22 (northeastern Australia; J. B. Jukes). Actinometra solaris Brit. Mus., MS. A. H. Cuarx, Smithsonian Misc. Coil., vol. 61, No. 15, 1913, p. 22 (Prince of Wales Channel; 5-7 fathoms). Craspedometra aruensis REICHENSPERGER, Abh. Senck. naturf. Ges., vol. 35, Heft 1, 1913, p. 82 (Aru Islands), p. 99 (west of Ngaiguli, 14 meters; detailed description and comparisons), fig. 8, p. 100 (lower pinnules). Craspedometra anceps REICHENSPERGER, Abh. Senck. naturf. Ges., vol. 35, Heft 1, 1913, p. 82 (Aru Islands), p. 99 (details of the locality; notes; comparisons), fig. 8, p. 100 (lower pinnules). Heterometra crenulata A. H. Cuarx, Unstalked crinoids of the Siboga-Exped., 1918, p. 77 (in key; range), p. 79 (synonymy; notes; stations 273, 274), p. 275 (listed) —H. L. Cuarx, The echino- derm fauna of Torres Strait, 1921, pp. 5, 6 (history), pp. 21, 192 ff. (distribution).—GisLién, Zool. Bidrag Uppsala, vol. 9, 1924, p. 215 (disk)—A. H. Cuarx, Journ. Linn. Soe. (Zool.), vol. 36, 1929, p. 638 (Baudin Island; 8-15 fathoms); Treubia, vol. 14, livr. 2, 1933, p. 212 (localities; notes).—Guistin, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 22, 27; Kungl. Fysiogr. Sallsk. Lund Férh., vol. 7, No. 1, 1936, p. 4 (French Indochina), p. 6 (range), p. 9 (localities; notes).—H. L. Charx, Mem. Mus. Comp. Zool., vol. 55, 1988, p. 33 (Roebuck Bay, 5-8 fathoms; Broome; Norwest Islet, Capricorn group; notes). Heterometra variipinna Gistin, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1984, pp. 7, 16, 27. Diagnostic features—This very variable species is easily recognized by the tnlarged lower pinnules, which are very strongly prismatic, the outer portion of the prismatic ridges on each segment being raised into conspicuous broad rounded eriangular processes so that the profiles of the pinnules are very strongly serrate, the teeth of the serration having broadly rounded ends. The animal is usually (though not always) stout and robust, with the dorsal surface of the elements of the division series and of the brachials swollen and with well-developed broadly rounded synarthrial and articular tubercles. The enlarged lower pinnules are usually (but not always) much stiffened and sometimes spinelike. Notes.—The specimen from Albatross station 5481 is small with 12 arms. The specimen from Albatross station 5358 has 11 arms 150 mm. long. The cirri are XI, 34-41, from 35 to 37 mm. long, rather more slender than usual. The processes on the segments of the proximal pinnules are less developed than is usually the case, and the general appearance of the animal is less rugged than in typical examples. The individual probably is not quite mature. The two specimens from Albatross station 5157 upon which Craspedometra aliena was based proved to be unusually slender and poorly developed examples of crenulata. The centrodorsal is thick discoidal with the bare polar area flat, 2-4 mm. in diameter. The cirrus sockets are arranged in a single or partially double alternating marginal row. 256 BULLETIN 82, UNITED STATES NATIONAL MUSEUM The cirri are XV-XVII, 29-36 (usually 35 or 36), 30 mm. long. They are moderately stout basally but taper very gradually to a slender tip. The first segment is short and those following gradually increase in length to the fourth or fifth, which is about as long as broad, the remainder being slightly longer than broad. The terminal 10 to 14 segments may have a slight dorsal carination. The opposing spine is small but prominent, terminal or subterminal, and is equal to one-third the width of the penultimate segment in height. The terminal claw is somewhat longer than the penultimate segment and is slender and slightly curved. The segments in the distal half or two-thirds of the cirri have purple saddle-shaped markings. The radials may be entirely concealed in the midradial line, or they may be equal to half the length of the IBr,. The IBr, are oblong, short, five or six times as broad as long, and laterally in contact. The IBr, (axillaries) are short, almost or quite triangular, two and one-half to three times as broad as long. The IIBr series are 4(3-+-4)—in one of the specimens two of them are 2—and are in close lateral apposition and flattened against their neighbors. One of the two specimens possesses a single I1IBr 2 series, which is developed internally. The 15-20 arms are 140 mm. long, and are elongated, slender, and evenly tapering. The first two brachials are subequal in size, slightly wedge-shaped, three to four times as broad as long in the middorsal line. The first syzygial pair (composed of brachials 3-4-4) is oblong, two and one-half or three times as broad as long. The next five brachials are approximately oblong, nearly four times as broad as long in the median line. The following brachials are triangular, somewhat over twice as broad as long with the longer side somewhat convex, after the proximal fourth of the arm becoming wedge-shaped, about three times as broad as long, and slightly longer terminally. Syzygies occur between brachials 3--4, again from between brachials 14+-15 to between brachials 29-+30, and distally at intervals of 11 to 19 muscular articulations. P,, is small and weak, about 6 mm. long, and composed of 13-16 segments, all of which are about as long as broad. P, is similar but slightly larger, 7 or 8 mm. long. P, is 9 to 12 mm. long with 18 segments, much stouter and stiffer than the preceding pinnules, and ending abruptly with a stiffened tip more or less as in the species of Stephanometra. The first two segments are not so long as broad, the third is about as long as broad, and the remainder are longer than broad, becoming twice as long as broad distally. The pinnule may be slightly carinate basally, and after the second or third the distal dorsal ends of the segments become thickened and project strongly, forming lateral processes resembling those in the type of crenulata. P3 is slightly stouter and slightly longer than P., but similar to it though usually with a few less segments. P, is 9 mm. long, comparatively slender, and its segments do not bear lateral processes; the segments become about as long as broad on the fourth or fifth and about twice as long as broad terminally. The following pinnules are similar, gradually increasing to 10 mm. in length, and then becoming more slender and slowly decreasing to 8 mm. in length. The color in alcohol is white, the segments in the distal half or two-thirds of the cirri with purple saddle-shaped markings. The type specimen of Antedon crenulata was thus described by Carpenter: The centrodorsal is a thick convex disk 5 mm, in diameter with a single or partially double row of cirri. The cirri are about XX, 30+. The sixth segment is longer A MONOGRAPH OF THE EXISTING CRINOIDS 257 than broad and those following are tolerably equal. The eleventh segment and its successors bear strong forward projecting dorsal spines which diminish again after about the twenty-fifth segment. The radials are partially visible, and have crenulated distal edges. The IBr, are trapezoidal, closely united laterally, and rising rather sharply toward the middle of their junction with the pentagonal axillaries, which are about twice their length. The postradial series are very close together, and divide twice, or sometimes three times. The IIBr series are 4(3+4) and the IIIBr series are 2. The first 2 segments beyond each axillary form a slight tubercular elevation (synarthrial tubercle) at the middle line of their junction. The first brachials are rhomboidal and are closely united laterally. The second brachials are longer and are sharply wedge-shaped. The first syzygial pair (composed of brachials 3++4) and the five or six following brachials are short and oblong, and those succeeding are triangular, more than twice as broad as long, and overlapping, gradually becoming smoother and more oblong in the middle and outer parts of the arms. Syzygies occur between brachials 3+-4, again from between brachials 13-++14 to between brachials 14+15, and distally at intervals of 8-11 muscular articulations. The pinnules are variable. P, seems to have been less stout than P, P, is small, usually less than half the length of P,. On the outer arms of the postradial series the fifth, sixth, and seventh brachials bear large pinnules (P2, P,, and P3) like that on the second (P;). Sometimes P, and sometimes P; is the largest. The pinnules of the next pair, on the eighth and ninth segments (P, and P,), are smaller again. On the inner arms the fifth brachial bears a small pinnule (P,) like that on the fourth (P,), and the next two pairs are large, the second pair, on the eighth and ninth bra- chials (P, and P,) being the larger. Sometimes, however, the only two really large pinnules are those of the seventh and eighth brachials (P; and P,), the preceding pair being smaller, but not especially so, like P,. These long lower pinnules consist of about 25 stout segments the distal ends of which have forward-projecting lateral processes. In the large lower segments these are chiefly limited to the outer side, but they appear on both sides in the later segments. Similar but less marked processes are visible on the cylindrical segments of the remaining pinnules, which increase again in size after the fifth pair but never become so long as the large lower pinnules. The color of the skeleton seems to have been white and that of the perisome purplish. Carpenter said that the preceding description was based upon some dried frag- ments of a moderately large Antedon, which appeared to be new. He said that it is altogether a larger species than variipinna, from which it is readily distinguished by its crenulated radials, tubercular arm bases, and smoother arms, while the inequality in the sizes of the lower pinnules is not of the same character in the two species. In the specimen from Tonkin the cirri have 44-48 segments, the outer with dorsal knobs. The IJBr series are 4 (3+4). Of the eight IJIBr series, four are 2 and four are 4 (3-4). The single IVBr series is 3. One of the IBr series is 6 (5+6), with a pinnule on the fourth element. There are 27+ arms. P; is the longest pinnule, 30 mm. long with 33 segments. The distal segments have the distal portion everted and spinous, and a little angular. There are faint tubercles at the distal outer borders. The Annam specimen has 32-36 cirrus segments, each with a dorsal longitudinal keel tending to form an indistinct dorsal tubercle. The IIBr series are 4 (3+-4). Of the three IIIBr series, two are 2 and one is 4 (3+4). The radials are visible; 258 BULLETIN 82, UNITED STATES NATIONAL MUSEUM they are crenulated. There are 20 arms. The arm bases are less rugged than those of the preceding specimen. P is about 20 mm. long with 22 segments; the eversion and spinosity of the distal segments is not very prominent. The individual is young. In the specimen from Cambodia the cirri have been lost. The IIBr series are 4 (3-4). Of the five I1[Br series present, three are 2 and two are 4 (3+4). P, is about 30 mm. long with 27 segments. The everted ends of the segments and the tubercles are well developed. The four specimens from north of central Java are all small. One of the specimens from north of western Java has 10 and the other has 11 arms. ‘The lateral processes on the segments of the proximal pinnules are strongly developed but slender. One of the specimens from off Cape Jabung, Sumatra, has 22 arms. Of the ten I1Br series present, nine are 4 (3+4) and one is 2. There is one IIIBr 4 (3+4) series following a I1Br 2 series, and one IIIBr 2 series following a I1Br 4 (3+4) series. Both of the II1Br series are internally developed. One of the specimens collected by Svend Gad at Singapore has 19 arms 80 mm. long. The cirri are XV, 31-34, 20mm. long. P2is 13 mm. long, with 20-23 segments. Six IIBr series are present on four postradial series of which five are 4 (8-+-4) and one is 2. There is one IIIBr 2 series, developed internally. In alcohol this specimen is flesh colored, with the arms crossed by purple bands. Another specimen from Singapore has 12 arms 80 mm. long and the cirri XIX, 32-35, 23 mm. long. There are two IIBr series developed on adjacent postradial series. The color in alcohol is dark brown. A third specimen from Singapore has 15 arms 80 mm. long. There is a single IBr series on each postradial series. Of the five I1Br series four are 4 (8-+4) and oneis2. The cirri are XVIII, 28-30, 20 mm. long. A fourth example resembles that immediately preceding and is of exactly the same size with 16 arms. All the II[Br series are 4 (3+4). The cirri are XVIII, 30-35. Both of these specimens are flesh colored. A specimen from Singapore collected by Mr. Gad and dated October 23, 1910, has 14 arms 62 mm. long. One of the I[Br series is 2. One postradial series bears two IIBr series, both 4 (3+4). Another specimen with the same date has 14 arms 75 mm. long. The cirri are XIX, 31, 20 mm. long. The dorsal pole of the centrodorsal is convex, 2.5 mm. in diameter. Prof. F. Jeffrey Bell’s Antedon decipiens was based upon specimens from the Arafura Sea in 32-36 fathoms, Dundas Strait, and the Prince of Wales Channel. He provisionally associated with it, as a variety, two specimens from Alert station 144 (probably Thursday Island). The characters he gave for Antedon decipiens are the following: The centrodorsal is small, with the cirri arranged on three levels. The cirri are about XX, 25, about 16 mm. long. The fourth-tenth segments are longer than broad and those following, which gradually become shorter, are provided with a well-marked spine. These spines decrease toward the end of the cirri, but the penultimate one is larger again. The radials are quite distinct. The IBr, are oblong, three times as long as broad (he meant as broad as long), partly in contact. The IBr, (axillaries) are A MONOGRAPH OF THE EXISTING CRINOIDS 259 almost triangular. The arms may or may not divide. Where the arms divide the IIBr series are 4(3+-4). The arms are about 14-18 in number and about 70 mm. long. The first brachials, or the first elements in the IIBr series, are always broad, and touch. The first brachials, which have sharp overlapping distal edges, are pretty regularly oblong; at about the seventh they become alternatingly wider on either side, to again become more regular later on. The first brachial syzygy is between brachials 3-++4, the second is from between brachials 13-14 to between brachials 16-++17, the third is from between brachials 24+25 to between brachials 27-28, and those following are separated by from 11 to 13 muscular articulations. P, (when present) and P, short. P, to Pa are much longer with the basal segments very stout, the free ends very deli- cate, and their outer side produced into a well-marked conical process. The succeed- ing pinnules are shorter, and these again increase in length; they are not composed of a large number of segments. The disk is small. Bell said that in the two varietal specimens from Alert station 144 the cirri are rather more numerous and more jointed, the whole animal appears to be more slender and delicate, and the color is ashy gray. Bell’s Antedon irregularis was based upon specimens from the Prince of Wales Channel and from Torres Strait. He described it as follows: The centrodorsal is flattened and small; the cirri are arranged in two marginal rows. The cirri are about XXYV (but there may be not more than XV), 30-35, the longest 24 mm. long. The basal segments are short, the fourth-ninth are longer than broad, the segments then shortening again. There are no dorsal spines, except on the penultimate segment, and that is exceedingly small. The radials are not at all, or are barely, visible. The IBr, are broad, in contact, with a median convex protuberance (a synarthrial tubercle). The IBr, (axillaries) are almost perfectly triangular. The I[Br series are 4(3+4), and the IIIBr series, when present, are 2. The 11-22 arms are about 85 mm. long. The earlier brachials have a well-rounded convex dorsal surface and are broader than long; soon, however, they become very markedly wedge-shaped and form a prominent projection alternating on either side (the articular tubercles). Toward the end of the arms these disappear. Syzygies occur between brachials 3+4, 11+12, and 21+-22; or 3-+4, 12+13, and 23-+24; or3+4, 14+15, and 23+24; or 3+4, 15+16, and 23+24; the distal intersyzygial interval is 7 to 11 muscular articulations. The first pinnules are very short; the third and fourth are the stoutest and longest, quite stiff, with well-developed broad lower segments, each of which has a marked protuberance on each side. The succeeding pinnules are shorter, and then again longer. The diameter of the disk is 6 mm. The color is pale flesh color, occasionally with a dark band here and there, especially at the syzygies. Sometimes there is a good deal of brown. The cirri are typically banded purple and white. Bell said that Antedon irregularis has some resemblance to A. decipiens, but it may be distinguished from the latter by the absence of spines from the segments of the cirri, the broader lower pinnules, and the greater length of the more distal pinnules. The specimens secured by the Challenger at station 186 and at the Aru Islands were thus described by Carpenter: The centrodorsal is a low and slightly convex disk. The cirri are XV-XXX, 20-35. Some of the lower segments are longer than broad; the later are usually somewhat compressed laterally and rather sharply carinate in consequence, but they sometimes bear well-marked spines. The radials are never 260 BULLETIN 82, UNITED STATES NATIONAL MUSEUM altogether invisible in lateral view and are often comparatively large and granulated externally. The IBr, are short, broad, and laterally united, forming more or less of a prominence at the middle of their junction with the broadly pentagonal IBr, (axillaries). The postradial series divide twice and sometimes three times. The IIBr series are 4(3+4) and the I1IBr series, when present, are 2. The dorsal surface of the elements of the division series is often considerably arched. The arms vary in number from 11 (probably even 10) to 25 or more; they are 100 mm. long and may have 180 brachials. The first six or eight brachials are relatively short and broad, nearly oblong in outline, and often much rounded dorsally. The following brachials are more triangular, with a variable tendency to overlap dorsally, and their broader ends project alternately on opposite sides of the arm to a greater or less extent. Farther out they become more quadrate again, but remain relatively short and broad and more or less overlapping until almost the very endofthearm. Thefirst brachial syzygy occurs between brachials 3+4, and the second from between brachials 11-++12 to between brachials 16+17; the distal intersyzygial interval is 7-13 (usually 10 or 11) muscular articulations The first pinnules are comparatively small and consist of about 20 short seg- ments of which the lowest are broad and slightly carinate. P,, if present, is smaller than P;, and so is P,. The following pinnules may reach nearly 15 mm. in length, with as many as 25 segments, which are both longer and stouter than in the lower pinnules. The relative sizes of the pinnules vary greatly. In the outer arm from each IJBr series the largest pinnules are generally P, and P,, but on the inner arm the largest pinnules are P, and P; (on the seventh and eighth brachials), while on arms borne directly on the IBr axillaries P; and P, are usually the longest. All but the lowest of these large pinnules have strong and blunt lateral processes at their distal ends. The pinnules of the next pair are generally considerably smaller, with rela- tively shorter segments, which gradually become more elongated in the slender distal pinnules but lose the lateral processes at their ends. The disk is 10 mm. in diameter naked, and much incised. The sacculi are small but abundant. The color in alcohol is ashy gray, white, or pale flesh color, with frequent bands or patches of purple or yellowish brown; sometimes purple with whitish bands. Carpenter said that this is in some respects the most remarkable species of Antedon that he had seen, for it has a very considerable range of variation and has been described under four different names. He called it Antedon variipinna, but unfortunately the type specimen of variipinna from Borneo represents quite a different species with cylindrical instead of sharply prismatic lower pinnules. He said that the Challenger collection contained two individuals from the entrance to Prince of Wales Channel in Torres Strait that agree with the type of crenwlata in several points, but they have no IIIBr series at all. One of them has spines on the distal cirrus segments though the cirri of the other are merely carinate. At first sight, he said, it did not seem advisable to unite these two forms under one specific name, since one had IIIBr series and the other lacked them. He therefore gave the Challenger specimens the manuscript name bidentata, which he later sup- pressed. He remarked that the same course was taken two years later by Bell who described Antedon decipiens with spiny cirri and no IIIBr series, and A. irregularis with unarmed cirri and IJJBr series. But their other characters, he noted, especially the short brachials and the lateral projections on the lower pinnules, agree very A MONOGRAPH OF THE EXISTING CRINOIDS 261 closely with those of Antedon crenulata. He said that Bell appears to have regarded the absence of IIIBr series in Antedon decipiens and of cirrus spines in A. irregularis, which has IIIBr series, as sufficient to separate both these types from A. crenulata. When he began to revise the endocyclic comatulids with the I[Br series 4(3+-4) included in the Challenger collection, the descriptions of which had been written five . or six years before, Carpenter found that figures of a form closely allied to Bell’s Antedon decipiens and also of an example of Bell’s A. irregularis, but without [Br series, had already been drawn.’ Both had been obtained in Prince of Wales Channel and had formerly seemed to him, as the Alert specimens from the same locality had seemed to Bell, to represent two different specific types which could not be referred either to Antedon variipinna or to A. crenulata. A third form from the Aru Islands also appeared to be new, and Carpenter figured it in the Challenger report under the name of Antedon dubia, not being quite clear in his mind whether the occurrence of a IIIBr series was normal or merely due to the regeneration of a 10-armed form as is so often the case in Antedon bifida and in other species. Later Carpenter made a critical study of all the Alert material and also recon- sidered his descriptions of Antedon variipinna and of A. crenulata. The result was that he found himself unable to discover any characters sufficiently constant to be of specific value as distinguishing Antedon irregularis from A. decipiens, or any of these three from A. variipinna and A. crenulata. He remarked that Bell gave no details respecting the relative size of the lower and distal pinnules in Antedon decipiens and A. irregularis and that after examining his material he found difficulty in attributing the difference to anything more than the size of the individual specimens, those of A, decipiens being generally smaller than those of A. irregularis. The presence of spines on the cirrus segments of A. decipiens and their absence on the more numerous segments of the cirri in A. irregularis seemed, however, to be good specific characters. But when he came to examine the gray specimens from Prince of Wales Channel (in reality from Alert station 144), which Bell had provisionally regarded as a variety of the white individuals obtained at the same locality because of their cirri being ‘rather more numerous and more jointed,” he found the cirrus segments to be also unprovided with definite spines, although they have the same sharply carinate appear- ance as those of A. irregularis. The radials of these individuals are also mostly concealed, as is the case in A irregularis, though in the type of A. decipiens they are “quite distinct’’ as described and figured by Bell; but they are much less distinct in the white individuals from Prince of Wales Channel. In all the specimens from the last-mentioned locality, therefore, the radials resemble those of Antedon irregularis rather than of A. decipiens; but some of them had spiny cirri as in the type of A. decipiens, while in others the cirrus segments are only sharply carinate as in A. irregularis. The arms and pinnules of all the specimens, however, are most like those of A. decipiens. Carpenter said that it would seem impossible, therefore, to make any distinction between the two species on the basis of the characters presented by the arms, the radials, or the cirri, and this conclusion is confirmed by the following considerations: 3 Challenger Report, Zoology, vol. 26, pt. 60, 1888, pl. 48, figs. 3-5 (form allied to A. decipiens), pl. 49, figs. 1, 2 (A. irregularis). 262 BULLETIN 82, UNITED STATES NATIONAL MUSEUM The two individuals from the Aru Islands, which he formerly referred to a new species, Antedon dubia, have about 35 cirrus segments with the later ones carinate as in A. irregularis. But they have relatively large and conspicuous radials with a sculptured surface exactly as in Bell’s figured specimen of Antedon decipiens, which, like these, has no IIIBr series. The arm bases of the smaller individual from the Aru Islands resemble those of A. decipiens, while those of the larger show more of the characters of A. irregularis. On the other hand, A. variipinna and A. crenulata both © have IJIBr series and 30 or more spiny cirrus segments, while the radials are fairly distinct, those of A. crenulata being more or less sculptured. Neither species has specially rounded arm bases, like those of A. irregularis, though the general outline of the ossicles is the same in all the types. Carpenter said that the variation in the characters of all these different forms may be conveniently expressed by letters as follows: Niwmper oF cirrus/seementss. ee 30't0' do=-seeee= A 20 eee eee a Character of cirrus segments_--_-__-------------- Distinctly spiny-B Sharply carinate__-_b Radights.22 eee Seo pie eo om cknin gaan Distincts=-—- === Cc Mostly concealed_-_--_¢ Rr serige os eee he ta be nk ee Present... --=— D Absent sy. 5 24 d ATM 'DARCS). We ee a an once e onan Much rounded__E Less rounded_--__---- e UE nn ee ate Sa eee ano sae a oe Serrate_.____.-- F Fairly smooth---_---- f He noted that we then get the following expressions to denote the eight forms of this specific type, five of which have been regarded as representing different species: [livAntedonwartipiniad...2-> =) a2n2us Jee b= bots seen. ABCDeF Canton] Be Antenons ChANULOIG ate. fo SS foe ee oot ABCDef Borneo 3. Antedon decipiens type (Alert) ._.--.----------------- aBCdef Arafura Sea 5, VAMAGON GECVinene Vala CAT) = — nn as anno ane ee Abedef Prince of Wales Channel 5. Antedon decipiens var. (Challenger)___..-.__---------- ABedeF Prince of Wales Channel 6. Antedon irregularis (Challenger) __.-...--..-.-------_- Abed EF Prince of Wales Channel 7. Antedon irregularis (Alert). .._.----22..---222------- AbeDEF Torres Strait and Prince of Wales Channel §. Antedon autre (ChOUCnger) = oo = oe eee ee gee AbCdEF Aru Islands Carpenter said that with these facts before us it is difficult to avoid the conclusion that we are dealing with but one specific type, and this conclusion is confirmed by the fact that in all these different forms the general shape of the brachials and the char- acters of the pinnules are respectively identical, though the latter vary considerably in the degree of development. The distal brachials have the same shape throughout the whole series, as shown in the Challenger examples from the Aru Islands and from Torres Strait. On the other hand, the alternating lateral projections of the ossicles in the lower parts of the arms are very marked in the form from Torres Strait, which Bell called Antedon irregularis, and it is fairly distinct in those from the Arafura Sea and from the adjacent Aru Islands. But it is comparatively insignificant in the other form from Torres Strait, which has much less convex [Br and IIBr series than the irregularis form from the same locality. Carpenter said that another universal character of all the different varieties he has referred to this species is the large size of the pinnules on the fourth (that is, fifth) and the two or three following brachials, and the lateral projections at the distal ends of their component segments. Pp, when present, is comparatively small, but its successor, P;, is somewhat larger, though that on the next brachial (P,) is smaller A MONOGRAPH OF THE EXISTING CRINOIDS 263 again. Beyond this point, however, there is much variation. The pinnules of the next three or four brachials are considerably longer and stouter than P,, being the largest pinnules on the arm. In those arms that spring directly from the IBr axillary, so that there is no Pp, the largest pinnules are usually P; and P,. When, however, a IlI[Br axillary is present, the arm borne on its inner face usually has its largest pin- nules on the sixth and seventh brachials, while on the outer arm they are usually on the fifth and sixth (P, and P,).{ But this arrangement is very far from being a con- stant one. The next two pinnules after the large pair may also be of considerable size and composed of somewhat elongated segments, but in other cases they show a considerable alteration both in the size and in the character of their component segments. Carpenter noted that the double row of lateral projections on the segments of these proximal pinnules is developed in rather a singular manner. Their basal seg- ments are somewhat flattened against the arm and the upper edge of their broad dorsal surface is sharpened and more or less carinate, while its distal end is marked by a median process of variable prominence. As the following segments lose their flattened appearance and become more rounded the carination of the upper edge develops into a strong blunt process at the distal end of the segment on its inner side, while the mediodorsal prominence passes into a corresponding process on the outer side. There is much variation, however, in the exact nature and mode of development of these processes. It was remarked by Carpenter that the frequency of the division of the post- radial series in this species, and therefore the number of the arms, is subject to great fluctuations. He said that IIIBr series occur only in the single specimen, which he named Antedon crenulata, and sometimes also in the form described by Bell as Antedon irregularis. A large number of individuals were obtained by the Alert, and the major- ity of them have two or more IIIBr series, though in others, as in the Challenger specimen, IIIBr series are entirely absent. Bell gave the number of arms as ranging from 11 to 22, but Carpenter said he seemed to have overlooked one example in which there are 25. The occurrence of an individual with only 11 arms makes it quite possible that a 10-armed variety of this protean type may eventually be discovered. In fact, according to Carpenter, the two individuals that he formerly called Antedon dubia are not improbably of this nature. One of them has two IIBr series, and the other only one. But in each case they result from regeneration of the arm at the syzygy between brachials 34-4. This may perhaps have originally supported a I1Br axillary, or it may have given rise to only one arm, which was replaced by two after fracture as it so often the case. Under these circumstances Carpenter thought it safer to assign this species a place among the 10-armed species to which it can defi- nitely be referred if ever an individual is found in which IJBr series are entirely absent. He said that no harm will be done if it never turns up, and should it do so it will run less risk of being baptised as a new species. The specimens referred to by Carpenter as Antedon dubia are simply young individuals undergoing adolescent autotomy. But Carpenter was not aware of this curious and wasteful method of arm reduplication. Carpenter said that there is one point relating to the extremely variable char- acters of this species that seemed to him to be of special importance. The variations 264 BULLETIN 82, UNITED STATES NATIONAL MUSEUM he described are not altogether due to difference of locality. Varieties 5 and 6 were found associated by the Challenger in Prince of Wales Channel. Variety 7, which is only variety 6 with II1Br series, was found associated with it by the Alert both in Torres Strait and in Prince of Wales Channel, and in the latter locality variety 4 was obtained as well. It was therefore evident to Carpenter that the cause of these remarkable variations in one and the same specific type must be attributed to some- thing more than a mere change of local conditions. The specimen recorded by Dr. Hubert Lyman Clark from off Double Island Point, Queensland, is a fine adult individual with 28 arms. There are about XXV cirri, the longest with 45 segments. The color, dry, is uniformly brownish white. The three specimens from 10 miles southwest of Mapoon are especially fine. One has 19 arms 150 mm. long, and the cirri XVI, 41-43, from 25 to 30 mm. long. Another has 21 arms 150 mm. long and the cirri with 40-42 segments, 30 mm. long. The third has 17 arms 130 mm. long and the cirri XXIII, 41-44, 30 mm. long In these three specimens the processes on the segments of the lower pinnules are strongly developed and typical. The cirri are more or less carinate distally. The synarthrial tubercles are prominent, so that the animals are characteristically rugose. Of the two specimens from Port Curtis one has 14 arms 80 mm. long, and the cirri XXIII, 35-40, 30 mm. long. The other has 26 arms 100 mm. long, and the cirri XXIX, 35-45 (usually 40-41), 30 mm. long. The centrodorsal is 5 mm. in diameter at the dorsal pole. The arm bases are exceptionally rugged with the syn- arthrial tubercles large, prominent, and bluntly conical. The lower brachials are very short but much swollen, with strongly projecting distal edges. The division series are just in apposition laterally. The distal cirrus segments are very strongly and sharply ecarinate, this carination standing out as a broad spine of which the outer edge is almost or quite parallel with the longitudinal axes of the segments. The specimen from Mast Head Island is a fine example with 22 arms. One of the specimens from Baudin Island has the lateral processes on the seg- ments of the proximal pinnules greatly exaggerated. The other is small. Of the specimens from Holothuria Bank six are large, very rugged, and typical examples with the lateral processes on the segments of the proximal pinnules excep- tionally well pronounced. These have 10, 11, 12, 13, 15, and 17 arms. One small specimen has 15 arms. There are also four small 10-armed specimens. The specimen from the vicinity of York Sound is small. Dr. H. L. Clark, basing his conclusions on 50 specimens, 30 dredged at various points near Broome in 9-15 meters, 18 from on or near Pearl Shoal in 9-15 meters, and 2 from Norwest Islet in the Capricorns, said that the number of arms ranges from 16 to 30, but the great majority of individuals with arms over 60 mm. long have 20 to 25. The cirri are XX-~XXXII, 34-46, but small individuals often have fewer segments. The largest specimens have the arms 75-90 mm. long. There is great diversity in the development of the projections on the segments of the basal pinnules; often they may best be designated as spines, but in other cases they must be char- acterized as “‘wings.’’ Light-colored individuals seem on the whole to tend toward wings, while the dark-colored ones are more spinous. The correlation is, however, very imperfect. A MONOGRAPH OF THE EXISTING CRINOIDS 265 Dr. Clark said that one very small comatulid, apparently a Heterometra, and hence almost surely this species, was taken in 1932 at Broome. There are 13 arms about 15 mm. long. The division series are 4(3-++4), but the cirri are only X, 13. The color is very pale brown, the cirri with a very slight tinge of violet. The single specimen from Mjéberg’s station 7 and the two from his station 11 were thus described by Dr. Torsten Gislén: The centrodorsal is thickly discoidal with an almost flat papillated dorsal surface which is from 3 to 4 mm. in width. The cirrus sockets are arranged in two rows. In two of the specimens the cirri are XXIX, 33-46, the marginal cirri being 38-46 and those of the distal row having 33-39 seg- ments. The length of the cirri is 25 to 30 mm. Included in the number of the cirri are some empty fresh sockets from several of which young cirri are regenerating. In the marginal cirri dorsal carination begins at the nineteenth-twenty-first segment (in an exceptional case at the twelfth) while in the cirri of the distal row it begins at the fourteenth-sixteenth. The third specimen has the cirri XXXV, 35-40; the dorsal carination first appears on the very last segments. The proximal cirrus seg- ments are laterally flattened, and the distal are laterally compressed. The longest proximal segments are one-third again as broad as long, and the distal segments are twice as broad as long. The first segment is short, four times as broad as long. The opposing spine is usually a little stronger than the spines on the preceding segments. The terminal claw is curved, and is at least as long as the penultimate segment. The anterolateral angles of the radials project as small triangular portions between the bases of the cirri and the IBr;. Their surface is strongly granulated— they are almost completely concealed by the centrodorsal, however, in the specimen from station 7. The IBr are laterally in contact. The IBr, and IBr, are united by synarthry, each having a median tubercle. This points directly up on the IBr, but is bent backward on the IBr2 so that the two tubercles converge to a joint tip. This is repeated on all the pairs of ossicles united by synarthry. The IIBr series are 4(3-++4), in exceptional cases four arranged in two synarthrial pairs; in the latter case each pair of ossicles shows a synarthrial tubercle. In the specimen from station 7 two of the IIBr series are 2. The IIIBr series are 2, in a few cases 2(1+2). The ossicles following the axillaries are united interiorly. The syzygies in the division series unite a very short discoidal hypozygal and an irregularly triangular epizygal— the axillary. The division series and arm bases are in close lateral apposition, but they are little flattened laterally. The elements of the division series and the proximal brachials are plump and swollen. The three specimens have 23, 24, and 30 arms, 85to100 mm.inlength. The first 10 brachials are discoidal and those following become wedge-shaped, and again dis- coidal in the distal portion of the arms. From the seventh or eighth brachials onward | there are thick transverse dorsal prominences (articular tubercles), which are devel- oped most strongly on the side of the segment that carries the pinnule, thus alternating on the right and left sides. The distal segments are collar-shaped. Gislén gave the distribution of the syzygies in a typical postradial series. The IBr series is 2. The two IIBr series are 4(3-+4). There are two IIIBr series, both internally developed, one 2 and the other 2(1+2). On one of the outermost arms, arising from a IIBr axillary, the syzygies are between brachials 3+-4, 12+13, and 2082444018 266 BULLETIN 82, UNITED STATES NATIONAL MUSEUM 20-+21; on the other outermost arm they are between brachials 3+4, 13+14, and 26+27. On one of the two arms arising from the IIIBr 2 series the syzygies are between brachials 3-+-4, 10+11, and 18+19, and on the other they are between brachials 3-+-4, 15-+-16, and 26+27. On one of the two arms arising from the IITBr 2(1-+-2) series the syzygies are between brachials 3+-4, 11412, and 24+25, and on the other they are between brachials 3+4, 15416, and 27+28. The distal intersyzygial interval is from 13 to 21 muscular articulations. Ppis 6 mm. long with 21 segments of which the first is short and the second-tenth are about as long as broad with prominences directed outward (toward the arm tip); the distal segments are shorter with the inner margin provided with low knots. The whole pinnule is compressed and twisted a quarter of a turn. P, is 7 mm. long, with 21 segments, very much stouter and stronger than Pp. The first two segments taper distally, and are somewhat longer than broad. From the second segment onward there is a sharply marked crest directed toward the arm tip. From the sixth segment the carination branches, and from this point to the fifteenth segment the pinnule is rather pronouncedly two or three edged. The sixteenth—-twentieth seg- ments are more rounded with thickened spinous borders distally. P, is 8-10 mm. long, with 18 segments of which the first two are without out-turned prominences and the remaining segments have large and very spiny processes which on the last eight form a spiny garland around the distal part of the segments. This garland also appears, though still more delicately developed, up to Py. Ps; is 11 or 12 mm. long, with 16-18 segments. P, is 7-9 mm. long, with 12 segments. P,, is 7 mm. long, with 15 segments. P, is rather short, corresponding to Pp. The longest pinnule is P,. P, is a little shorter than P,. The disk is without calcareous con- cretions or plates. The color in life was minium red; in alcohol it is the same but fainter. This description is not altogether clear, but it cannot apply to any species other than the one under consideration. Gislén discussed the relationship between varii- pinna and crenulata at considerable length, but in his discussion makes no mention of the differences in the proximal pinnules by which the two species are easily distinguished. The specimens from Siboga station 273 exactly resemble others from northwestern Australia. The arms are stout, basally strongly flattened laterally, and very rugose. The lateral processes on the pinnule segments are very strongly developed. One very fine typical example has 22 arms 120 mm. long. All the IIBr series are 4(3-++4), and both the IIIBr series are 2 and are internally developed. The color is white, with the outer part of the pinnules and the cirri violet. Another has 20 arms 110 mm. long. Two of the IIBr series are absent, the eight present being 4(3+4); the two I11Br series, both of which are internally developed, are 2. The color is similar to that of the preceding, but there are in addition a few narrow purple bands and blotches on the outer part of the arms. A third has 16 arms 120 mm. long. There are five IIBr 4(3+-4) series and one IIIBr 2 series, the latter internally developed. A fourth has 16 arms 115mm. long. There are six II]Br 4(3+-4) series. The three remaining examples are similar. A MONOGRAPH OF THE EXISTING CRINOIDS 267 The two specimens from Siboga station 274 are small. The larger has 13 arms 75 mm. long. There are three I[Br 4(3-+4) series on three separate postradial series. The animal is flesh colored with regular narrow purple bands on the arms. The specimen collected by Dr. Merton in the Aru Islands was thus described by Reichensperger: The centrodorsal is somewhat hemispherical, convex, the free dorsal pole about 3 mm. in diameter and sculptured with fine pits. The cirri are arranged in two or three irregular rows. The cirri are XX XIII, 40-42, about 35 mm. in length. None of the cirrus segments are longer than broad; the middle ones are about as long as broad and they bear toward the end a prominent dorsal keel. The antepenultimate segment is toothed. The terminal claw is short and not very stout, but sharp. When viewed from above the cirri are seen to be somewhat less sharply pointed than those of Craspedometra acuticirra, although they have in general great similarity to the cirri of that species. The radials are visible for the most part only in the interradial angles of the calyx where their adjoining anterolateral corners are to be seen. The IBr, are very broad and short, about five times as broad as long. The IBr, (axillaries) are shortly pentagonal. The IIBr series are always 4 (8+4); they are composed of short rugose elements which appear as if laterally compressed at the articulations. This lateral compression is continued to the eighth or tenth segment of the free undivided arms. When present the IIIBr series are 2; the first segment is short discoidal, evenly convex dorsally, and the axillary is half again as long and pentagonal. There are no further division series. The first ele- ments of each division series are laterally united. There are 22 arms which are up to 130 mm. long. The arms gradually taper distally. The 9 or 10 lowest brachials are discoidal, short, and laterally compressed; those following become more and more wedge-shaped and distally markedly convex dorsally, with the distal ends laterally and dorsally produced so that the middle portion of the arm is very uneven. The distal brachials are much more even and are short and discoidal. The first syzygy is between brachials 3+-4, and the second is somewhere between brachials 12+13 and brachials 15+16. Pp is short, 10-12 mm. long, and is composed of 25-28 segments, which at first are broader than long, in the middle of the pinnule about as long as broad, and later broader again. The pinnule is but little stiffened and becomes slender and some- what moniliform distally. P; is longer than Pp, of similar form but stiffer, 16-17 mm. long; the component segments are relatively markedly longer than those of Pp. Pz is stiff, 19-20 mm. Jong, with 24-26 segments. It tapers gradually and the distal portion is recurved and finally pointed. The three basal segments are broader than long, the fourth is about as long as broad, and those following, if the pinnule is viewed laterally, are longer than broad. P3 is similar to P, but markedly stouter and prominently keeled. It is up to 23 mm. long and is composed of 26-28 seg- ments. Py to P; are basally, and P; to P; are also farther out, prominently carinate. The latter have on the outer side of the segments from the third onward lateral processes involving the distal half of the segments which increase in size distally. P, and a number of the succeeding pinnules are straight, rather stiff, of approxi- mately the same length, 11 mm., and composed of 16-18 segments which except for three or four of the basal are longer than broad. On individual arms P; may be the smallest pinnule, only from 9 to 10.5 mm. long. From this point the size very slightly 268 BULLETIN 82, UNITED STATES NATIONAL MUSEUM increases to about 13 mm. in the distal portion of the arms. All the pinnules in about the proximal half of the arms seem to be rather stiff, and this, taken in connection with the uneven dorsal sculpture of the arms, gives the animal a robust and spiny appearance. The disk is lacking. The color is white, the tips of the cirri dorsally and the ambulacral grooves tinged with violet. This specimen was recorded by Dr. Reichensperger under the name of Craspedo- metra anceps, but he referred it to this species with reservations and suggested that if it proved to represent a new species, as he was fairly sure it would, it might bear the name of Craspedometra aruensis. He said that it could not be confused with acuticirra (from Sydney) or with amboinae, but that it bears a certain similarity to anceps although the cirri are about twice as numerous and are composed of more numerous segments, the pinnules are markedly longer with more numerous segments and are quite different in their relative length and stoutness, while it seemed to him that the lateral compression of the brachials and the strong convexity of the distal portion of the arms are in no previously described species so pronounced as they are in thisspecimen. The figure he gave of the lower pinnules shows that these are of the type characteristic of, and peculiar to, H. crenulata, and the description in all its details applies perfectly well to certain forms of this highly variable species. Professor Koehler’s specimen from Biliton was much mutilated, most of the arms having been broken off while only a single cirrus remained intact. There were XXII cirri. The radials are barely visible. There are one or two IIIBr series on each postradial series—six II]Br series in all. Presumably, therefore, there were 26 arms. The dorsal surface of the elements of the IBr series, the two first elements of the II Br series, the 111 Br series, and the first two brachials is strongly sculptured. The lateral borders of the brachials are swollen and prominent. The second syzygy is at about the sixteenth brachial. P, is very small, with at the most 25 segments. The pinnules following, which are larger, bear lateral processes on their segments which have the form of those figured by Bell for his Antedon irregularis. The general color is gray. The specimen from Pocock Island has 20 arms 110 mm. long. Three IIBr series are lacking, but this deficiency is compensated by the development of three 111Br series. The seven I1Br series are extraordinarily irregular. One is 8 (83+4; 5+6; 7+-8), one is 2, one is 4 with the two distal elements united by synarthry, and four are 4 (3-++-4). The three IIIBr series are also irregular; one is 1, one is 2, and one is 4 (3+-4), the last two being on a single postradial series on which the IITBr 2 series is internal and the II]Br 4 (3-++4) series external. This example agrees with others from Singapore in the collection of the Copenhagen Museum, and with others from Australia in the collection of the Australian Museum. The specimen from South Nilandu recorded by Bell I have not seen. He said that it was taken among living corals at the seaward end of a passage, but gave no notes on it. I have grave doubts concerning the correctness of the determination, as the Maldive archipelago lies considerably west of the Andamans, which represent the western limit of the range as outlined by specimens I have been able to examine. Color in life-—Dr. H. L. Clark wrote that some half dozen specimens dredged on Pearl Shoal outside Roebuck Bay seemed at the moment the most lovely comatulids he had ever seen. The arms were a pure milk-white, the cirri bright rose-red in sharp A MONOGRAPH OF THE EXISTING CRINOIDS 269 contrast. The somewhat rigid arms remained curved outward like the petals of a flower, and the old book name “sea-lily” was most appropriate. He subsequently dredged other specimens the coloration of which was so different that they were not even suspected of being the same species. In 1932 he met with Heterometra often, and the diversity of color led to the supposition that he was collecting three different comatulids. Dr. Clark says that according to coloration three forms may be recognized, but with sufficient material it is obvious that they intergrade completely. At one extreme are the milk-white forms with rose-colored cirri. When preserved in alcohol or dried these lose much of their beauty, the white becoming dingy or tinged with purple and often small spots or bloches of a purplish shade appear. At the other extreme are the individuals, usually large ones, that are deep red-purple or dark crimson with few or no lighter markings; these commonly have a coarser and more rugged struc- ture than the white ones. The third group includes the more or less handsome indi- viduals with banded arms; these may be white, or at least very light, as to ground color with the arms banded with red or purple of some shade, or they may be red or purple with whitish bands and markings. The number, width, and shade of the arm bands show endless diversity. The cirri are always red or purple of some shade, ranging from light rose to a deep red-violet. One very small specimen with 13 arms about 15 mm. long was very pale brown, the cirri with a very light tinge of violet. Abnormal specimens.—In the specimen from Pocock Island one of the JIBr series is 8(3-+4; 5+6; 7-++8), one is composed of four elements united in two synar- thrial pairs, and one is 2; of the three IIIBr series one is 1, one is 2, and the third is 4(3+-4). Gislén recorded a specimen from Mjéberg’s station 7 with a IIBr series consisting of four ossicles united in two synarthrial pairs and a IIIBr 2(1+-2) series. In the specimen from Tonkin described by Gislén one of the [Br series is 6(5+6). In the specimen from Annam an external arm arising from a I[Br 4(3+4) axil- lary has the first syzygy between brachials 4+-5; another arm on the same postradial series arising from the inner side of a I][Br 2 axillary has the first syzygy between brachials 7-8; a third arm on the same postradial series arising from the [Br axillary has syzygies between brachials 3+-4, 5+6, and 7-+8. Remarks.—This species reaches its extreme development (form irregularis) on the coast of Australia, where the individuals are usually large and rugged with nu- merous arms and with the characteristic features of the enlarged proximal pinnules highly developed and often exaggerated. Northward the individuals become less extremely developed and are usually smaller with fewer arms, more slender, and less rugged, with the characteristic features of the proximal pinnules moderately devel- oped. This is the typical form (crenulata) that occurs throughout the range of the other forms, except perhaps in the extreme south, and is the prevailing form at Singa- pore and in the Philippine Islands. From the Aru Islands to the Philippines the species occurs in a still more slender form (aliena [=aruensis]) in which the cirri are merely slightly carinate distally and the processes on the segments of the enlarged proximal pinnules are only very feebly developed. Bell’s name decipiens and Carpenter’s names bidentata and dubia were applied to young individuals of the highly developed form called irregularis by Bell. 270 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Basing his conclusions on specimens from Tonkin, Annam, and Cambodia Gislén says that evidently northward this species tends to pass over into H. variipinna, from which it seems scarcely to be distinguishable as a good species. Localities —Albatross station 5481; between Samar and Leyte, Philippines, in the vicinity of Surigao Strait; Cabugan Grande Island (N.) bearing N. 86° W., 3.8 miles distant (lat. 10°27/30’ N., long. 125°17’10’’ E.); 111 meters; sand, shells, and gravel; July 30, 1909 [A. H. Clark, 1911]. ?Near Bohol, Philippines; Prof. Carl Semper [von Graff, 1877, 1884; P. H. Carpenter, 1888; A. H. Clark, 1912]. Albatross station 5385; Jolo (Sulu) Sea; Sandakan Light bearing S. 34° W., 19.7 miles distant (lat. 6°06’40’’ N., long. 118°18’15’’ E.); 71 meters; mud; January 7, 1909 [A. H. Clark, 1911]. Albatross station 5157; Tawi Tawi group, Jolo Archipelago; Tinakta Island (N.) bearing S. 80° W., 3.3 miles distant (lat. 5°12’30’’ N., long, 119°55’50’’ E.); 33 meters; fine sand; February 21, 1908 [A. H. Clark, 1908, 1909] (2, U.S.N.M., 25516, 35197). Philippine Islands [A. H. Clark, 1911, 1912]. This refers to the four preceding records. Neighborhood of Borneo [P. H. Carpenter, 1882, 1888; A. H. Clark, 1912]. Off southeastern Borneo (lat. 3°12’ S., long. 116°38’ E.); 14-19 meters [A. H. Clark, 1933] (1, Buitenzorg Mus.). Along Bay, Tonkin; Dr. C. Dawydoff [Gislén, 1936]. Nha’trang Bay, Binh Damon (Tre Island), Annam; Dr. C. Dawydoff [Gislén, 1936]. Off Phuquoc, Duougdong, Cambodia; Dr. C. Dawydoff, February 11, 1930 (Gislén, 1936]. North of central Java (lat. 5°22’ S., long. 109921’ E.); November 21, 1907 [A. H. Clark, 1933] (4, Buitenzorg Mus.). North of western Java (lat. 5°22’ S., long. 107°42’ E.); 23 meters; November 15, 1907 [A. H. Clark, 1933] (2, Buitenzorg Mus.). Off Cape Jabung, northern coast of southern Sumatra (lat. 1°03’ S., long. 104°35’ E.; July 3, 1908 [A. H. Clark, 1933] (2, Buitenzorg Mus.). Singapore; Svend Gad [A. H. Clark, 1909, 1911, 1912] (6, C.M.). Northeastern Australia; Prof. J. Beete Jukes; H.M.S. Fly, 1843-1847 [A. H. Clark, 1913] (1, B.M.). Alert station 144 (probably Thursday Island) [Bell, 1884]. Alert; Torres Strait; 18 meters; sand; also without mention of depth [Bell, 1882, 1884; P. H. Carpenter, 1888; A. H. Clark, 1911, 1912, 1913] (16, B. M.). Alert; Prince of Wales Channel, Torres Strait; 9-13 meters*; 13 meters, sand; 13-16 meters; and without mention of depth [Bell, 1882, 1884; P. H. Carpenter. 1888; A. H. Clark, 1911, 1912, 1913] (74, B. M.). Challenger station 186; Prince of Wales Channel (lat. 10°30’ S., long. 142°18’ E.); 15 meters; coral mud; September 8, 1874 [P. H. Carpenter, 1888; A. H. Clark, 1911, 1912, 1913] (2; B. M.). * Given in Mem. Australian Mus., vol. 4, 1911, p. 765, as ‘Torres Strait, in 5-7 fathoms.” A MONOGRAPH OF THE EXISTING CRINOIDS Dial: Challenger station 187; off Booby Island (lat. 10°36’S., long. 141°55’ E.); 11 meters; coral mud; September 9, 1874 [P. H. Carpenter, 1888].4 Albany Island [A. H. Clark, 1912]. Three miles north by west of Double Island Point, Queensland (South of Mary- borough); 46-47 meters [H. L. Clark, 1916] (1, Australian Mus.). Ten miles southwest of Mapoon; 18 meters [A. H. Clark, 1911] (3, Australian Mus.). Port Curtis, Queensland [A. H. Clark, 1911, 1912] (2, Australian Mus.). Masthead Island, Capricorn group (off Port Curtis) [A. H. Clark, 1911] (1, Australian Mus.). Norwest Islet, Capricorn group; 11 meters; Boardman and Livingstone, December 1930 [H. L. Clark, 1938]. Alert; Dundas Strait (entrance to Van Diemen Gulf) northern Australia; 31 meters; mud [Bell, 1882, 1884; P. H. Carpenter, 1888; A. H. Clark, 1911, 1912, 1913] @._B: M.). Alert station 160; Arafura Sea; 58-66 meters; mud, sand, and shells [Bell, 1882, 1884; P. H. Carpenter, 1888; A. H. Clark, 1912, 1913] (1, B. M.). Baudin Island, northwestern Australia [Bell, 1894; A. H. Clark, 1911, 1912, 1913] (1, B. M.). Baudin Island; 15-27 meters [A. H. Clark, 1929] (1, B. M.). Holothuria Bank, northwestern Australia; 44 meters [Bell, 1894; A. H. Clark, 1911, 1912, 1913] (6, B. M.). Holothuria Bank; 69 meters [Bell, 1894; A. H. Clark, 1911, 1912, 1913] (4, B. M.). Northeast of Holothuria Bank; 27-36 meters [Bell, 1894; A. H. Clark, 1911, 1912, 1913] (1, B. M.). Northwestern Australia [A. H. Clark, 1912]. This refers to the 5 preceding records. Vicinity of York Sound, northwestern Australia (lat. 14°50’ S., long. 125°40’ E.) [A. H. Clark, 1913] (1, B. M.). Broome, Western Australia; dredged at various points in 9-15 meters; H. L. Clark, June 1932 [H. L. Clark, 1938]. Pearl Shoal, off Broome; 9-15 meters; H. L. Clark, September 1929 [H. L. Clark, 1938]. Dr. Eric Mjéberg’s station 7; northwestern Australia; Cape Jaubert 45 miles westsouthwest; 13 meters; July 15, 1911 [Gislén, 1919]. Dr. Eric Mjoberg’s station 11; northwestern Australia; Cape Jaubert 45 miles westsouthwest; 22 meters; July 17, 1911 [Gislén, 1919]. Challenger; Aru (Arrou) Islands [P. H. Carpenter, 1888; A. H. Clark, 1912, 1913] (3, B. M.). Siboga station 273; anchorage off Pulu Jedan, eastern coast of the Aru Islands (pearl banks); 13 meters; sand and shells; December 23-26, 1899 [A. H. Clark, 1918] (7, U.S.N.M., E. 468; Amsterdam Mus.). 4 This locality is not included in the list of localities given under Antedon cariipinna in the Challenger report, nor is the species given in the list of species secured at station 187. But on page 97 of that report, under Antedon multiradiata, Antedon bidentata was given as “the other species dredged at that locality” [station 187]. 272 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Siboga station 274; off the northeastern coast of the Aru Islands (lat. 5°28’12’’ S., long. 134°53’54’’ E.); 57 meters; sand, shells, and stones; December 26, 1899 [A. H. Clark, 1918] (2, Amsterdam Mus.). Aru Islands; Dr. H. Merton’s station 1; west of Ngaiguli; 14 meters; coarse yellow sand; February 18, 1908 [Reichensperger, 1913). Biliton, Sunda Islands; M. Korotnev, 1885 [Koehler, 1895; A. H. Clark, 1912 (as Sunda Islands)]. Investigator; Pocock Island, southeast of Cape Price, the northernmost point of Great Andaman Island (lat. 13°33’40’’ N., long. 93°00’30’’ E.); 36 meters [A. H. Clark, 1912] (1, I. M.). ?South Nilandu, Maldive archipelago; 38 meters [Bell, 1902]. Erroneous locality —Mergui Archipelago [A. H. Clark, 1911]. This is an error for Andaman Islands. Geographical range-—From the Philippine Islands southward to Australia, reach- ing Double Island Point, Queensland, and Cape Jaubert, Western Australia, and west- ward to the Andaman Islands, and possibly to the Maldive archipelago. Bathymetrical range-—From the shoreline down to 111 meters. The average of 21 records for which a definite depth is given is 35 meters. History —Prof. Ludwig von Graff in 1877 recorded some myzostomes from a comatulid from Bohol in the Philippines, which had been given the manuscript name of Comatula dubia by Prof. Carl Semper, who had collected it a few years before. What this Comatula dubia was is not at all clear. Prof. von Graff referred to it again as Antedon dubia in 1884, and Carpenter later used the name Antedon dubia for the present species as will appear below. Previously, therefore, I have assumed that Semper’s Antedon dubia was the same as Carpenter’s Antedon dubia. But as Carpenter never mentioned the former, did not include the Philippines among the localities from which the species under consideration was known to him, and gave the Aru Islands as the only locality for the form which he called Antedon dubia, | now believe that the dubia of Semper and the dubia of Carpenter must have been two differ- ent species and that the former should not be included in the synonymy of the present species, but instead should be placed in the list of unidentifiable forms. Semper’s dubia is mentioned here, however, because it has so long been, on what I now believe to have been quite insufficient grounds, included in the synonymy of this species. The present species was first described by Dr. P. H. Carpenter under the name of Antedon crenulata on September 26, 1882, his description having been based upon some dried fragments he had studied in the Hamburg Museum. Five days after this, on October 1, 1882, Prof. F. Jeffrey Bell published specific formulas for two new species which he called Antedon decipiens and Antedon irregularis, but he gave no indication of their habitat or of their origin. In the report upon the collections of H. M. S. Alert published in 1884 Professor Bell described and figured Antedon decipiens from specimens secured in the Arafura Sea (32-36 fathoms), Dundas Strait, and Prince of Wales Channel. He provisionally associated with this species, as a variety, two specimens that had been dredged at station 144, identified some pages farther on (page 167) in a footnote under Actinometra robusta as probably Thursday Island. Antedon decipiens, he said, presents some resem- blances to Carpenter’s Antedon pinniformis. In the Alert report he also described A MONOGRAPH OF THE EXISTING CRINOIDS 273 and figured Antedon irregularis from specimens secured in the Prince of Wales Channel and in Terres Strait. He remarked that irregularis has some resemblance to decipiens, and gave what he considered the distinguishing characters. In 1884 Professor von Graff described the myzostomes found on a specimen from Challenger station 186, the name of the host having been given him by Carpenter as Antedon bidentata. This name appeared again in a notice of myzostomes published in the Challenger narrative in 1885, and in a second contribution on the myzostomes by Professor von Graff published in 1887. Carpenter also used it in an article containing an account of the sacculi in the comatulids published in 1887. In the Challenger report upon the comatulids published in 1888 Carpenter gave an exhaustive account of this species—quite the best account that he gave of any comatulid. His description was drawn up from two specimens from Challenger sta- tion 186 and two more collected by the Challenger at the Aru Islands. He used the name Antedon varvipinna for the species, including as synonyms under this name his own Antedon crenulata and the Antedon decipiens and A. irregularis of Bell. Unfor- tunately he overlooked the fact that in the descripion of Antedon variipinna he had said that “the distal ends of the cylindrical joints of the large lower pinnules are raised into slight spines,” which indicates a species quite different from a form with sharply prismatic pinnules such as his A. crenulata and the two forms described by Bell. In his account of Antedon multiradiata (=Zygometra microdiscus) he mentioned that Antedon bidentata had been secured at the same station (station 187), but he did not include this station among the localities under Antedon variipinna. At the very end of his account of Antedon variipinna he said that the single example of A. variipinna, var. 5 (given in the list of varieties as ““Antedon decipiens, var. [Chal- enger]’’) from the Prince of Wales Channel was serving as host to 14 individuals of myzostomes, and that the name of the host was given in Professor von Graff’s report as Antedon bidentata, this being the manuscript name he had applied to it before he had become convinced of its identity with Antedon variipinna or had had the oppor- tunity of identifying it with Bell’s A. decipiens. Von Graff, however, gave the locality of this specimen as station 186 instead of station 187 as it was given by Carpenter. Carpenter mentioned Antedon dubia several times. He gave it as an example of a 10-armed species occasionally varied by the intercalation of I[Br 4(3+4) series, and as a species in which the second and third pairs of pinnules are especially character- ized by their large size, and he figured it under that name on plate 36, figures 1-6. While he had regarded it as a distinct species when the earlier pages of the Challenger report were going through the press, before the account of Antedon variipinna was ready to be printed he had reached the conclusion that it is merely a synonym of that species. Under Antedon variipinna he said that he had formerly referred the two individuals from the Aru Islands to a new species; Antedon dubia, and discussed them at some length. In 1894 Prof. Jeffrey Bell recorded this species, under the name of Antedon variipinna, from northwestern Australia in 9-38 fathoms, and in 1895 Prof. René Koehler under the same name recorded a specimen collected at Biliton by M. Korotney in 1885 and gave notes on the specimen. Professor Bell in 1902 recorded it from South Nilandu in the Maldives. 274 BULLETIN 82, UNITED STATES NATIONAL MUSEUM In my first revision of the old genus Antedon published in 1907 I assigned this species to the new genus Himerometra, calling it Himerometra variipinna, and in 1908 I recorded a specimen from Albatross station 5157 under that name. In a revision of the genus Himerometra published in 1909 this species was re- ferred to the new genus Amphimetra as Amphimetra varvipinna. Later in the same year it was recorded from Singapore under this name and notes were given on the specimens. In 1909 I described a new species based upon two specimens from Albatross station 5157 which I called Craspedometra aliena; this supposed new species later proved to be nothing but very slender examples of this species. In 1911 I recorded specimens from Albatross stations 5358 and 5481 under the name of Amphimetra variipinna, and in the same paper, under the heading Craspe- dometra anceps, said that an examination of the Challenger types in London had shown me that Carpenter’s Antedon clemens is the same thing as his A. anceps and my Craspedometra aliena. In a memoir on the recent crinoids of Australia published in 1911 I recorded and gave notes upon a number of specimens in the collection of the Australian Museum, and in giving the range of the species I included the details of the localities repre- sented in the British Museum. In this report the locality Mergui Archipelago was given by mistake for Andaman Islands, Pocock Island, taken from the label of an unrecorded specimen in the Indian Museum, being the place referred to. In a memoir on the collection of the Hamburg Southwest Australian Expedition pub- lished at the same time I included a detailed summary of the distribution of this species on the Australian coasts, also including the data from the specimens in the British Museum. In my memoir on the crinoids of the Indian Ocean published in 1912 I recorded and gave notes upon a specimen from Pocock Island in the Andamans, and included in the synonymy of this species for the first time Comatula dubia von Graff, 1877, Antedon bidentata, and Antedon dubia P. H. Carpenter, 1888. A complete list of the known localities was given. In the same year, having examined the type of Carpenter’s Antedon variipinna at the Hamburg Museum in 1910, I published a redescription of it showing that it does not represent the species previously called variipinna, but instead, so far as I could see, represents the species subsequently called Antedon brockii by Hartlaub. I therefore called this species Amphimetra crenulata, remarking at the same time that I had unfortunately overlooked the type of this form while at Hamburg. In a supplementary report on the crinoids collected by the Hamburg Southwest Aus- tralian Expedition published in 1913 I repeated this information. Dr. August Reichensperger in 1913, misled by my erroneous disposition of Craspedometra aliena in 1911, determined as Craspedometra anceps an interesting specimen collected by Dr. H. Merton in the Aru Islands. He described this speci- men in detail and figured the proximal pinnules. He was somewhat doubtful in regard to the correctness of this determination, and was rather of the opinion that it would turn out to be a new species for which he suggested the name Craspedometra aruensis. A MONOGRAPH OF THE EXISTING CRINOIDS 25 Under the name Amphimetra crenulata I published in 1913 notes on the 17 lots of specimens, all but one of which had previously been studied by Bell or Carpenter, or both, in the British Museum, and in 1915 I again discussed the distribution of this species on the Australian coasts. In 1916 Dr. Hubert Lyman Clark recorded a fine adult from off Double Island Point, Queensland, the southernmost known locality on the eastern coast of Australia. In the report on the unstalked crinoids of the Siboga expedition published in 1918 I assigned crenulata to the genus Heterometra, and in a footnote to the key to the species of Heterometra said that crenulata includes Comatula dubia von Graff, 1877, Antedon decipiens and A. irregularis Bell, 1884, Antedon bidentata von Graff, 1884, Antedon dubia P. H. Carpenter, 1888, and Craspedometra aliena A. H. Clark, 1909, with the exception of the 10-armed specimen mentioned in the description of the last named. I also said that this species has usually been recorded as Antedon (or Amphimetra) variipinna, but it is not Antedon variipinna P. H. Carpenter, 1882. Under Heterometra crenulata I recorded and gave notes on nine specimens from Siboga stations 273 and 274. Dr. Torsten Gislén in 1919 recorded and gave notes on specimens from Dr. Eric Mjéberg’s stations 7 and 11 in northwestern Australia. He called the species Amphimetra vartipinna and included in the synonymy Antedon variipinna P. H. Carpenter, 1882, A. crenulata P. H. Carpenter, 1882, A. decipiens Bell, 1884, A. irregularis Bell, 1884, A. dubia P. H. Carpenter, 1888, and A. brockii Hartlaub, 1890, and discussed the status of variipinna and crenulata at considerable length; but he overlooked the fact that Carpenter mentioned the cylindrical segments of the large lower pinnules in variipinna. Under the name Amphimetra crenulata he discussed the structure of this species at considerable length in 1924. In 1929 I recorded a specimen, as Heterometra crenulata, from Baudin Island in 8-15 fathoms. This specimen is in the British Museum. In 1933 I recorded and gave notes on nine specimens in the Buitenzorg Museum from Borneo, Java, and Sumatra. In 1936 Gislén recorded and gave notes on a specimen from Tonkin, another from Annam, and a third from Cambodia. In 1938 Dr. H. L. Clark recorded and gave notes on 48 specimens collected by himself at Broome, Western Australia, and 2 collected by Boardman and Livingstone at the Capricorn Islands, Queensland. HETEROMETRA PRODUCTA (A. H. Clark) PLATE 26, Fiagure 106 Himerometra producta A. H. Cuark, Proc. Biol. Soc. Washington, vol. 21, 1908, p. 224 (description; Singapore). Amphimetra producta A. H. Cuark, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 7 (listed); Vid. Medd. Nat. Foren. Kgbenhayn, 1909, p. 157 (description; Singapore), p. 193 (collected at Singapore by Svend Gad); Ann. Mag. Nat. Hist., ser. 8, vol. 10, 1912, p. 31 (compared with A. [Heterometra] propinqua); Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p. 17 (compared with A. variipinna [Heterometra crenulata]); Crinoids of the Indian Ocean, 1912, p. 40 (= Antedon laevissima Bell, 1902, in part, and Antedon milberti Bell, 1902, in part), p. 111 (synonymy; Singapore; Fadiffolu, Maldives); Smithsonian Mise. Coll., vol. 61, No. 15, p. 23 (Fadiffolu; characters) . 276 BULLETIN 82, UNITED STATES NATIONAL MUSEUM ? Antedon laevissima Brit. Mus., MS. A. H. Cuark, Smithsonian Mise. Coll., vol. 61, No. 15, 1913, p. 23 (Fadiffolu, Maldives). Heterometra producta A. H. Cuarx, Unstalked crinoids of the Siboga-Exped., 1918, p. 76 (in key; range).—GIst6N, Kungl. Fysiogr. Siallsk. Handl., new ser., vol. 45, No. 11, 1934, p. 22.—A. H. Cuark, John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1936, p. 99 (range), p. 103. Diagnostic features—The enlarged proximal pinnules are rather slender and become flagellate distally; their basal segments are sharply carmate, and the seg- ments from the third onward have the entire distal end somewhat produced and over- lapping, the outer portion of the prismatic ridges being produced into a blunt spine which is directed obliquely forward. The cirri have 23-26 segments of which the longest are half again as long as broad, and the outer are about as long as broad. The 10-13 arms are 50 mm. long. Description.—The centrodorsal is low hemispherical, almost discoidal; the polar area is moderate in size with a raised and roughened center. The cirrus sockets are arranged in two crowded alternating rows, beyond which are one or two rows of more or less obliterated sockets. The cirri are XVI, 24-26, slender, 15 mm. long. The first three segments are very short, the fourth is not quite twice as broad as long, the fifth-eighth are half again as long as broad, and those following gradually decrease in length, after the tenth or eleventh being about as long as broad. The ninth and following segments are sharply carinate dorsally, and on some of the cirri bear small subterminal dorsal spines. The opposing spine is terminal, rather large, and arises from the whole dorsal surface of the penultimate segment, its distal edge being a continuation in a straight line of the distal edge of that segment. The terminal claw is somewhat longer than the penultimate segment, rather slender and moderately curved. The radials are short, with the middorsal line parallel to the dorsoventral axis of the animal. The IBr, are very short, four or five times as broad as long. The IBr, (axillaries) are broadly pentagonal, rather over twice as broad as long, rising to a slight smooth rounded tubercle with the IBr,. Three IIBr series are present in the type specimen, all 4(3-++4). The 13 arms in the type specimen are 50 mm. long. The first two brachials are wedge-shaped, the next six are oblong, three or four times as broad as long, and those following are short wedge-shaped, after about the middle of the arm becoming oblong again and very short. P, is slender, becoming flagellate distally, 5 mm. long with 17-20 segments of which the first is not so long as broad, the second-fifth are about as long as broad, and those following become elongated and about twice as long as broad distally; the first seven or eight segments are somewhat carinate. P, is similar to P,, 7 mm. long, with the distal segments somewhat longer and the carination of the basal segments more marked. In addition to the dorsal carination there is a lateral carination, which is only slightly marked on P,, especially evident toward the end of each segment. P; is comparatively stout and stiff, 8 mm. long with 12 segments of which the first is not quite so long as broad, the second is about as long as broad, and those following become elongated, being about three times as long as broad distally. The second-fifth seg- ments have a sharp dorsal keel, and from the third onward the median portion of the outer surface of the distal ends of the segments is produced, forming a moderate lateral Lol A MONOGRAPH OF THE EXISTING CRINOIDS 277 process of the type seen in H. crenulata, which is repeated, less developed, on the oppo- site side of the pimnule. In the distal portion of the pinnule the entire distal end of the segments is somewhat overlapping. P, and the following pinnules are 4 mm. long with 12 segments increasing in length to the third, which is about as long as broad, then increasing distally to about three times as long as broad, the three or four ter- minal being shorter again. The distal pinnules are 6 mm. long. The color in alcohol is flesh color, clouded with violet. Notes.—I am not wholly convinced that the specimen from the Maldive Islands should be referred to this species. It has only 10 arms. There are 23-25 cirrus segments, which from the tenth onward bear small dorsal spines. The segments of the proximal pinnules have slightly projecting and spinous distal ends. It is possible that the specimens from the Aru Islands described by Reichensperger under the name of Amphimetra variipinna (see page 281) should be referred to this form, as well as the specimen he described under the name of Craspedometra anceps (C. aruensis, n. sp.?). Remarks.—This species appears to be most closely related to H. crenulata, but it differs in its cirri, which are more slender, especially distally, with rudimentary dorsal spines and longer terminal segments, the total number of the segments being less; in its lower pinnules, which are not nearly so much enlarged and of which the com- ponent segments have less produced distal edges; and in its general build, which is considerably more slender. The general ruggedness so very characteristic of H. crenulata is quite absent in this species, and the overlapping of the brachials is much less marked. Localities Singapore; Svend Gad [A. H. Clark, 1908, 1909 1918] (1, C. M.). Fadiffolu, Maldive archipelago; J. Stanley Gardiner [A. H. Clark, 1912, 1913] (1, B. M.). History—This species was described in 1908 under the name of Himerometra producta from a single specimen from Singapore. On the establishment of the new genus Amphimetra in 1909, producta was assigned to it. Later in 1909 the features distinguishing it from H. crenulata were given in detail. In 1912 it was compared with H. propinqua (see below), and in another paper with the type specimen of H. variipinna (see p. 278). In 1913 it was recorded, as Amphimetra producta, from the Maldive Islands on the basis of a specimen in the British Museum that had been labeled ?Antedon laevissima by Prof. F. Jeffrey Bell. In the report on the unstalked crinoids of the Siboga expedition published in 1918 producta was transferred to the genus Heterometra. HETEROMETRA PROPINQUA (A. H. Clark) Puate 25, Figures 97-100 Amphimetra propinqua A. H. Cuarx, Ann. Mag. Nat. Hist., ser. 8, vol. 10, 1912, p. 31 (description; Siboga station 318). Heterometra propingua A. H. Cuarx, Unstalked crinoids of the Siboga-Exped., 1918, p. vii (dis- covery by the Siboga and its significance), p. 76 (in key; range), p. 80 (detailed description; stations, 318, 320), p. 276 (listed), pl. 18, figs. 33, 34.—GisLin, Kungl. Fysiogr. Sallks. Handl., new ser., vol. 45, No. 11, 1934, p. 22. Diagnostic features —This species very closely resembles H. producta, but the cirri are slender and tapering with all the component segments approximately subequal in 278 BULLETIN 82, UNITED STATES NATIONAL MUSEUM length, about twice as long as broad at the ends. The cirri have 24-36 (usually 30-33) segments. The 10-13 arms are 90-120 mm. long. Description —The cirri are VIII-XIII, 24-36 (usually 30-33), 26 to 32 mm. (usually about 30 mm.) long, very slender, tapering gradually to the end of the proximal third and being especially slender from that point onward, recalling the cirri of Leptometra celtica. All the cirrus segments are approximately subequal in length, about twice as long as broad at the ends, though those in the distal third of the cirri are slightly carinate, which makes them appear slightly shorter, and those in the proximal half are slightly longer with slightly expanded ends. On the outermost segments there is a slight indication of dorsal tubercles. The 10 to 13 (usually 11-13) arms are 90-120 mm. long. P; is 7.5 mm. long, with 18 segments, which, gradually increasing in length, become about as long as broad on the fourth or fifth and twice as long as broad ter- minally; the second-eighth have a strong, though rounded, supplementary ridge. P, is 10.5 mm. long, with 17 segments, slightly stouter basally than P,, tapering less gradually and composed of longer segments; there is a supplementary ridge on the second-ninth segments. PP; is similar to P,, 9 mm. long. P, is 6 mm. long, with 11 segments and has a supplementary ridge on the second-fourth or -fifth. P. and P, are not stouter basally than the succeeding pinnules, but they taper with slightly less rapidity; the distal angles of their component segments are only slightly swollen and produced. ‘The distal pinnules are 10 mm. long and are composed of 20 segments. The color in alcohol is brownish white with occasional narrow bands of rusty on the arms; the cirri are white. Notes.—The characters of the six specimens from Siboga station 318 are included in the preceding description. One of the specimens from Siboga station 320 has 11 arms 120 mm. long with a single I1Br 4(3+-4) series; the cirri are 30 mm. long and are composed of 33 segments. The other is smaller with 12 arms about 55 mm. long, both the IIBr series being 2, and the cirri 20 mm. long. Remarks.—This species is most closely related to H. producta, but it is even more slender and delicate. The elongate distal cirrus segments serve to distinguish it at once. Localities.—Siboga station 318; south of eastern Borneo (lat. 6°36’30” S., long. 114°55’30”’ E.); 88 meters; fine yellowish gray mud; February 22, 1900 [A. H. Clark, 1912, 1918] (6, U.'S.N.M., E. 431; Amsterdam Mus.). Siboga station 320; south of eastern Borneo (lat. 6°05’00’’ S., long. 114°07/00’ E.); 82 meters; fine yellowish gray mud; February 22, 1900 [A. H. Clark, 1918] (2, U.S.N.M., E. 420; Amsterdam Mus.). History.—This species was first described in 1912 from specimens from Siboga station 318. In 1918 it was redescribed and figured, and specimens were recorded also from Siboga station 320. HETEROMETRA VARIIPINNA (P. H. Carpenter) Priate 26, Ficures 104, 105 Antedon variipinna P. H. Carpentrr, Journ. Linn. Soc. (Zool.), vol. 16, 1882, p. 506 (description; Canton, China); Proc. Zool. Soc. London, 1882 (1883), p. 746 (specific formula); Challenger A MONOGRAPH OF THE EXISTING CRINOIDS 279 Reports, vol. 26, Zoology, 1888, p- 256 (specimen from Canton; the description and other records refer to H. crenulata)—Harriaus, Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 17 (characters; based chiefly upon A. crenulata). Antedon brockit Hartuaus, Nachr. Ges. Gottingen, May 1890, p. 183 (description; Amboina) ; Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 11 (collected by Brock at Amboina), p. 18 (in key), p. 23 (detailed description and comparisons), p. 113 (in Géttingen Mus.), pl. 1, figs. 4,, 12, 18, pl. 2, fig. 17— Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1582 (listed) —A. H. Cuarx, Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p. 16 (synonym of Antedon variipinna P. H. Carpenter, 1882); Crinoids of the Indian Ocean, 1912, p. 36 (= Heterometra brockit); Unstalked crinoids of the Siboga-Exped., 1918, p. 76, footnote 2 (= Heterometra variipinna). Himerometra brockti A. H. Cuarx, Smithsonian Mise. Coll., vol. 50, 1907, p. 356 (listed). Himerometra vartipinna (part) A. H. Cuarx, Smithsonian Mise. Coll., vol. 50, 1907, p. 356 (listed). Amphimetra vartipinna A. H. Cuarx, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 7 (listed [the species contemplated is H. crenulata]); Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p- 16 (type specimen redescribed; includes Antedon brockii; not variipinna of authors, which is covered by the name crenulata); Crinoids of the Indian Ocean, 1912, p. 110 (in part; record from Canton; other records refer to H. crenulata); Die Fauna Siidwest-Australiens, vol. 4, Lief. 6, 1913, p. 314 (Australian records all refer to H. crenulata); Proc. Biol. Soc. Washington, vol. 26, 1913, p. 179 (range in east Asia).—ReIcHENSPERGER, Abh. Senck. naturf. Ges., vol. 35, Heft 1, 1913, p. 92 (Aru Islands; details of localities and notes) —A. H. Cuarx, Journ. Washington Acad. Sci., vol. 5, 1915, p. 214 (Malayan species; range and its significance); MorrENnsEn, Hong Kong Nat., Suppl. No. 3, 1934, p. 5 (Canton). Heterometra brockit A. H. Cuanrk, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 11 (listed); Crinoids of the Indian Ocean, 1912, p. 36 (identity), p. 127 (synonymy; Amboina). Heterometra variipinna A. H. Cuarx, Unstalked erinoids of the Siboga-Exped. 1918, p. 76 (in key; range; synonymy).—GisLEéN, Kungl. Fysiogr. Sillsk. Lund Férh. vol. 7, No. 1, 1936, p. 10 (relation to H. crenulata). Antedon brocki Gisu£N, Kungl. Svenska Vet.-Akad. Handl., vol. 59, No. 4, 1919, p. 23 (synonym of varitpinna). Diagnostic features.—The enlarged lower pinnules are stout and are composed of cylindrical segments of which the distal ends are everted and finely spinous; in fully developed specimens P, is the longest pinnule and is composed of 22-35 segments. The cirri are 17-30 mm. long and composed of 28-38 segments of which a few are about as long as broad or slightly longer than broad, and the outer are about one-third again as broad as long; long and sharp dorsal spines are developed from the ninth-twelfth onward. The 20-28 arms are 50-140 mm. long. The disk is more or less extensively plated, especially on and about the anal tube. Description.—According to Carpenter the centrodorsal is a moderately thick convex disk with the cirri in a double row. The cirri are about XXV, about 30. The segments are tolerably uniform, the fifth being about as long as broad, the next two or three a trifle longer than broad, and those following shorter again. From about the twelfth onward the segments have rather sharp dorsal spines, that on the penultimate (the opposing spine) being considerably larger than its predecessors. The radials are partially visible. The IBr, are oblong, short and broad, partially united laterally. The IBr, (axillaries) are also short, but little longer than the IBr,, and pentagonal with very open angles. The IIBr series are 4(3+-4), the elements having somewhat uneven edges. The IIIBr series are 2. There are 23 arms probably about 50 mm. long. The first segments after each axillary are rhomboidal and are closely united laterally. The second brachials are 280 BULLETIN 82, UNITED STATES NATIONAL MUSEUM bluntly wedge-shaped. The first syzygial pair (composed of brachials 3+4), and the four next brachials are short and oblong. The following brachials are short, bluntly wedge-shaped, and overlap rather strongly so as to give the arms a serrate appear- ance; the brachials become more oblong again at about the middle of the arm. Syzygies occur between brachials 3-++4, again from between brachials 20+-21 to between brachials 26-+27, and distally at intervals of 8-13 (usually 10 or 11) muscu- lar articulations. The size of the lower pinnules varies considerably according as they are on the outer or inner arms of the postradial series. Pp is moderately long and stout at the base but tapers rather rapidly. P, on the outer arms is both longer and stouter than Pp, sometimes very much so, while P, is equal to, or slightly longer than, P,. P, and P, are both smaller than the corresponding pinnules on the outside of the arm, and the size decreases to P, and Py, after which it gradually increases again. On the inner arms of the postradial series the fifth brachial usually bears the largest pinnule, those on the second and seventh brachials being about equal. The distal ends of the cylindrical segments of the large proximal pinnules are raised into slight spines. The disk is naked and considerably incised; it is 10 mm. in diameter. Sacculi are closely set along the pinnule ambulacra. The color is purple with whitish bands. Notes—My notes on Carpenter’s type specimen, which I examined at the Hamburg Museum in 1910, are as follows: The centrodorsal is moderate in size with a slightly convex dorsal pole 2 mm. in diameter. The cirri are about XX, 28-33, from 17 to 20 mm. in length. The longer proximal segments are about as long as the distal width, sometimes very slightly longer. The outer segments are about one-third again as broad aslong. The earlier segments have slightly thickened and prominent distal dorsal edges. The ninth-twelfth (usually the latter) and following develop long and sharp dorsal spines. Of the 10 IJBr series 7 are 4(3+4) and 3 are 2. There are present 3 II] Br 2 series, all internally developed. The 23 arms were probably about 50 mm. long. The longest pinnules have about 22 segments. The distal edges of the third and following segments of the proximal pinnules are strongly everted and very finely spinous instead of being with broad lateral processes as in H. crenulata. The disk is completely covered with a pavement of small irregular plates. The type specimen of Antedon brockii from Amboina was described by Hartlaub in the following terms: The centrodorsal is rather large and thick with a bare flat dorsal pole. The cirri are arranged in two and a partial third marginal rows. Cirri about XXX, 30-37, reaching 30 mm. in length, stout, and not entirely confined to the border of the centrodorsal. Some of the cirrus segments may be a little longer than broad. Strong dorsal spines are developed from the tenth or twelfth segments onward. The radials are only very slightly visible. The IBr, are short, partially united laterally. The IBr, (axillaries) are short and pentagonal. The IIBr series are 4(3+-4). The I11Br series are 2, developed only internally. Feeble synarthrial tubercles are developed, the best developed being between the first two brachials. The 28 arms are about 140 mm. long and composed of short brachials of which the first 9 or 10 have smooth articulations and are longer than those following. The first syzygial pair (composed of brachials 3-++4) is about as long as broad. The few immediately following are discoidal and form with each other feeble alternating eleva- A MONOGRAPH OF THE EXISTING CRINOIDS 281 tions and depressions, rendering the surface of the arm as a whole somewhat uneven. There then follows a row of somewhat wedge-shaped brachials, which soon pass over into discoidal brachials. The brachials remain short to the arm tips. Because of the produced distal ends of the brachials the dorsal profile of the arms is gently serrate. Syzygies occur between brachials 3+4, again from between brachials 23+24 to between brachials 27+-28, and distally at intervals of from 8 to 10 muscular articu- lations. Pp, and P; on arms arising from a [Br axillary, is slender and about 11 mm. long. P, on other arms is almost as long but not quite so stout as P., which reaches 20 mm. and is composed of about 30-35 segments. P3 is markedly shorter, and the length of the two following decreases. P, varies in size but is mostly as long as Py. Sometimes it is almost as long as P,, which is always a good deal shorter than the long, slender P,, which becomes very slender in its outer portion. The distal pinnules reach 12 mm. in length. All the lower pinnules are remarkable for the somewhat produced finely toothed distal edges of their segments. The disk is 15 mm. in diameter and strongly incised. The disk bears conspicuous small conical lateral papillae opposite the first forking of the ambulacral grooves. The anal tube is plated. The color is deep blackish brown with a reddish cast. From Dr. Merton’s station 10 Reichensperger recorded 18 specimens with 11-20 arms from 50 to 80 mm. long. The cirri are XV—-XXIV, 35-38, up to 23 mm. in length. From about the tenth onward the segments are provided with prominent sharp dorsal spines. The radials and the IBr, are easily visible. Synartbrial tubercles are developed on the articulation between the elements of the IBr series. The IIBr series are almost without exception 4(3-+-4); but in one 20-armed specimen there are exceptionally two IIBr 2 series. There are no division series beyond the IIBr series. The distal brachials are exceedingly short, sometimes almost triangular. Syzygies occur between brachials 3+4, from between brachials 13+14 to between brachials 17-+18, and distally at intervals of 11-15 muscular articulations. P; is the longest pinnule and is composed of 11 or 12 segments, which are longer than broad and bear rounded and mostly finely dentate processes on the distal edge. Similar processes are borne by the pinnules of the fifth, sixth, and seventh brachials. The color is varie- gated, broadly banded with light to dark violet and brick red; the cirri are mostly dark violet. In another lot from Merton’s station 10 there are eight specimens representing two color varieties. Two specimens are grayish white with the cirri and ambulacral grooves mostly dark violet. Six specimens are wholly brick-red. There are 12-18 arms which in the largest are 90 mm. long. All the IJBr series examined by Reich- ensperger are 4(3-++4). The third lot from Merton’s station 10 consisted of three large specimens that had been fixed in corrosive sublimate and had been red when alive; three red and violet examples; and 10 calices from animals which in life were cross banded red brown and white, the disk with triangular white patches situated singly between the ambul- acral grooves. From Dr. Merton’s station 13 there were four specimens of which one had the arms 45 mm. long, and the smallest had 11 arms only 30 mm. long, the single IIBr 208244—40-—19 282 BULLETIN 82, UNITED STATES NATIONAL MUSEUM series being 4(3-+-4). The two other specimens had 11 and 14 arms up to 90 mm. in length. The color was dark violet and white, variously divided. From between Batu Kapal and Meriri there were two specimens each with 11 arms, which in one were 90 mm. and in the other 50 mm. long. The larger specimen, in addition to XI normal cirri, possesses a large number of very small flagellate taper- ing cirri—cirri in process of regeneration. From Dr. Merton’s station XII there are three specimens, two with 13 and one with 16 arms, the arm length being up to 70 mm. ‘The color is the same as in those first described from station 10. From Dr. Merton’s station 17 there is a strikingly slender regular 10-armed speci- men with the arms 70 mm. long. The cirri are very short and compressed, as in the others, up to 7 mm. long with 16-18 segments most of which from the base outward are feebly spined. The synarthrial tubercles are scarcely marked, and the distal brachials in comparison with the other specimens are broader and less triangular. The color in life was banded dark violet-brown and yellow. Remarks.—I am not quite sure that Reichensperger’s specimens from the Aru Islands really belong to this species. The description of P; certainly agrees much more closely with P; in H. producta than with P, in H. variipinna, while the number of arms and the size also suggest that species. Localities —Canton, China; Werner [P. H. Carpenter, 1882, 1883, 1888; A. H. Clark, 1907, 1912, 1918; Th. Mortensen, 1934] (1, H. M.). Amboina; Dr. J. Brock [Hartlaub, 1890, 1891; A. H. Clark, 1907, 1912, 1918}. Aru Islands; Dr. H. Merton’s station 10; north of Penambulai; 8 meters; stony bottom; April 2, 1908 [Reichensperger, 1913]. Aru Islands; Dr. H. Merton’s station 13; Sungi Barkai (eastern half); 15 meters; rocky bottom; April 9, 1908 [Reichensperger, 1913]. Aru Islands; between Batu and Meriri; 10 meters; March 30, 1908 [Reichen- sperger, 1913]. Aru Islands; Dr. H. Merton’s station XII; off Mimien; 15 meters; coarse sand; April 8, 1908 [Reichensperger, 1913]. Aru Islands; Dr. H. Merton’s station 17; Sungi Manumbai (Kapala Sungi); 20 meters; rocky bottom; May 5, 1908 [Reichensperger, 1913]. Geographical range-—Known from Amboina and the Aru Islands. The record from Canton needs confirmation as Canton is more or less of a center for “‘curios’’ gathered over a wide area. Bathymetrical range-—From the shoreline down to 20 meters. The average of 6 records is 11 meters. Remarks.—Hartlaub said that at first sight Antedon brockii resembles Antedon ludovici (that is, Heterometra amboinae) in its black color, its centrodorsal, its cirri, and the form of the brachials. But a comparison of the pinnules serves immediately to differentiate the two species, for the carination so characteristic of the proximal pinnules in H. amboinae is entirely lacking in Antedon brockii. He noted that the small conical protuberances on the disk may represent the soft parts of the lowest pinnules, which have become detached from the pinnules and developed independently. Simi- lar protuberances are found to be strongly marked in the type specimen of Lovén’s Hyponome sarsii (= Zygometra microdiscus) in the Hamburg Museum and also on the A MONOGRAPH OF THE EXISTING CRINOIDS 283 figures of Ant. multiradiata Lam. (in reality of P. H. Carpenter) (=Zygometra micro- discus) given in the Challenger report on the stalked crinoids, volume 11, part 32, 1884, plate 55, figures 3, 4. The type specimen of variipinna from Canton represents a species related to H. producta from which it differs in having long spines on the outer cirrus segments in- stead of aslight carination. Itis also a stouter form than producta, the cirri especially being stouter and longer without the distal taper seen in those of producta. The lower pinnules are more cylindrical—less prismatic—than are those of producta, and the spinous eversion of the distal ends of the segments is not increased at the prismatic angles, but is uniform all around the dorsal side. So far as I can see the type specimen of Antedon variipinna P. H. Carpenter differs from the type specimen of Antedon brockii Hartlaub only in its smaller size. In both specimens the disk is plated, which is most unusual in the genus Heterometra. History.—This species was first described by Dr. P. H. Carpenter in 1882 from a specimen labeled Canton, China, in the Hamburg Museum under the name of Antedon variipinna. He gave a specific formula for the species in the succeeding year. In the Challenger report on the comatulids published in 1888 Carpenter placed in the synonymy of Antedon variipinna his Antedon crenulata described at the same time, and Bell’s Antedon decipiens and A. irregularis described in 1884. These last three —crenulata, decipiens, and irregularis—really represent a quite different species with sharply prismatic proximal pinnules bearing conspicuous processes on the distal ends of the prismatic ridges of their component segments. The manuscript names Comatula dubia, Antedon dubia, and Antedon bidentata published by von Graff also apply to this latter species. Subsequent authors have accepted Carpenter’s interpretation of these forms and have without exception used the name variipinna for the species herein called H. crenulata. In 1890 Dr. Clemens Hartlaub described Antedon brockii from a specimen col- lected by Dr. J. Brock at Amboina in 1884-85, redescribing and figuring it in 1891. In 1910 I examined the type specimen of Carpenter’s Antedon vartipinna at Ham- burg and was surprised to find that it was quite a different thing from the species that I, following Carpenter’s Challenger report, had been accustomed to consider as variipinna. Carpenter had said in his original description that the distal ends of the cylin- drical segments of the large proximal pinnules are raised into slight spines, and this was found to be the case. The lower pinnules are very different from those of the other forms which had been considered as synonyms of variipinna. On the other hand, no characters could be found whereby Carpenter’s variipinna, as represented by the type specimen, could be distinguished from Hartlaub’s brockii. So in the author's paper on the crinoids in the Hamburg Museum published in 1912 brockii was placed in the synonymy of variipinna, and from this synonymy all the various forms previ- ously included were removed. In 1913 Dr. August Reichensperger recorded and gave notes upon numerous specimens of this species which had been collected by Dr. H. Merton in the Aru Islands in 1908. In a memoir on the echinoderms of Hongkong published in 1934 Dr. Th. Morten- sen listed Amphimetra variipinna with the locality, Canton, given by Carpenter. 284 BULLETIN 82, UNITED STATES NATIONAL MUSEUM HETEROMETRA ASTYANAX, sp. nov. Piate 34, Fiacures 152-156 Diagnostic features.—The cirri taper conspicuously distally so that the terminal third or fourth is only about half as broad in lateral view as the proximal portion; the cirri are almost smooth, the last 10-15 segments bearing only slight and inconspicuous dorsal tubercles; the distal cirrus segments are always longer than broad, varying from slightly longer than broad to twice as long as broad; and the cirri are long, reach- ing rather more than one-fourth the arm length. The cirri are 30-33 mm. long, with 37-39 segments. The 19 arms are 120 mm. long. * Description "The centrodorsal is discoidal, rather thin, with the broad dorsal pole, which is 5 mm. in diameter, almost flat. The cirri are arranged in a single slightly irregular marginal row. The cirri are XX VI, 37-39, from 30 to 33 mm. long; they are moderately stout at the base and taper slowly distally so that the terminal third or fourth is only about half as broad in lateral view as the proximal portion. The first segment is about twice as broad as long, and those following slowly increase in length, becoming about as long as broad on the twelfth and from half again to twice as long as broad on the last three or four, on which the cirrus usually tapers with increased rapidity; insome cases these last segments taper more slowly, and they may be scarcely longer than broad. The last 10 to 15 segments bear a subterminal dorsal tubercle. which is scarcely perceptible except on the last five or six. The opposing spine is well de- veloped, slender, and directed obliquely forward. The terminal claw is nearly half again as long as the penultimate segment and is slender and moderately curved. The radials are entirely concealed. The IBr, are concealed in the median line but are visible as long narrow triangles extending inward so that their apices are more or less near the midradial line. The IBr, (axillaries) are low triangular, three to four times as broad as long, with the lateral angles truncated and thus forming short straight sides which are somewhat more than half as long as the narrow lateral bases of the triangles representing the TBr,. The elements of the IBr series and the first two brachials are in close lateral contact and are sharply flattened laterally. The I1Br series are 4(3-+4) and the IIIBr series are 2; there are in the type specimen five I1Br and four II)Br series, making 19 arms in all. The 19 arms are 120 mm. long. The brachials beyond the proximal quarter of the arms are exceedingly short and almost oblong, as in the species of Himerometra. P, is 11-13 mm. long and composed of 17-20 segments. The pinnule is rather stout proximally but tapers rapidly at the end of the proximal half and is flagellate in the distal half. A rather prominent ridge runs the entire length of the pinnule on the outer side, and the second-seventh or-eighth segments are provided with strong cari- nate processes having the distal edge straight and parallel to the axis of the pinnule, which make the pinnule appear disproportionately broad basally. The earlier seg- ments are much broader than long, those succeeding becoming about as long as broad on the ninth or tenth and half again as long as broad terminally. P, is 16 mm. long, with 26 segments, of which the earlier are about twice as broad as long, those following becoming as long as broad on the eleventh or twelfth, and one-third again A MONOGRAPH OF THE EXISTING CRINOIDS 285 as long as broad terminally. The pinnule is about as stout basally as P,, but it tapers very gradually and regularly distally and hence appears much stouter. A ridge runs the entire length of the outer side of the pinnule and the first 10 segments bear prominent carinate processes which from the seventh outward decrease slowly in height. P; is 19 mm. long, with 27 segments. It resembles P, and is of about the same width basally, but it tapers less rapidly, especially in the distal half, so that it appears stouter. The carinate processes on the earlier segments are slightly smaller than those on P,. P, is 13 mm. long, with 25 segments. It is slightly less stout basally than P; and tapers gradually, becoming very slender in the terminal third. The earlier segments are about twice as broad as long, those following becoming about as long as broad on the eleventh and twice as long as broad terminally. As in P, the lateral ridge is conspicuous and is more or less tuberculated. The second-seventh segments bear narrow, though very prominent, carinate processes. P; is 10 mm. long, with 22 segments, and is much smaller and more slender than the preceding pinnules. The carinate processes on the second-seventh segments are rather more conspicuous than they are on the preceding pinnules, causing the pinnule to appear rather disproportionately broad in the basal half. The succeeding pinnules are similar, the next three or four slowly decreasing in size. Locality.—Danish Expedition to the Kei Islands; Dr. Th. Mortensen; station 66; off western Java (lat. 5°54’ S., long. 106°12’ E.); 24 meters; sand and shells; July 27, 1922 (1, C. M.). Remarks.—This species seems to approach Craspedometra in the marked terminal taper of the cirri. It is almost intermediate between Craspedometra acuticirra and such species of Heterometra as H. amboinae. HETEROMETRA AFFINIS (Hartlaub) Antedon affinis HartLtaus, Nachr. Ges. Géttingen, May 1890, p. 184 (description; Amboina) ; Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 11 (collected by Brock at Amboina), p. 18 (in key), p. 25 (detailed description and comparisons; Amboina), p. 113 (in Géttingen Mus.), pl. 1, fig. 14, pl. 2, figs. 18, 24-—Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1582 (listed) —A. H. Cuark, Crinoids of the Indian Ocean, 1912, p. 37 (identity). Himerometra afinis A. H. Cuark, Smithsonian Misc. Coll., vol. 50, 1907, p. 356 (listed). Heterometra affinis A. H. Cuark, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 11 (listed) ; Crinoids of the Indian Ocean, 1912, p. 37 (identity), p. 127 (synonymy; habitat); Unstalked crinoids of the Siboga-Exped., 1918, p. 79 (in key; range), p. 82 (station 99; notes), p. 272 (listed).— Gist£n, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, p. 22. Diagnostic features.—The enlarged proximal pinnules are smooth with the earlier segments carinate, and P, is the largest and longest. The longest cirri have not more than 30 segments of which a few are slightly longer than broad and the outermost bear small dorsal spines. The 15-18 arms are 65-120 mm. long and are composed of wedge-shaped brachials. Description —The centrodorsal is a rather small disk with the cirri arranged in two irregular rows. The cirri are not confined to the margin but leave only a rather small irregular central portion of the dorsal surface bare. The cirri are about XXIV, 20-30, the longest 20 mm.long. A few of the segments are slightly longer than broad. ‘The distal segments are strongly compressed laterally 286 BULLETIN 82, UNITED STATES NATIONAL MUSEUM and keeled; they strongly overlap the bases of the succeeding segments so that here the dorsal profile of the cirri appears serrate; the outermost segments have small dorsal spines. The radials are only partially visible. The IBr, are partially united laterally. The IBr, (axillaries) are rather short and pentagonal. One of the [Br series bears two undivided arms. The four others bear on one side an undivided arm and on the other a I1Br 4(3+4) series. Each IJBr series bears a single IJ[Br 2 series. The first elements of the I1Br series are short and of the same length as the first brachials of the adjacent arms with which they are laterally united. The 18 slender and rather smooth arms are apparently about 65 mm. long. The basal brachials are short, the first being shorter than the second. The first syzygial pair (composed of brachials 3-+-4) is also short. The following brachials are short disks with feebly produced ends. From the eighth or ninth brachial onward there is a series of brachials that are triangular and a good deal larger; about the thirtieth the brachials become shorter and blunter, wedge-shaped, farther on short discoidal, and finally more squarish. Syzygies occur between brachials 3-+4, again from between brachials 11-+-12 to between brachials 14+-15, and distally at intervals of from 9 to 11 muscular articula- tions. The second syzygy in arms which arise from the IBr axillaries is between brachials 8+-9 or 9-+10, and the intersyzygial interval beyond is 5 or 6 muscular articulations. The lowest pinnules, whether on the IIBr series or on the arms, are very small and delicate with carinate and rather large basal segments. P, when following a pinnule on a I1Br series is a good deal larger than these and measures about 9 mm. in length. P, is very small. The largest pinnule is P,, which measures 12 mm. The pinnules following decrease in length to Ps. The pinnule of the sixth brachial (P,) is smaller than that of the fifth (P,). The pinnule of the seventh brachial (P,) on arms arising directly from a [Br axillary is almost as long as that of the fifth (P,). The pinnules of about the first 20 brachials have a few carinate basal segments. The length of the succeeding pinnules reaches 7 mm. The disk is 13 mm. in diameter, and is naked and incised. Sacculi are very numerous on the arms and pinnules. The color is light chocolate-brown, with the disk dark brown. The preceding description is adapted from Hartlaub’s description of the type specimen. Notes —One of the specimens from off Jolo has 18 arms about 120 mm. long. The other is very smell with 11 arms, one ITBr 4(3+4) series being developed. The specimen from Siboga station 99 has 15 arms about 70 mm. long; each IBr series bears on the right hand side (as viewed dorsally) a IIBr 4(3+-4) series; the two additional arms on the left anterior ray are as yet only 22 mm. in length, measured from the axillary. The longest cirrus has 29 segments. This appears to be a young individual just completing the adolescent autotomy. Of the specimens from Singapore one has 16 arms 50 mm. long and another has 15 arms 55 mm. long. Localities —Dr. Th. Mortensen’s Pacific Expedition 1914—’16; off Jolo, Philip- pines; about 36-55 meters; sand and coral; March 19, 1914 (2). A MONOGRAPH OF THE EXISTING CRINOIDS 287 Siboga station 99; anchorage off North Ubian, southern part of the Jolo (Sulu) Archipelago (lat. 6°07'30’’ N., long. 120°26/00” E.); 16-23 meters; lithothamnion bottom; July 28-30, 1899 [A. H. Clark, 1918] (1, Amsterdam Mus.). Singapore; shallow water; Mr. Svend Gad (3, U.S. N. M., E. 1082; C. M.). Amboina; Dr. J. Brock [Hartlaub, 1890, 1891; A. H. Clark, 1907, 1909, 1912]. Geographical range-—From the Philippines to Singapore and southward to Amboina. Bathymetrical range —From the shoreline down to 36 (255) meters. History —This species was first described under the name of Antedon affinis by Dr. Clemens Hartlaub in 1890, and was redescribed and figured by him in the following year. He remarked that this new species shows much similarity to Antedon biparti- pinna P. H. Carpenter and A. ludovici P. H. Carpenter—apparently representatives of the same species—which are also forms combining the presence of IIIBr series with slender carinate pinnules in the proximal arm region. As his specimen was not sexually mature Hartlaub said it was not impossible that the new species may prove to be simply a young stage of Antedon ludovici, although to him the characteristic features seem to point more to specific distinctness. He said that Antedon affinis differs from A. ludovici in its much smaller size and in its very slender build. In addition, its cirri have a relatively much finer structure and a smaller number of segments. Also, the synarthrial tubercles so characteristic of A. ludovici are lacking and the general surface of the first eight or nine brachials has only a suggestion of the unevenness which we find so marked in that species. Lastly the triangular brachials that follow the proximal brachials are markedly longer in Antedon affinis than the corresponding brachials in A. ludovici, which also are not triangular. The agreement between the two forms, according to Dr. Hartlaub, lies in the carination of the proximal pinnules and in the form of the proximal brachials, especially that of the first syzygial pair, and finally in the arrangement of the arms and the manner of the division of the post- radial series. In my first revision of the genus Antedon published in 1907 this species was assigned to the new genus Himerometra, and in the revision of the family Himero- metridae published in 1909 it was assigned to the new genus Heterometra. In 1918 a specimen from Siboga station 99 was recorded and notes on it were given. HETEROMETRA PULCHRA A. H. Clark PLATE 36, Ficures 164, 165 Heterometra pulchra A. H. Cuark, Rec. Indian Mus., vol. 7, pt. 3, No. 26, July 1912, p. 269 (nomen nudum; Arrakan coast); Crinoids of the Indian Ocean, November 1912, p. 317 (detailed descrip- tion; localities; comparisons); Unstalked crinoids of the Siboga-Exped., 1918, p. 78 (in key; range); Treubia, vol. 14, livr. 2, 1933, p. 208 (not included in Szboga collection), p. 212 (near Deli).—GisLin, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, p. 22. Diagnostic features——The brachials are distinctly wedge-shaped with the ends never quite parallel and are not exceedingly short. The enlarged lower pinnules are smooth, with the earlier segments keeled. The cirri are 25-35 mm. long with 32-45 segments of which the longest are slightly longer than broad and the outer are slightly broader than long and bear rather small, though sharp and prominent, dorsal spines. The 24-32 arms are 85+ mm. long. P; is the largest and longest pinnule, 14-22 mm. long with 22-25 segments. 288 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Description.—The centrodorsal is low hemispherical, with the bare dorsal pole slightly convex, finely pitted, 3.5 mm. in diameter. The cirrus sockets are arranged in 2 irregular rows. The cirri are XX, 32-39 (usually nearer the latter), 25 mm. long. The longest cirrus segments are slightly longer than broad. The short distal segments are but little broader than long. Rather small, though sharp and prominent, dorsal spines are developed from the twentieth segment onward. The radials are just visible beyond the rim of the centrodorsal. The IBr, are very short, bandlike, about six times as broad as long. The IBr, (axillaries) are nearly three times as broad as long, with the lateral edges slightly more than half as long as those of the [Br,. Eight I1Br series are present in the type specimen, six of which are 4 (3-+4) and two of which are 2. There are six IIIBr series, all 2, and all developed internally except one, which is developed externally by the side of an internal series. The division series are well rounded dorsally, in lateral apposition and laterally flattened as far as Pp, with the apposed edges slightly produced outward. The elements of the I/Br and IJIBr series are interiorly in lateral apposition through slightly produced edges. There are in the type specimen 24 arms about 85 mm. long, perfectly smooth and well rounded dorsally. Pp is 6 mm. long, with 22 short segments, none of which are longer than broad; the second-fifth are strongly carinate dorsally, and from the sixth segment onward the pinnule is rather strongly prismatic. P, is 12 mm. long, with 27 segments, all of which are short, the distal being scarcely twice as long as broad; the second-sixth are strongly carinate, and the pinnule is prismatic from the seventh onward. P, is 13 mm. long, with 25 segments, resembling P, but with the distal segments more elongate; the second-sixth segments are carinate. P3 is the largest and longest pinnule, 14 mm. in length, slightly stouter throughout than P,, with 22 segments, of which the second- seventh are carinate; a supplementary carination extends along the pinnule from the sixth segment to the tip. P, is 9 mm. long, with 18 segments. P; is 6 mm. long. The carination of the earlier segments is very strong on the fifth-seventh pinnules but nearly or quite obsolete from that point onward. The color in alcohol is flesh color, with purple bands at the articulations. The cirri are yellow. Notes.—In the specimen from off Jolo the cirri are 35 mm. long and are com- posed of 39-45 segments. There are 32 arms. All the I] Br series are 4 (3+4), and all the following division series are 2. P; is 22 mm. long and is composed of 24 or 25 segments. This specimen seems to belong to H. pulchra, the slight differences presumably being due to greater size as compared with the type. Remarks.—While in its general habitus this species perhaps most nearly resembles H. savignii, it is undoubtedly most closely related to H. singularis, from which it differs most strikingly in having P; instead of P, the largest pinnule. Localities.—Investigator station 387; southwest of the mouths of the Irrawaddy river, off Cape Negrais, Burma (lat. 15° 25’ N., long. 93° 45’ E.); 73-89 meters; sand and coral; November 16, 1909 [A. H. Clark, 1912] (1, I. M.). A MONOGRAPH OF THE EXISTING CRINOIDS 289 Investigator; Arrakan coast, Burma [A. H. Clark, 1912] (3, U.S.N.M., 35077, 35194, 35212). Near Deli, northern Sumatra (lat. 3° 40’ N., long. 99° 10’ E.); 16-18 meters; June 25, 1908 [A. H. Clark, 1933] (2, Buitenzorg Mus.). Dr. Th. Mortensen’s Pacific expedition, 1914-16; off Jolo, Philippines; about 27 meters; March 21, 1914 (1, C. M.). Geographical range-—From the Philippines to the coast of Burma. Bathymetrical range.—From 27 to 73 (?89) meters. History.—This species was first mentioned in July 1912, when two small and broken specimens were recorded from the Arrakan coast of Burma. It was formally described in November of the same year from a single specimen dredged by the Investigator at station 387. In 1933 I recorded two additional specimens in the Buitenzorg Museum from near Deli, Sumatra. HETEROMETRA GRAVIERI A. H. Clark Heterometra graviert A. H. Cuarx, Proc. U. 8. Nat. Mus., vol. 40, 1911, p. 8 (southeastern Africa), p- 25 (description; Zanzibar); Bull. Mus. Hist. Nat., Paris, 1911, No. 4, p. 251 (detailed descrip- tion; Zanzibar); Crinoids of the Indian Ocean, 1912, p. 127 (synonymy; habitat); Unstalked crinoids of the Szboga-Exped., 1918, p. 78 (in key; range)—GuisLin, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 22, 48.—A. H. Ciark, John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1936, p. 100 (range), p. 104. Diagnostic features——The brachials are distinctly wedge-shaped with the ends never quite parallel and are not exceedingly short. The enlarged lower pinnules are smooth, with the earlier segments keeled. The cirri are 21 mm. long with 36-39 segments of which the longest are from one-third to one-half again as long as broad and those in the distal half are about twice as broad as long and bear short dorsal spines. The 19 arms are 80 mm. long. P, is the longest pinnule and is composed of 24 segments. P; is slightly smaller than P., but much larger than P;. P, is about as large as P). Description —The cirri are XXVIII, 36-39, 21 mm. long, more slender than those of H. joubini. The first segment is short and the following gradually increase in length to the fifth or sixth, which is from one-third to one-half again as long as broad, and after the eighth or ninth begin to decrease again so that those in the distal half of the cirri are about twice as broad as long. The segments in the outer half of the cirri bear short dorsal spines. Nine IIBr series of 4(3-++4) are present. The division series have produced lateral borders. The 19 arms in the type specimen are 80 mm. long. The brachials are only very slightly overlapping. The arm structure resembles that of H. jowbini. P, is small and weak. P, is half again as long as P,, slender, and flagellate distally, with 24 segments of which the second-fifth are carinate. P; is slightly smaller than P,; but much larger than P,. P,is about as large as P;. The following pinnules are slightly shorter than P;. The basal segments of all the lower pinnules are distinctly carinate. Remarks.—Though of the same size as H. joubini, this species is as a whole more slender and delicate. 290 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Locality —Zanzibar; M. Rousseau, 1841 [A. H. Clark, 1911, 1912, 1918] (1, P. M.). History.—Dr. P. H. Carpenter examined this specimen in the Paris Museum, placing with it a label indicating that it was an undescribed form differing from H. joubini (see p. 312). It was described in 1911 both in a paper on the recent crinoids in the Paris Museum and in a memoir on the crinoids of the coasts of Africa. HETEROMETRA SINGULARIS A. H. Clark Piate 36, Figures 161-163 [See also vol. 1, pt. 2, fig. 269 (arm and pinnules), p. 207; figs. 455, 456 (pinnule tip), p. 261.] Heterometra singularis A. H. Cuarx, Proc. U. 8. Nat. Mus., vol. 36, 1909, p. 638 (description; south- ern portion of Malacca Strait); Vid. Medd. Nat. Foren. Kgbenhayn, 1909, p. 164 (Singapore; description of a specimen), p. 193 (collected at Singapore by Svend Gad); Crinoids of the Indian Ocean, 1912, p. 128 (synonymy; detailed description; localities), fig. 11, p. 129; Unstalked crinoids of the Siboga-Exped., 1918, p. 78 (in key; range); Journ. Linn. Soe. (Zool.), vol. 36, 1929, p. 638 (80 miles northwest of Penang; notes); Rec. Indian Mus., vol. 34, pt. 4, 1932, p. 551 (listed), p. 556 (Investigator station 548, 549; notes).—GisL6N, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, p. 22. Diagnostic features —The brachials are distinctly wedge-shaped with the ends never quite parallel and are not exceedingly short. The enlarged lower pinnules are smooth, with the earlier segments keeled. The cirri are about 35 mm. long with 36-39 segments of which the longest are about as long as broad and the outer are slightly broader than long and bear conspicuous dorsal spines. The 12—20 arms are 90-150 mm. long. P, is the largest pinnule, P; being of the same length as P,. Description.—The centrodorsal is discoidal, the bare polar area flat, 1.5 mm. in diameter. The cirrus sockets are arranged in a single crowded, more or less irregular, marginal row. The cirri are XVII, 21-25, 12 mm. long. The first segment is short, the second is about twice as broad as long, the third is somewhat longer, and the fourth is about as long as broad. The next two are slightly longer than broad, and those following gradually decrease in length so that the terminal 15 are one-third to one-half again as broad as long. On the seventh subterminal dorsal spines begin to develop and soon become long and prominent. The opposing spine is large and long, much larger than the spines on the preceding segments, triangular, the apex terminal, arising from the whole surface of the penultimate segment and about equal to the width of that segment in height. The terminal claw is nearly twice as long as the penultimate seg- ment, slender, abruptly curved proximally, becoming nearly straight distally. The radials are short, oblong, the dorsal surface with numerous prominent rounded tubercles. The IBr, are short, oblong, slightly over four times as broad as long, in close lateral apposition. The IBr, (axillaries) are broadly pentagonal, almost triangular, twice as broad as long, the lateral edges shorter than those of the IBr. The II Br are 4(3+-4); the ossicles up to and including the second brachial exteriorly and the fourth interiorly, as well as the first two segments of the first three pinnules, are in close apposition and sharply flattened, the lateral edges somewhat produced. The 11 arms are 40 mm. long. The first two brachials are subequal, wedge- shaped, about twice as broad as the exterior length, the first interiorly united. The first syzygial pair (composed of brachials 3+4) is slightly longer interiorly than A MONOGRAPH OF THE EXISTING CRINOIDS 291 exteriorly, nearly three times as broad as the interior length. The next four brachials are oblong, nearly four times as broad as long, those following becoming almost tri- angular, about three times as broad as long, then gradually lengthening (though remaining almost triangular) to about twice as broad as long, and at a point some- what beyond the proximal third of the arm rather quickly becoming wedge-shaped, almost oblong, about two and one-half times as broad as long. From about the ninth onward the brachials have prominent distal ends, though these do not overlap the bases of the succeeding segments. Syzygies occur between brachials 3+-4, 9+ 10, and 15+16 (the second sometimes omitted), and distally at intervals of 7-10 muscular articulations. Pp is 4.5 mm. long, moderately stout basally but tapering rapidly in the proximal half and becoming slender distally, composed of 20segments, which at first are about twice as broad as long, becoming about as long as broad after the eighth; the second-seventh segments are rather strongly carinate. PP, is similar, very slightly longer and stouter. P, is 6 mm. long, considerably stouter and stiffer than the preceding and rather more strongly carinate basally, with about 20 segments of which the first 7 (except for the carinate process) are squarish and the remainder slightly longer than broad, becoming about half again as long as broad distally; the ridge in the distal half of the outer side is but little marked. P,; is 3 mm. long, much smaller than any of the preceding pin- mules, composed of about 12 segments, which at first are broader than long, becoming as long as broad at about the fifth and nearly twice as long as broad distally; the second- fifth segments are carinate. The following pinnules are similar and of about the same length, the segments becoming gradually longer and the basal carination gradually less. The distal pinnules are 5 mm. long. On the arms arising from a [Br axillary P;, P2, and P; are usually as described for Pp, P;, and P2; and P, is much smaller, as described for P;; but occasionally P, is enlarged and similar to P; as described, instead of being small like P,. The disk bears a few calcareous granules in the anal area, especially on the anal tube. Notes.—The specimens from Jnvestigator station 549 have 15 and 27 arms. The specimen from Jnvestigator station 548 has 20 arms. Of the nine IIBr series present six are 4 (3+4) and three are 2. One IIIBr 2 series is present, developed internally on a IIBr 2 series. The cirri are XV, with several additional rudimentary, 41-44, up to 35 mm. in length. One of the six specimens collected by the Patrol 80 miles northwest of Penang may be thus described. The centrodorsal is discoidal, moderately thick, with the bare dorsal pole flat, though with an irregular surface, 4 mm. in diameter. The cirrus sockets are arranged in two alternating marginal rows. The cirri are XXII, 37-39, about 35 mm. long. The first segment is short, and those following gradually in- crease in length to the fifth or seventh, which is about as long as broad. In the longest cirri the next 10 segments are subequal, somewhat longer than broad, and those succeeding slowly decrease in length so that the distal are broader than long, though never much so. On a segment somewhere between the fifteenth and the nineteenth a subterminal dorsal tubercle appears, which, two segments beyond, becomes a con- spicuous dorsal spine and, in lateral view, appears as a narrow triangle with the sharp apex terminal, arising from the entire dorsal surface of the segment and in height 292 BULLETIN 82, UNITED STATES NATIONAL MUSEUM equal to about one-third the width of the segment. The distal edge of the radials is just visible beyond the rim of the centrodorsal; it usually bears a single row of small and regular beadlike tubercles, but it may be simply roughened. The division series are broad, in close lateral contact, and laterally flattened. The lateral borders of the elements of the division series are slightly produced, the production forming a narrow margin with a straight outer edge. Of the eight [Br series six are 4(3+4) and two are 2. The two ILIBr series, which are both developed on the same IBr series, are 2. The 20 arms are 150 mm. in length. The earlier brachials are wedge- shaped, short, twice as broad as the maximum length; beyond the proximal third of the arm the brachials become almost oblong and about four times as broad as the median length. Py is 10 mm. long and is composed of 32 segments. It tapers rapidly on the first nine segments, beyond which it is very slender. The first segment is much broader than long, but those succeeding gradually increase in length so that the eighth and following are about as long as they are broad. The first six segments are sharply carinate, the carination being narrow with the crest straight and parallel with the longitudinal axis of the segments. The first two or three segments are flattened against the corresponding segments of the adjacent pinnule on the neighboring arm. P, is about 13 mm. long and is composed of 34 segments. It tapers rather rapidly in the first eight segments, becoming slender and flagellate distally. The segments become as long as broad on the third to the sixth, then longer than broad, the last 10 being about twice as long as broad. The first seven segments are sharply carinate, the carination being highest on the second and third, and disappearing at the end of the seventh. P, is 19 mm. long and is composed of 33 segments, which become as long as broad on the fourth and on the last 10 are twice as long as broad. The second-seventh segments bear a sharp knifelike carinate process which is narrow with a straight crest parallel to the longitudinal axis of the segments. The pinnule is only moderately stout and tapers with extreme slowness from the base to the tip, which gives it a straplike appearance. P; is 13 mm. long, with 27 segments, which beyond the fifth are all about as long as broad. The second-sixth segments are sharply, though narrowly, carinate. P, is 9.5 mm. long, with 23 segments, less stout at the base than P; and tapering rather rapidly in the proximal half so that the distal half is slender and flagellate. The segments become as long as broad on the sixth and twice as long as broad terminally. The second-sixth segments are narrowly carinate. Another specimen from 80 miles northwest of Penang resembling the preceding has 15 arms, there being four I[Br 4 (3+4) series and one IIIBr 2 series, the latter developed internally. ‘The longest cirri have 36-39 segments. A third specimen from the same locality has 12 arms 90 mm. long, there being present two I] Br 4 (3+4) series. The longest cirri have 32-33 segments. The three other specimens from 80 miles northwest of Penang are more grayish in color and slightly smoother in appearance. One has 14 arms 140 mm. long; there are three I1Br 4 (3-++4) series and a single I1IBr 2 series, the latter internally developed. Another has 13 arms 130 mm. long; there are two I1Br 4 (3+-4) series and one I11Br 2 series, internally developed. A MONOGRAPH OF THE EXISTING CRINOIDS 293 The last example has 13 arms 100 mm. long; there are two IIBr 4 (3+-4) series and one internal IIIBr 2 series. The longest cirri have 41-43 segments and reach a length of 35 mm. Although these six specimens are much larger and better developed than the type- specimen of H. singularis, there can be no doubt but that they represent that species. The specimen from Singapore is in much better condition than the badly broken type specimen. It has 12 arms 65 mm. long, two IIBr 4 (3+4) series being developed on adjacent postradial series. The cirri are XV, 26-31, from 15 to 18 mm. long. This specimen was compared directly with the type and the only difference found was the slightly greater number of cirrus segments in the former. None of the cirri remaining in the type, however, appear to be quite mature. Localities —Investigator station 549; Mergui Archipelago, northwest of King Island (lat. 12°48’00’’ N., long. 98°16/10’’ E.); 44 meters; mud, sand, and shells; October 21, 1913 [A. H. Clark, 1932] (2, I. M.). Investigator station 548; Mergui Archipelago, northwest of King Island (lat. 12°49’23’’ N., long. 98°23’30’’ E.); 42 meters; mud, sand, and shells; October 21, 1913 [A. H. Clark, 1932] (1, I. M.). HKighty miles northwest of Penang; 73 meters; cable repair ship Patrol, Eastern and Associated Telegraph Co., through Colonel Grant, I. M.S.; taken in May, 1923, from a cable laid two years previously [A. H. Clark, 1929] (6, B. M.). Investigator; southern portion of Malacca Strait [A. H. Clark, 1909, 1912] (1, I. M.). Singapore; Svend Gad [A. H. Clark, 1909] (1, C. M.). Geographical range—From the Mergui Archipelago southward to the Malacca Strait, between the Malay Peninsula and Sumatra. Bathymetrical range——From the shoreline down to 73 meters. History.—This species was first described in 1909 from a specimen obtained by the Royal Indian Marine Surveying steamer Investigator in the Malacca Strait. Another specimen collected at Singapore by the Danish consul, Svend Gad, was recorded in the same year. In 1912 the type specimen was redescribed and figured. In 1929 six more specimens were recorded from 80 miles northwest of Penang, where they had been brought up by the cable repair ship Patrol, and in 1932 three more specimens were recorded that had been dredged by the Investigator in the Mergui Archipelago in 1913. HETEROMETRA COMPTA A. H. Clark Puate 33, Ficgures 144-148 [See also vol. 1, pt. 2, fig. 268 (arm and pinnules), p. 207; figs. 453, 454 (pinnule tip), p. 261.) Heterometra compta A. H. Cuarx, Proc. U. 8. Nat. Mus., vol. 36, 1909, p. 636 (description; Pedro Shoal); Crinoids of the Indian Ocean, 1912, p. 124 (synonymy; detailed description; Pedro Shoal), fig. 10, a, b, p. 126; Unstalked crinoids of the Siboga-Exped., 1918, p. 79 (in key; range) ; Smithsonian Mise. Coll., vol. 72, No. 7, 1921, pl. 15, fig. 51 (arm).—Gisi£Nn, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 22, 23, 30. Diagnostic features —The brachials are distinctly wedge-shaped, with the ends never quite parallel, and are not exceedingly short. The enlarged lower pinnules are 294 BULLETIN 82, UNITED STATES NATIONAL MUSEUM smooth, with the earlier segments keeled. The cirri are 23-25 mm. long with 31-35 segments of which the longest are about as long as broad or slightly longer than broad, and the outer are from half again to twice as broad as long and bear dorsal tubercles. The 16-25 arms are 125mm. long. P, and P, are of the same length and similar with 26 segments and are much longer and slightly stouter than Pp. Ps is scarcely longer than Pp. Description.—The centrodorsal is discoidal, with the bare polar area flat, slightly convex or slightly concave, about 5 mm. in diameter. The cirrus sockets are arranged in a single more or less irregular marginal row. The cirri are XVIII-XXIII, 31-35, from 23 to 25mm. long. The first segment is very short, the next three are nearly two and one-half times as broad as long, and those following gradually increase in length to the sixth or seventh, which is about as long as broad. The next five to seven segments are usually slightly longer than broad, sometimes squarish, the following gradually decreasing in length so that the terminal fifteen or rather more are from half again to twice as broad as long. At about the fifteenth segment dorsal tubercles are developed, at first involving only the distal portion of the dorsal surface, later arising in a slightly convex line from near the proximal end, the apex being subterminal; these tubercles are narrow, laterally occupying only a small portion of the median part of each segment, and are slightly rounded dorsally. On the last three segments the tubercles become somewhat sharper, more erect, and move to a median position. The opposing spine is small, though larger than the tubercle on the preceding segment, blunt, arising from the entire dorsal surface of the segment, the apex median or sub-median in position, in height equal to about one-third the lateral width of the penultimate segment. The terminal claw is somewhat longer than the penultimate segment, rather stout and strongly curved. There are no basal rays. The radials are concealed by the centrodorsal. The IBr, are very short and band- like. The IBr, (axillaries) are short, almost triangular, two and one-half times as broad as long. The I1Br are 4 (3+-4), in apposition laterally, though not laterally flattened; the I1Br, are entirely united interiorly. The IIIBr series are 2, rarely 4 (3+4). The IVBr series are 2, but are rarely present. The 16-25 arms are 110 mm. long. The first two brachials are wedge-shaped, three times as broad as long exteriorly, the first interiorly united. The next four or five brachials are oblong, about four times as broad as long, and those following gradu- ally becoming wedge-shaped, almost triangular, about three times as broad as long, and less oblique and somewhat longer in the outer portion of the arms. The distal portion of the arms is perfectly smooth dorsally. Syzygies occur between brachials 3+4, again between brachials 13+14 to 20+-21 (usually in the vicinity of the fifteenth) and distally at intervals of 7-11 (most commonly 8 or 9) muscular articulations. Py is 7.5 mm. long, moderately stout basally but tapering rather rapidly in the proximal half and becoming slender distally, with 25 segments, which at first are twice as broad as long, becoming as long as broad after the tenth; the first four seg- ments are strongly carinate, this carination decreasing distally and disappearing after the middle of the pinnule. P, is 13 mm. long, slightly stouter than Py basally, taper- A MONOGRAPH OF THE EXISTING CRINOIDS 295 ing gradually and becoming slender in its distal third, with 26 segments, which at first are twice as broad as long, becoming as long as broad after the ninth and somewhat longer than broad in the terminal portion; the first seven or eight segments are rather strongly carinate, and in addition have a low sharp ridge running along the exterior surface at the base of the carinate processes. P; is similar to P, and of the same length, but the low ridge just described may be traced to about the twelfth segment. P; is 9 mm. long, with 19 segments, similar to the two preceding pinnules but slightly less stout: P, is small, 6 mm. long, tapering rapidly in the proximal half and becoming very slender distally, with 16 segments, which at first are twice as broad as long, becom- ing as long as broad on about the ninth, and longer than broad distally; the first six segments are carinate like those of the preceding pinnules. P; is similar, 5.5 or 5.0mm. long. P, and the following pinnules are 6 mm. long with 17 segments, which at first are twice as broad as long, becoming about as long as broad on about the eighth, and twice as long as broad terminally; the pinnules are about as stout basally as the two preceding, tapering rapidly in the proximal half and becoming very slender distally; the carination of the proximal segments is slightly marked on the first four, later becoming restricted to the second and third segments only, and disappearing entirely from the pinnules in the outer half of the arm. Near their tips the ambulacral lappets contain very minute calcareous reticula- tions of one, two, or three meshes; the perisome between the sacculi and the pinnulars is unplated. The tentacles contain two narrow lines of spicules, one distal and one proximal, which extend to about the end of their proximal half. Locality.— Investigator; Pedro shoal, off the western coast of India, west of Man- galore [A. H. Clark, 1909, 1912] (9, U.S.N.M., 35088 [original No. 3F], 35178 [origi- nal No. 6F], 35184 [original No. 11F], 35192 [original No. 7F], 35209 [original No. SF], 35210 [original No. 2F]; I. M.). History.—This species was first described in 1909 and was redescribed and figured in 1912. It is as yet known only from the nine specimens collected by the Investigator on Pedro Shoal. HETEROMETRA MADAGASCARENSIS (A. H. Clark) PuateE 29, Fieures 120-122 Craspedometra madagascarensis A. H. Cuark, Proc. U.S. Nat. Mus., vol. 40, 1911, p. 8 (southeastern Africa), p. 23 (description; Madagascar); vol. 43, 1912, p. 381 (cotype in the U.S.N.M.), p. 384 (original reference), p. 393 (Madagascar; description of specimens other than the type); Crinoids of the Indian Ocean, 1912, p. 119 (synonymy; Madagascar). Heterometra madagascarensis A. H. Cuarx, Unstalked crinoids of the Siboga-Exped., 1918, p. 78 (in key; range); John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1936, p. 99 (range), p. 104. Heterometra madagascariensis GisLiN, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 22, 48. Diagnostic features —The brachials are distinctly wedge-shaped, with the ends never quite parallel, and are not exceedingly short. The enlarged lower pinnules are smooth, with the earlier segments keeled. The cirri are 25-30 mm. long with 34-36 segments of which the longest are somewhat longer than broad and the outer are broader than long and bear small but prominent dorsal spines. The 16 arms are 130 mm.long. P,is 12mm. long, with 27 segments. P; is similar to P, but is very slightly larger and longer. 296 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Description.—The centrodorsal is thick discoidal with the bare polar area slightly convex, 4 mm. in diameter. The cirrus sockets are arranged in two closely crowded irregular marginal rows. The cirri are XVI, 34-36, from 25 to 30 mm. in length, stout basally, and tapering slightly in the proximal half. The first segment is short, and those following slowly increase in length to the eighth or tenth, which varies from only slightly broader than long to one-third again as broad as long, the distal segments being slightly shorter again. From the eleventh to the fifteenth (usually from about the fourteenth) on- ward small but prominent dorsal spines are developed. The opposing spine is larger than the spine on the preceding segment, triangular in lateral view with the apex subterminal, arising from the whole dorsal surface of the penultimate segment, and equal in height to about one-half of its width. The terminal claw is somewhat longer than the penultimate segment, moderately slender, especially in the distal two-thirds, and rather strongly curved proximally but becoming straighter distally. The radials are concealed by the centrodorsal. The [Br, are very short, band- like, in apposition laterally. The IBr. (axillaries) are very broadly pentagonal, twice as broad as long, with the lateral borders only half as long as those of the IBr,. The IIBr series are 4(3-++4). The division series and the first brachials are in close lateral apposition and laterally flattened, their lateral borders being moderately produced. The synarthrial tubercles are obsolete. The 16 arms of the type specimen are 130 mm. long. The first brachials are slightly wedge-shaped, twice as broad as long exteriorly, entirely united interiorly. The second brachials are of about the same size, but are more obliquely wedge-shaped. The first syzygial pair (composed of brachials 3+-4) is slightly longer interiorly than exteriorly, twice as broad as the exterior length. The next three or four brachials are oblong, three times as broad as long, those following becoming very obliquely wedge-shaped, twice as broad as long, after the proximal fourth of the arm gradually becoming less obliquely wedge-shaped, but never oblong. The eighth and ninth and following brachials have slightly produced distal edges, this feature gradually dying away after the middle of the arm. Py is 9.5 mm. long, more slender than P,, tapering more rapidly from the base and therefore more slender in its outer portion, and is composed of 32 segments of which the earlier are short, becoming about as long as broad on the twelfth and following. The second-fourth segments are strongly carinate, and those following are narrowly carinate to about the middle of the pinnule. P, is slender, becoming very delicate in the distal half, 10 mm. long with 26-28 segments of which the first is short and those following gradually increase in length so that the eighth and follow- ing are about as long as broad. The second-fourth segments are rather strongly carinate. P, is 12 mm. long, very slightly stouter basally than P, but tapering much more gradually and so appearing considerably stouter, with 27 segments, which become about as long as broad on the sixth and from one-third to one-half again as long as broad distally. P, is similar to P, but very slightly larger and longer. P, is 8 mm. long, resembling P, and P,; but tapering more rapidly and therefore more slender in the distal half. P,; is 6 mm. long, small and weak, with 16 segments all but the outermost of which are broader than long. The following pinnules are similar, gradually increasing in length and in the length of the component segments. A MONOGRAPH OF THE EXISTING CRINOIDS 297 The distal pinnules are exceedingly slender, 8 mm. long. On the large lower pinnules the second-fifth segments are rather strongly carinate, this feature decreasing rather rapidly distally, being soon confined to the second-fourth and then to the second- third, disappearing at the end of the proximal fourth of the arm. The color in alcohol is brownish white, the cirri being lighter and tinged with purple. Notes.—A second specimen has 18 arms about 90 mm. long. Six I[Br series are present, of which five are 4(3+4) and one is 2. One of the IIBr 4(3+-4) series bears internally a IIIBr 2 series. The IIBr 2 series bears a IIIBr 4(3++4) series externally. The color is purple. The third example has 13 arms 65 mm. long, the three IIBr series being 4(3-+4). The cirri are XVI, 32-33, 20 mm. long. The color is brownish white, the cirri light purple. Remarks.—This species is easily distinguished from Heterometra africana by the tapering cirri which bear shorter dorsal spines, by the much more slender lower pinnules, and by the longer wedge-shaped brachials. Locality Madagascar; Doctor Voeltzkow [A. H. Clark, 1911, 1912, 1918] (3, U.S.N.M., 35216; Berl. M., 5348). History.—This species was described as Craspedometra madagascarensis in 1911, and in 1912 notes were given on two additional specimens that had been collected with the type. In the key to the species of the genus Heterometra given in the report on the unstalked crinoids collected by the Siboga published in 1918 this form was included as a member of that genus. HETEROMETRA AMBOINAE (A. H. Clark) Puate 29, Figure 123; Puate 31, Figures 131-136 Antedon ludovict (part) Hartuaus, Nova Acta Acad. German., vol 58, No. 1, 1891, p. 6 (with A. [Stephanometra] oxyacantha the most usual Antedon species of Amboina), p. 11 (collected by Brock at Amboina), p. 15 (very common at Amboina), p. 19 (in key), p. 29 (detailed description of specimens from Amboina; discussion and comparisons), p. 113 (in Géttingen Mus.), pl. 1, figs. 7, 8, 11— Korutrr, Rev. Suisse Zool., vol. 3, 1895, p. 286 (Bay of Amboina).—?D6pERLEIN, Denkschr. med.-nat. Ges. Jena, vol. 8, pt. 5, 1898, p. 477 (Amboina)—Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, pl. 12, fig. 5 (from Hartlaub). Craspedometra amboinae A. H. Cuark, Crinoids of the Indian Ocean, 1912, p. 35 (= Antedon ludovici P. H. Carpenter, 1888, in part), p. 119 (=A. ludovict Hartlaub, 1891, in part; description [from Hartlaub]; Amboina).—RetcuensrercerR, Abh. Senck. naturf. Ges., vol. 35, Heft 1, p. 83 (Amboina; Ceylon), p. 99 (description of a specimen from Ceylon collected by Sarasin), p. 100 (comparisons). —A. H. Cuarx, Unstalked ecrinoids of the Siboga-Exped., 1918, p. 78, footnote 1. ?Heterometra reynaudii (part) H. L. Cuarx, Spolia Zeylanica, vol. 10, pt. 37, 1915, p. 85 (3 larger specimens). Heterometra amboinae A. H. Cuark, Unstalked crinoids of the Siboga-Exped., 1918, p. 78 (in key; range).—GisLéin, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 22, 23. Diagnostic features—The brachials are distinctly wedge-shaped with the ends never quite parallel, and are not exceedingly short. The enlarged lower pinnules are smooth, with the earlier segments keeled. The cirri have 28-40 (usually 30-35) subequal segments all of which are about twice as broad as long. The 13-23 (usually 208244—40 20 298 BULLETIN 82, UNITED STATES NATIONAL MUSEUM 17-19) arms are 105-180 mm. long. P, and P; have 26-31 segments; P3 is usually, though not always, longer than P,. Description.—The centrodorsal is large, slightly convex, with the dorsal pole often finely pitted. The cirrus sockets are arranged in two irregular rows. The cirri are about XXV, 35-40, about 35 mm. long; none of the cirrus segments are longer than broad; the outer bear small dorsal tubercles, and the terminal some- times prominent spines. The radials are plainly visible in the angles of the calyx, or entirely concealed. The IBr, are entirely united laterally. The IBr, (axillaries) are short and pentagonal. The I1Br series are usually 4(3-+-4), more rarely 2. IIIBr series, usually developed only interiorly, are 2, but 4(3+4) when following a IIBr 2 series. Synarthrial tubercles are moderately developed. The division series are laterally free, though sometimes very close together. The 16-22 elongated arms are 180 mm. long, some of them arising direct from the IBr axillaries. The brachials are short; the first is shorter than the second; the first syzygial pair (composed of brachials 3+-4) is very short; following this there is a series of discoidal brachials, and then a series of wedge-shaped brachials, the latter becoming progressively shorter, and short-oblong after the middle of the arm. The brachials are only slightly overlapping. The bases of the arms are rugose. Syzygies occur between brachials 3-+4, again somewhere between the eleventh and nineteenth brachials, and distally at intervals of from 7 to 11 muscular articu- lations. In the arms arising directly from a [Br axillary the second syzygy is between brachials 9+10, more rarely between brachials 16+-17, and the distal intersyzygial interval is usually 7 or 8 muscular articulations. P, or Pp is 8 or 9 mm. long. P, is usually much longer, 18 mm. long, almost as long as P;. P, is variable, but usually smaller. The following pinnules decrease in length to P;, which is 7 or 8 mm. long. The distal pinnules are 14 mm. long. The proximal segments of the pinnules in the basal third of the arm are strongly carinate. The enlarged lower pinnules are moderately stiffened. PP, is composed of about 20 sharply carinate segments, none of which are longer than broad. Pp and P, are very slender and flagellate after the first few segments. Notes.—The specimen from the Danish Expedition to the Kei Islands station 90 has 19 arms 165 mm. in length. The cirri have 34-36 segments. P, has 29 segments. P, is smaller with 26 segments. On one arm the thirty-second brachial bears an irregular series of small ossicles which collectively serve as an axillary; beyond this the main arm continues at a slight angle and another arm, of almost the same size, extends outward dorsally at right angles to the main arm stem. In one of the specimens from the Danish Expedition to the Kei Islands station 31 there are 21 arms; the 10 IIBr series are all 4(3+4) and there is a single IJIBr 2 series developed internally. Another specimen has 19 arms; of the nine IIBr series present eight are 4(3+-4) and one is 2. The cirri are XVIII, 31-33. A third example has 18 arms; of the eight ITBr series present one is 4(3+-4), six are 2, and one consists of a single axillary only. The cirri are XVIII, 28-31. The fourth specimen has 13 arms, there being 3 IIBr 4(3+4) series. The cirri are XVII, 28-30. One of the specimens from the Danish Expedition to the Kei Islands station 38 has 22 arms 110 mm. long. There are 10 I]Br 4(3+4) series and two IIIBr 2 A MONOGRAPH OF THE EXISTING CRINOIDS 299 series, both internally developed. The centrodorsal has the dorsal pole slightly convex, 4.5 mm. in diameter. The cirri are XVI, 28-31. In another specimen from station 38 the 21 slender arms are 135 mm. long. Of the 10 IJBr series eight are 4(3-++4) and two are 2. The two IIBr 2 series are on the same postradial series; one of them bears internally a IIIBr 2 series. The centro- dorsal is discoidal with the flat dorsal pole 5 mm. in diameter. The cirri are XX, 29-34; the segments are subequal, about twice as broad as long; there are small but prominent dorsal tubercles on the last 12 to 15. P, is composed of 44 segments, tapering rapidly on the first seven or eight segments and then becoming flagellate; the basal segments are short, and those in the distal third are about as long as broad; the second-seventh segments are narrowly but sharply carinate. P, resembles P, but is slightly longer and tapers more regularly so is less slender distally; the segments become as long as broad on the eighth; the second-seventh segments bear a high carination with a straight crest, and the second segment also has a lateral carinate process. P; is about 18 mm. long; it is stouter than the preceding pinnules and tapers more slowly; it is composed of 31 segments which become as long as broad on the eighth and twice as long as broad distally; the second-seventh segments are carinate, the crest of the keel, which disappears in the middle of the seventh segment, being straight and parallel to the axis of the segments. The third specimen from station 38 has 13 arms 105 mm. long. There are three IIBr 4(3+4) series. The cirri are XV, 29-30, from 20 to 24 mm. long. One of the specimens collected by the Danish Expedition to the Kei Islands at Amboina Bay has 18 arms 165 mm. long. Of the six IIBr series present four are 4(3+4) and two are 2. The two II1Br 2 series are both internally developed. The centrodorsal is discoidal with the dorsal pole flat, 4 mm. in diameter; the cirrus sockets are arranged in two rows. The cirri are XXIII, 33,30 mm. long. The cirrus segments are subequal, all being about twice as broad as long. The 12 to 16 segments preceding the penultimate have a low blunt median carination the crest of which rises distally so that in lateral view the outer portion of the cirri appears roundedly serrate. The opposing spine is prominent, much larger than the process on the preceding segment, conical, arising from the entire dorsal surface of the penultimate segment, and directed slightly forward. Occasionally it is much shorter with a chisel-shaped edge, or forked. Another specimen from Amboina Bay has 23 arms. There are six IIBr 4(3+4) series and seven IIIBr 2 series; of the latter six are internal and one is external, developed by the side of an internal one. The centrodorsal is flat, discoidal, with the dorsal pole 3 mm. in diameter; the cirri are in two irregularly alternating rows. The cirri are XIX, 33-38, from 30 to 35 mm. long, slightly slenderer than in the specimen described immediately above. A third specimen from Amboina Bay has 19 arms 120 mm. long. There are six IIBr 4(3+4) series and three IIIBr 2 series, all the latter internally developed. The cirri are XVII, 33, 30 mm. long. The color is dark purple, the outer two-thirds of the arms banded purple and flesh color in bands about 3 mm. broad. The arm bases are more or less mottled and banded with flesh color. The cirri are flesh color in the proximal half or two-thirds, purple distally, sometimes also purple just at the base. 300 BULLETIN 82, UNITED STATES NATIONAL MUSEUM A fourth specimen from Amboina Bay has 21 arms 105 mm. long. The cirri are XIX, 28-30, from 20 to 25 mm. long. A fifth example from Amboina Bay has 17 arms 120 mm. long. The six I/Br series are 4(3-+-4), and the single internally developed IIIBr series is 2. A sixth specimen from Amboina Bay has 18 arms about 100 mm. long. Of the seven IIBr series present four are 4(3-+-4) and three are 2. The single IIBr 2 series is internally developed. The seventh specimen from Amboina Bay has 14 arms. Of the four IIBr series two are 4(3+-4) and two are 2. The cirri are XXI, 34-39, from 28 to 32 mm. long. The five specimens from the breakwater at Amboina have 21, 18, 18, 18, and 12 arms, which in the largest are 120 mm. long. The description of the species included herein is that given by Hartlaub for the 12 specimens collected by Brock at Amboina. Professor Strubell’s specimen from Amboina, recorded by Reichensperger, is much broken. It has about 15 armsand XXII cirri. The color in alcohol is chocolate brown. The specimen from Singapore has 13 arms, which were probably between 70 and 80 mm. long; there are three I1Br 4(3-++-4) series just appearing. The cirri are 20-25 mm. long and consist of 31-35 segments. This is apparently a young example of this species. Reichensperger said that in the specimen from Ceylon the centrodorsal is large with a slightly concave dorsal pole and bears cirri in two irregular rows. The dorsal pole bears isolated pits about its border. The cirri are XXIV, about 36, and are 30 mm. long. The segments are broader than long and from the thirteenth onward bear a blunt dorsal tubercle, which on the last 8 to 10 becomes a prominent spine. When viewed from above the cirri are seen to taper far less than those of Craspedometra acuticirra. The 16 arms are about 140 mm. long. The structure of the distal brachials and pinnules recalls that shown in Hartlaub’s figure. Pp reaches about 7 mm. in length and consists of about 20 segments of which the four lowest are broad and sharply keeled, the pinnule beyond that point becoming filiform. P, is longer, almost 12 mm. long, with 25 segments, which in the basal third of the pinnule are laterally and distally keeled. P, is the longest pinnule, about 17 mm. long with 26 segments, and is like- wise keeled. P, is smaller than P, but somewhat longer than P,, with about 23 segments. P, is 7 mm. long, with 17 segments; like the next following pinnules, it is keeled at the base. Reichensperger remarked that this species is strongly differen- tiated from Craspedometra acuticirra. . Professor Koehler said that the numerous specimens of this species sent to him by M. Bedot are not absolutely identical with those studied by Hartlaub. The cirri are XXV-XXVIII, with often 33-34, but rarely more than 35, segments. The IIBr series are sometimes 2, sometimes 4(3-+-4). The IIIBr series, which are rarely pres- ent, are 2. In contrast to what had been written by Hartlaub, Koehler found, though rarely, ITBr 2 series following I1Br 2 series. Very marked synarthrial tuber- cles are present. The arms are 14-22 in number and reach 150 mm. in length. The first syzygy is between brachials 3+-4, and the second is at about brachials 9+10 in arms arising directly from a IBr axillary, but at about brachials 17-++18 or 18+-19 in arms arising from a I[Br or IIJBr axillary. P, is short. P, is much A MONOGRAPH OF THE EXISTING CRINOIDS 301 longer, and a little smaller than P;, which is the longest pinnule. P, is nearly the size of P;, but the length of the pinnules decreases abruptly beyond P;. The color is dark purple-violet. The two specimens of Antedon ludovici recorded by Déderlein from Amboina probably represented this species and not Craspedometra acuticirra. Localities—Danish Expedition to the Kei Islands; Dr. Th. Mortensen; station 90; Sebesi Strait, between Java and Sumatra; 36 meters; hard bottom; August 1, 1922 (1). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; Station 31; about 50 meters; sand; April 19, 1922 (4). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; station 38; about 35 meters; sand; April 24, 1922 (3). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; Amboina Bay; about 50 meters; stones and sand; March 2, 1922 (7). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; Amboina; break- water; about 1 meter; February 28, 1922 (5). Amboina; Dr. J. Brock, 1884—’85 [Hartlaub, 1891]. Amboina; Professor Strubell, 1890 [Reichensperger, 1913]. Bay of Amboina; Maurice Bedot and C. Pictet [Koehler, 1895; A. H. Clark, 1912]. ?Amboina; Prof. Richard Semon [Déderlein, 1898]. Singapore; Svend Gad, December 12, 1906 (1, C. M.). Ceylon; Doctor Sarasin, 1886 [Reichensperger, 1913]. Geographical range-—From the Kei Islands, Amboina, and Singapore westward to Ceylon. Bathymetrical range.—From the shoreline down to about 50 meters. History.—In 1891 Dr. Clemens Hartlaub described in detail and figured a species represented by 12 specimens, which had been collected by Dr. J. Brock at Amboina in 1884-85. He had at first considered this form as new, but later he decided that it represented Carpenter’s Antedon ludovici, so he published the long description which he had prepared under that name. In 1895 Prof. René Koehler recorded, under the name of Antedon ludovici, and gave notes upon a number of specimens collected at the Bay of Amboina by Maurice Bedot and Camille Pictet. In 1898 Prof. Ludwig Déderlein recorded two specimens of Antedon ludovici that had been collected by Prof. Richard Semon at Amboina. These probably repre- sented the present species. A study of the type specimens of Carpenter’s Antedon ludovici, A. acuticirra and A. bipartipinna at the Hamburg Museum in 1910 had convinced me that all three represent the same species, and furthermore that this species is quite different from that represented by Hartlaub’s specimens from Amboina. Accordingly in my memoir on the crinoids of the Indian Ocean published in 1912 I named the species from Amboina Craspedometra amboinae on the basis of Hartlaub’s description, which I quoted. In 1913 Dr. August Reichensperger remarked that amboinae is clearly differen- tiated from Carpenter’s acuticirra, and under the heading Craspedometra amboinae he 302 BULLETIN 82, UNITED STATES NATIONAL MUSEUM recorded and gave notes on a specimen collected by Dr. Sarasin in Ceylon in 1886, and mentioned another collected by Professor Strubell at Amboina in 1890. In his account of the crinoids from Ceylon in the collection of the Colombo Museum (1915) Dr. Hubert Lyman Clark mentioned three specimens of Heterometra reynaudi that were noticeably larger with longer arms than the others. These may possibly have been specimens of H. amboinae. In 1918 amboinae was inserted in the key to the species of the genus Heterometra in the Siboga report but was not mentioned further as it was not secured by the Siboga. In a footnote under Heterometra ater as it appears in this key, I said that the specimen from Ceylon described by Reichensperger under the name of Craspedometra amboinae appears to belong to this species; at any rate it is closely related to it. I am now convinced that this is an error and that Reichensperger’s specimen was correctly identified. HETEROMETRA REYNAUDI (J. Miller) PLATE 32, Figures 137-143 {See also vol. 1, pt. 1, fig. 75 (lateral view of abnormal specimen), p. 128; fig. 257 (centrodorsal), p. 255; fig. 472 (radial pentagon), p. 361; pt. 2, figs. 45, 46 (radial pentagon), p. 26; figs. 451, 452 (pinnule tip), p. 261). ?Alecto horrida Leacu, Zool. Misce., vol. 2, 1815, p. 61, pl. 80.—ScHwerccER, Beobachtungen au naturhistorischen Reisen, 1819, p. 66, pl. 4, figs. 39-42.—Guitp1Ne, Zool. Journ., vol. 4, 1829, p. 173.—Levckart, Zeitschr. organ. Physik, vol. 3, Heft 4, 1833, p. 385, footnote —[ANonymovs], Penny Encylopedia, vol. 7, 1837, p. 391.—P. H. Carpenter, Trans. Linn. Soc. (Zool.), ser. 2, vol. 2, part 1, Dec. 1879, p. 4— A. H. Cuarx, Proc. U. S. Nat. Mus., vol. 34, 1908, p. 450; Crinoids of the Indian Ocean, 1912, pp. 2, 280. ?Comatula horrida J. Miuupr, Monatsb. preuss. Akad. Wiss., April 1840, p. 6; Archiv. Naturg., 1840, vol. 1, p. 311. 2G. horrida J. Miiuuer, L’Institut, 17 Sept. 1840, p. 394. Comatula (Alecto) reynaudi J. Mistter, Monatsb. preuss. Akad. Wiss., 1846, p. 178 (description; Ceylon); Abh. preuss. Akad. Wiss., 1847 (1849), p. 259 (redescribed). Comatula reynaudi Dusarvin and Hupf, Histoire naturelle des zoophytes, Hchinodérmes, 1862, p. 205 (synonymy; description; Ceylon). Antedon reynaudii P. H. Carnpentmr, Trans. Linn. Soc. (Zool.), ser. 2, vol. 2, 1879, p. 29 (listed as an Antedon). Antedon reynaudi Be.u, Proc. Zool. Soe. London, 1882, p. 533 (listed), p. 534 (specific formula).— P. H. Carpenter, Proc. Zool. Soc. London, 1882 (1888), pp. 743 ff (discussion of Bell’s method of formulation and corrected formula).—Bxruu, Sci. Trans. Roy. Dublin Soc., ser. 2, vol. 3, 1887, p. 645 (Ceylon).—P. H. Carpenter, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, pp. 43, 55, 252, 255, 366, 380 (discussion).—BrLt, Proce. Zool. Soc. London, 1888, p. 384 (Tuticorin, Madras), p. 387 (occurs in the Bay of Bengal)—Harriavus, Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 17 (in key)—Tuurstron, Madras Gov. Mus. Bull., No. 2, 1894, p. 114 (Pambdn; on stems of gorgoniae)—Cuapwickx, Rep. Ceylon Pearl Oyster Fisheries, pt. 2, Suppl. Rep. 11, 1904, p. 153 (occurs at Ceylon), p. 156 (station xxxrv; de- tailed description), plate, figs. 6-12—Hamann, Bronns Klassen und Ordnungen des Tier- Reichs, vol. 2, Abt. 3, 1907, p. 1582 (listed)—A. H. Ciarx, Crinoids of the Indian Ocean, 1912, pp. 35, 40 (identity of previous records). Antedon renaudi Tuurston, Madras Goy. Mus. Bull., No. 2, 1894, p. 106 (Gulf of Manaar; littoral). Antedon variipinna Cuapwick, Rep. Ceylon Pearl Oyster Fisheries, pt. 2, Suppl. Rep. 11, 1904, p. 153 (occurs at Ceylon), p. 157 (station II; characters). Himerometra reynaudi A. H. Cuarx, Smithsonian Misc. Coll., vol. 50, pt. 3, 1907, p. 356 (listed); Proc. Biol. Soc. Washington, vol. 22, 1909, p. 11 (listed); Proc. U. S. Nat. Mus., vol. 40, 1911, p. 25 (compared with H. joubint). A MONOGRAPH OF THE EXISTING CRINOIDS 303 Heterometra reynaudii A. H. CuarK, Bull. Mus. Hist. Nat., Paris, 1911, No. 4, p. 245 (identity), p. 251 (Ceylon; description of type; compared with H. joubini); Mem. Australian Mus., vol. 4, 1911, p. 765, footnote (recorded by Chadwick from Ceylon as Antedon variipinna); Rec. Indian Mus., vol. 7, pt. 3, 1912, No. 26, p. 268 (India; ?India; descriptions of specimens) ; Crinoids of the Indian Ocean, 1912, pp. 30, 35, 40 (identity of previous records), p. 121 (synonymy; new localities; descriptions; specimens deficient in lime; parasitic Eulimas), fig. 9 (p. 122), p. 317 (additional locality; descriptions); Smithsonian Mise. Coll., vol. 61, No. 15, 1913, p. 27 (pub- lished references to specimens in the British Museum; Tuticorin; notes).—H. L. Cuark, Spolia Zeylanica, vol. 10, pt. 37, 1915, p. 85 (Ceylon; 55 specimens, 3 noticeably larger with longer arms than the rest), p. 93 (occurs at Ceylon).—A. H. Cuarx, Unstalked crinoids of the Siboga- Exped., 1918, p. 79 (in key; range); Journ. Linn. Soe. (Zool.), vol. 36, No. 249, May 1929, p. 638 (South Male).—Gis.in, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 22, 27. Comatula (Alecto) reynaudii A. H. Cuarx, Bull. Mus. Hist. Nat., Paris, 1911, No. 4, p. 245 (identity) ; Crinoids of the Indian Ocean, 1912, p. 30 (same). Heterometra bengalensis A. H. Cuarx, Crinoids of the Indian Ocean, 1912, p. 130 (south of Ceylon; Andaman Is.; notes), p. 318 (off Gopalpore, 25-28 and 30-38 fathoms; ?India; notes), fig. 12, p. 131.—H. L. Crarx, Spolia Zeylanica, vol. 10, pt. 37, 1915, p. 93 (part; occurs at Ceylon).— A. H. Crarx, Unstalked crinoids of the Siboga-Exped., 1918, p. 79 (part; in key; range). Heterometra reynaudi PELSENEER, Bull. Soc. Zool. France, vol. 53, 1928, p. 173 (parasitized by Eulima [Sabinella], sp.).—A. H. Cuarx, Rec. Indian Mus., vol. 34, pt. 4, 1932, p. 551 (listed), p. 556 localities; notes)—A. H. Cuarx, John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1936, p. 100 (range), p. 103. Diagnostic features.—The brachials are distinctly wedge-shaped with the ends never quite parallel and are not exceedingly short. The enlarged lower pinnules are smooth, with the earlier segments keeled. The cirri are 25-27 mm. long and are composed of 30-47 (usually in fully grown individuals more than 40) subequal seg- ments, all of which are somewhat broader than long, commonly twice as broad as long, and of which the outer bear prominent sharp dorsal spines. The 14-25 (usually 17-20) arms are up to 110 mm. long. P, is the largest pinnule with 27 segments. P; is about the size of P,, both having 29 segments. Description.—The centrodorsal is a thick disk with a convex dorsal surface, bear- ing the cirri in two alternating rows on its sloping sides. The cirri are XXV, 35-36, all but a few immature ones being rather more than 20 mm. long. All the segments are slightly broader than long, the later ones being laterally compressed and bearing, from the fourteenth onward, a strong and distally directed dorsal spine. The relative length of the cirrus segments varies considerably. Usually all the cirrus segments are subequal, all being about twice as broad as long, but sometimes they are longer, the segments in the proximal half of the cirri being in extreme cases nearly as long as broad. The radials are entirely concealed. The IBr, are just visible above the bases of the cirri and are in almost complete lateral contact. The IBr, (axillaries) are broadly pentagonal, and are about half again as long as the IBr;. In the specimen described two of the postradial series bear two I[Br 4(3+4) series, two bear each a single IIBr 4(3+4) series, and the fifth bears two IIBr 2 series. Of the eight IIBr series present, therefore, six are 4(3-++4) and two are 2. The first elements of the IIBr series are in almost complete lateral contact, and there is a slight synarthrial tubercle on the articulation between the first and second elements. 304 BULLETIN 82, UNITED STATES NATIONAL MUSEUM The 18 arms are probably 110 mm. long and are composed of about 200 short and slightly overlapping brachials. The first brachials are in almost complete con- tact interiorly. The first eight brachials are discoidal and are succeeded by about 30 shortly triangular brachials, which are followed by discoidal ones, these continuing to the tip of the arm. The distal ends of the brachials are usually very finely spinous, though not produced. In specimens from the east coast of India they are occasionally slightly produced and rarely considerably produced and everted. Syzygies on six arms occur as follows: 1. Between brachials 3+4, 18+14, 21+22, 28+29, 37+38, 47+48, 56+57, 65-+66, and 74-++75. 2. Between brachials 3+4, 9+10, 14+15, 21422, 29+30, 33+434, 414-42, 49-+50, 63-+64, 72+73, 77+78, 84485, and 92+93. 3. Between brachials 3+-4, 15+16, 25+26, 34+35, and 42+43. 4, Between brachials 3+-4, 15+16, 25+26, 35+36, 50+51, and 60+61. 5. Between brachials 3+-4, 16+17, 23-+24, 31+32, and 38+39. 6. Between brachials 3-++-4, 7+8, and 31+32. Pp is about 10 mm. long and is composed of 31 segments of which the first seven are broad, flattened, and carinate, the later ones gradually becoming cylindrical, longer than broad, and with rounded ends. P, and P; are about equal in size, and P, and P, resemble them. P, and P; have about 29 segments. P, and P, are the largest pinnules on the arm, P, being about 15 mm. long, but having 27 segments only. The basal segments of the first three pairs of brachial pinnules are flattened and more or less carinate. P,; and P, are the shortest pinnules, the pinnules succeeding gradually increasing in length. The carination on the basal segments of the earlier pinnules, usually moderate or rather slight, is occasionally much exaggerated. The disk is 10 mm. in diameter and is considerably incised. The ambulacra are naked. Sacculi are abundant and closely set on the arms and pinnules, but are absent from the disk. In alcohol the disk is chocolate-brown, fading to white in the interambulacral angles. The brachials are pinkish, with the articulations in the proximal fourth of the arms marked with bands of deep reddish brown, which gradually fade into dots on the sides of the arms. The pinnules are irregularly banded with brown and white. The preceding description is adapted from Chadwick and is based upon the single specimen secured at the Ceylon Pearl Oyster Fisheries station XXXIV. Notes.—The two specimens from South Male are both small 10-armed individuals the identity of which must be considered as somewhat doubtful. Miiller’s description of the type specimen is as follows: The centrodorsal is flat, bearing marginal cirri. The cirri are about XX, 40; the segments gradually develop a dorsal tubercle. The radials are visible. The IIBr series are 4(3-+4). The 20 arms are 4 inches long. The distal intersyzygial interval is usually eight muscular articulations. P, and P; are longer than the other pinnules. The ventral side of the disk is soft. Miiller wrote in his description that there are three radials of which the third is axillary with a syzygy, and the same wording was given by Dujardin and Hupé in 1862. Obviously a part of what he wrote in his manuscript was omitted in the print- ing. What he intended to say was that there are three radials (that is, the radials A MONOGRAPH OF THE EXISTING CRINOIDS 305 are visible and are followed by the two ossicles of the [Br series) and the next division consists of three ossicles of which the third is axillary with a syzygy (that is, the I[Br series are 4[3-++4]). Carpenter in the Challenger report mentioned reynaudi as having all the outer arm divisions of two segments only, the axillaries being syzygial segments (that is, the outer arm divisions are 3[2+3)). At the time this was written he had evidently confused the type specimen of reynaudi with a specimen of Capillaster multiradiata from Trinquemale, Ceylon, which was also collected by M. Reynaud in 1829 (see Part 3, pp. 197, 204, 205). Farther on in the Challenger report, however, he cor- rectly placed reynaudi in the Savignyi group and assigned to it the following characters. The IIBr series are 4(3+-4). The cirri, which are without definite arrangement, have not more than 45 segments. The distal (that is, distichal, or I[Br) pinnule is generally smaller than its successors. The segments of the lower pinnules are without lateral processes. There are 40-45 cirrus segments which are mostly spiny. The distal intersyzygial interval is usually from 8 to 11 muscular articulations. In 1910 I examined Miiller’s type specimen at Paris. There are 41-44 cirrus segments, which are subequal, all being about twice as broad as long. There are 19 arms, nine II Br 4(3+-4) series being present. The brachials are slightly overlapping. Dr. Hubert Lyman Clark says that of the 55 specimens sent to him for study by the Colombo Museum three are noticeably larger than the rest with longer arms. Of the four specimens collected by the Investigator south of Ceylon in 62 meters two are of medium size and one is small. The fourth is curiously abnormal (see page 307). Of the two specimens collected by the Jnvestigator south of Ceylon in 58 meters one has the cirri XIII, 24-28, and the other, with 12 arms, has the cirri XII, 21-27. One of the 10 specimens from the entrance to Palk Strait has 14 arms, four I[Br 4(3+4) series being present. Another has 16 arms; of the six IJBr series present five are 4(3+-4) and one is 2. Two have 20 arms with eight of the 10 IIBr series 4(3+4) and two 2. One has 20 arms with seven of the ten IIBr series 4(3-+4) and three 2. One has 21 arms, with all the I[/Br series 4(3-++4) and all the IIIBr series, which are internally developed, 2._ The remaining four are badly broken. One of the four specimens from Palk Strait has 20 arms about 110 mm. long and the cirri XXV, 34-40, from 25 to 27 mm. long. Another has 21 arms (one IIIBr series being present) and the cirri XXI, 33-42. A third has 22 arms (with two III[Br series both internally developed). The fourth has 20 arms, one of the IIBr series being absent but its loss being compensated by the development of one IIIBr series which, however, is 4(3-++4) instead of the usual 2. The specimen from Investigator station 90 is medium sized. That from the Ganjam coast in 44-55 meters is small with 12 arms. Of the five specimens from off Gopalpore in 46-51 meters one has 14 arms 80 mm. long; of the four IIBr series three are 4(3-+4) and the fourth is 2. The cirri have 34-41 segments. The color is white, with the articulations banded with purple; the cirri are white with the distal portion purple dorsally. Three of the specimens have each 17 arms 80 mm. long; in all these three there are seven IIBr series. One has all seven of the IIBr series 4(3-+4). The cirri have 31 segments. The color is white with a row of small brown spots, in pairs, on the sides 306 BULLETIN 82, UNITED STATES NATIONAL MUSEUM of the proximal third of the arms. Another has six I1[Br 4(3-++4) and one IIBr 2 series. The cirri have 30-33 segments and are 20 to 25 mm. in length. The color is white with occasional blotches of brown on the arms. The third has five I[Br 4(3+-4) and two I1Br2 series. The color is white, the cirri, and well-separated narrow bands on the arms, being purple. The fifth specimen has 19 arms 80 mm. long. One of the IIBr series is absent. The dorsal pole of the centrodorsal is flat, 3 mm. in diameter. The cirri are XVIII, 28-32, 20 mm. long. The color is white, the cirri, narrow bands on the arms, and a series of small regular spots on the proximal third of the arms brown. Of the 10 specimens dredged by the Golden Crown off Gopalpore in 55-69 meters two have 17 arms, three have 18 arms, three have 20 arms, and two have 21 arms. Of the 17-armed specimens one has the arms 75 mm. long, and the seven II Br series 4(3-+-4); the cirri are 18 mm. long and are composed of 26-28 segments; the coloris white. The other has the arms 85 mm. long; the calyx and arm bases are white, the arms light yellow brown. Of the 18-armed specimens one has the arms about 75 mum. long; all the II Br series present are 4(3-+4); a I11Br 2 series, internally developed, is present on one of the postradial series; the cirri are 23 mm. long and are composed of 32-36 segments; the color is light yellow-brown. Another has the arms 90 mm. long; the cirri are 23 mm. long and are composed of 30 segments; the color is white. The third has the arms 110 mm. long; one postradial series is without I[Br series; of the eight I] Br series present seven are 4(3-+4) and one is 2; the cirri are 23 mm. long and composed of 30-33 segments. One of the 20-armed specimens has the arms 70 mm. long; the color is white, becoming yellow-brown on the arms, with the pinnules purple. Another has the arms 85 mm. long; the 10 I[Br series are 4(3 +4) ; the cirri are 20 mm. long and are composed of 25-27 segments; the color is white with the proximal portion of the arms spotted along the sides and the remainder of the arms narrowly banded with purple; the cirri are brown. ‘The third has the arms 115 mm. long; the 10 IIBr series are all 4(3+4); the cirri are XIV, 33-34, 25 mm. long; the color is white with faint purplish blotches on the arms, the cirri being purple. One of the 21-armed specimens has the arms 75 mm. long; the 10 IIBr series are 4(3+-4); the single internally developed IIIBr series is 2; the color is white, with the cirri deep purple. The other 21-armed specimen has the arms 80 mm. long; the 10 I1Br series are 4(3+-4); the single III Br series, which is developed internally, is 2, the calyx and arm bases are white, the arms are yellow-brown, and the cirri purple. The specimen from Puri Beach, Orissa, has 16 arms. The two specimens from Tuticorin in the British Museum are fine examples of the species. One has 18 arms 90 mm. long and the cirri 25 mm. long; one of the IBr series carries no II[Br series. The other has 17 arms. The specimen from Tuticorin Harbor collected by Dr. H. S. Rao has 17 arms. Of the three specimens from east of the Terribles two have each 25 arms, five IIIBr 2 series being present all developed internally. The third is smaller with only 13 arms. The two specimens from the Arrakan coast each have 22 arms, two IIIBr series being developed, both internally. One has 35-42 and the other 47 cirrus segments. The specimen presumably from the Arrakan coast is medium sized. A MONOGRAPH OF THE EXISTING CRINOIDS 307 The two specimens from the Andaman Islands are similar to those from south of Ceylon. One has 14 arms and 26 cirrus segments, and the other has 15 arms and 22 or 23 cirrus segments. Of the three specimens labeled India one has 20 arms about 65 mm. long; of the 10 IIBr series, nine are 4(3-++-4) and oneis2. Another has 12 arms 25 mm. long; there are 15-18 cirrus segments of which the fifth or sixth and following bear dorsal spines. The third is medium sized. Of the 10 specimens labeled ?India, which are probably from Ceylon, 6 are of medium size and 4 are very small. One of the latter has 10 arms 15 mm. long; the cirri are XIV, the longest with 17 segments of which the ninth and following bear dorsal spines, the shorter with 12 segments none of which bear dorsal spines. Another also has 10 arms; the longest cirri are 10 mm. long and are composed of 22 segments, dorsal spines being developed from the seventh segment onward; the smallest cirri are 2.5 mm. long with nine segments, quite without dorsal spines and exactly resembling the cirri of young examples of Antedon bifida. A third has 12 arms 25 mm. long. Abnormal specimen.—In a specimen from south of Ceylon the centrodorsal and division series are of normal size and shape. One cirrus remains which tapers to a point at the seventeenth segment; no dorsal spines are developed. The 19 arms, which are only 17 mm. long, are of normal size basally but rapidly taper to a point beyond which they are continued for a short distance in a slender and soft uncalcified process. The lower pinnules taper very rapidly for the first five or six segments, from that point onward being very slender and hairlike with little or no lime in their composition. Beyond the arm bases the pinnules are exceedingly slender with never more than the first or first two segments of normal size, usually with none, and commonly with traces of calcareous deposits showing segmentation, though often quite without any. With the reduction of the calcareous material in the pinnules comes a reduction in the pinnule sockets, the noncalcareous pinnules in the outer portion of the arm not being accompanied by any modification in the outer edge of the brachials whatever. Most of the arms of this specimen have been broken off and repaired at the syzygy between the third and fourth brachials. A marked shortening of the arms, which then taper abruptly, is occasionally noticed in multibrachiate comatulids, though it does not reach anything like the extent seen in this individual. Instances are seen in the specimen of Capillaster macrobrachius collected by the Siboga (see Part 3, p. 159, and compare figs. 26 and 25, pl. 8), and in the specimens of Himerometra robustipinna from Ceylon (recorded by Reichensperger as H. kraepelini; see page 199) and Akyab, Burma (the type of Hartlaub’s Antedon kraepelini; see page 198). Parasite—Two of the specimens from Palk Strait are parasitized by small mol- lusks of the genus Sabinella, which are attached to the cirri near the base (see Part 2, . 648). ; Localities —South Male, Maldive Islands; Prof. J. Stanley Gardiner [A. H. Clark, 1929] (2, B.M.) Ceylon; M. Reynaud, Chevretie expedition, 1829 [J. Miiller, 1846, 1849; Dujardin and Hupé, 1862; P. H. Carpenter, 1879, 1883, 1888; Bell, 1882; Hartlaub, 1891; A. H. Clark, 1907, 1909, 1911, 1912] (1, P.M.). Ceylon [Bell, 1887]. 308 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Ceylon [H. L. Clark, 1915] (55, M.C.Z., 603, 604, 610; Colombo Mus.). West coast of Ceylon; Ceylon Pearl Oyster Fisheries station II; from 7 to 14 miles north of Negombo, 5 miles off shore; 15-16 meters; coarse yellow sand, shells, stones, and small coral; temperature 25.28° C.; specific gravity 1.023 [Chadwick, 1904]. South coast of Ceylon; Ceylon Pearl Oyster Fisheries station XXXIV; Welligam Bay, various parts; 4-13 meters; shell sand and a little mud; temperature 25.44° C.; specific gravity 1.0225 (Chadwick, 1904] (1, U.S.N.M., 35211). Investigator; south of Ceylon (lat. 6° 01’ N., long. 81° 16’ E.); 62 meters [A. H. Clark, 1912] (4, U.S.N.M., 35091, 35106; I.M.). Investigator; south of Ceylon (lat. 6° 06’ 30’’ N., long. 81° 23’ E.); 58 meters [A. H. Clark, 1912, 1918; H. L. Clark, 1915] (2, U.S.N.M., 35187 [original No. 4D]; I.M.). Gulf of Manaar; littoral; Edgar Thurston [Thurston, 1894]. Investigator; entrance to Palk Strait, between northeastern Ceylon and India; Point Pedro bearing southsoutheast, distant about 3 miles; 11-15 meters; sandy bottom; 1893-1894 [A. H. Clark, 1932] (1, I.M.). Investigator; off northeastern Ceylon; entrance to Palk Strait; Point Pedro bearing southsoutheast, about 3 miles distant; 11-15 meters; sandy bottom [A. H. Clark, 1912] (10, U.S.N.M., 35762, 36256; I.M.). Investigator; Palk Strait, northeastern coast of Ceylon [A. H. Clark, 1912] (4, U.S.N.M., 35108 [original No. 4897-7], 35185 [original No. 5E], 35191, 35208 [original No. 4898-7]; I.M.). Investigator station 90; off the Ganjam coast, Madras presidency; 8 miles east- southeast of Kalingapatam lighthouse; 51-55 meters; sand and shells; February 17-21, 1890 [A. H. Clark, 1912] (1, U.S.N.M., 35078). Bengal Fisheries steamer Golden Crown; Ganjam coast, Madras Presidency; 44-55 meters; March 8-16, 1909 [A. H. Clark, 1912] (2, U.S.N.M., 35092; I. M.). Bengal Fisheries steamer Golden Crown; off Gopalpore; 46—51 meters [A. H. Clark, 1912] (5, U.S.N.M., 35103 [original No. ZEV 3624-7], 35113; I. M.). Bengal Fisheries steamer Golden Crown; off Gopalpore; 55-69 meters [A. H. Clark, 1912] (10, U.S.N.M., 35097, 35109, 35188; I. M.). Bengal Fisheries steamer Golden Crown; off Gopalpore; 55-69 meters; December 1909 [A. H. Clark, 1932] (1, I. M.). ; Bengal Fisheries steamer Golden Crown; Puri Beach, Orissa [A. H. Clark, 1932] (1, I. M.). Tuticorin, Madras; Edgar Thurston; littoral [Bell, 1888; Thurston, 1894; A. H. Clark, 1913] (2, B. M.). Tuticorin Harbor; shore collecting; Dr. H. S. Rao, February—March, 1926 [A. H. Clark, 1932] (1, I. M.). Pamban; littoral, on stems of gorgonians [Thurston, 1894]. Investigator; east of the Terribles (about 30 miles southeast % south from the south point of Western Borongo Island, Burma) (lat. 19°27’ N., long. 93°18’ E.); 24 meters [A. H. Clark, 1912] (3, U.S.N.M., 35076, 35204 [original No. 67B]; I. M.). Arrakan coast, Burma [A. H. Clark, 1912] (2, I. M.). ?Arrakan coast [A. H. Clark, 1912] (1, I. M.). Investigator; Andaman Islands [A. H. Clark, 1912, 1918] (2, I. M.). A MONOGRAPH OF THE EXISTING CRINOIDS 309 India [A. H. Clark, 1912] (3, I. M.). ?India [A. H. Clark, 1912] (8, U.S.N.M., 35090 [original No. 34H], 35093 [original No. 9H], 35105 [original No. 10H], 35214 [original No. 30H]; I. M.). Investigator; ?India (probably Ceylon) [A. H. Clark, 1912] (2, U.S.N.M., 35215 [original No. 55H], 35217; I. M.). Geographical range.—From the Maldive Islands and Ceylon northward along the western and northern shores of the Bay of Bengal and south along the eastern shore as far as Borongo Island, near Akyab, Burma. Bathymetrical range.—From the shoreline down to 62 (?69) meters. Most of the records are littoral or sublittoral. History.—lIt is probable that the Alecto horrida described by Dr. W. E. Leach in 1815 is in reality this species. Leach’s description is very short. He says “Rays simple; tentacules of the back smooth, with the joints moderately long and produced internally. Locality unknown. British Museum.” Carpenter noted that Schweigger’s figure of the disk of Alecto horrida shows clearly enough that the five trunks of the ambulacral grooves converge toward the center of the disk as in Antedon bifida and that Alecto horrida is therefore a true Antedon in the modern (1879) sense of the term, although belonging to that division of the genus in which the repetition of the bifurcation of the 10 primary arms is carried to a great extent. Leach’s figure shows a comatulid with 25 very flexible, slender, and slowly taper- ing arms about 90 mm. long. The drawing is sufficiently detailed to indicate that it represents, without any question, a species with the II Br series 4(3-+-4) and the III Br series 2. The cirri are numerous, slender, and rather long, all of them being shown incurved over the centrodorsal. They are the structures referred to as the “‘tenta- cules of the back,” and the expression ‘‘produced internally” indicates that the seg- ments carry dorsal processes. This specimen cannot belong to a species of Comasteridae, for no member of that family has such long and slender many jointed cirri and such evenly tapering and flex- ible arms. Nor do the four enlarged brachials and three enlarged pinnules shown suggest those of any species of Comasteridae. It is evidently an example of a species of Himerometridae belonging to the genus Heterometra. Among the species of Hetero- metra it agrees perfectly, so far as the details of the figure will permit us to form an opinion, with H. reynaudi of Ceylon, but not with any other species. There can be little doubt that Leach’s Alecto horrida, described in 1815, is the same species as Miiller’s Comatula (Alecto) reynaudi, although definite proof of the identity of the two is lacking. This species was first recognizably described in 1846 under the name of Comatula (Alecto) reynaudi by Prof. Johannes Miiller from a specimen that he had studied in the Paris Museum and that had been collected during the cruise of the Chevrette by M. Reynaud in Ceylon in 1829. It was redescribed by Professor Miller in 1849, and this redescription was translated and published in French by Dujardin and Hupé in 1862. In all three of these descriptions there is a curious omission of part of a sentence (see page 304), which renders them quite obscure. In the autumn of 1876 Dr. P. H. Carpenter examined Miiller’s type specimen in 310 BULLETIN 82, UNITED STATES NATIONAL MUSEUM the Paris Museum, and in 1879 he assigned this species to the genus Antedon as under- stood by him. In October 1882 Prof. F. Jeffrey Bell proposed a specific formula for this species, which was amended by Carpenter in April of the following year. In 1887 Professor Bell recorded additional specimens from Ceylon. In the Challenger report on the comatulids published in 1888 Carpenter included this species in his key to the Savignyi group and also said that its outer division series are of two joints with the axillary a syzygial joint (that is, are 3[2+3]). Prob- ably he confused the type specimen of reynaudi with a specimen of Capillaster multi- radiata also collected by M. Reynaud in Ceylon, but he may have been misled by Miiller’s original description. In 1888 Professor Bell recorded this species from Tuticorin, Madras, on the Indian shore of the Gulf of Manaar, whence specimens had been sent to the British Museum by Edgar Thurston of the Madras Government Museum. In 1891 Dr. Clemens Hartlaub inserted reynaudi in his key to the species of the Savignyi group of Antedon, using the same characters that were used by Carpenter in 1888. In 1894 Edgar Thurston recorded this species as living on the stems of gorgonians at Pamban, and also as being found along the shore in the Gulf of Manaar. In 1904 Herbert Clifton Chadwick recorded it from Ceylon, where it had been dredged at one of the stations occupied during the pearl oyster investigations of the Gulf of Manaar under the direction of Sir William A. Herdman. He gave an excellent description of the species and figured it. Later in the same year he recorded it in a supplementary report from another station, the specimen concerned having been listed in the original report as Antedon variipinna. In the first revision of the genus Antedon published by me in 1907, reynaudi was assigned to the new genus Himerometra, and in a revision of the family Himerometridae published in 1909 it was assigned to the new genus Heterometra. In 1911 I published notes on Miiller’s type specimen, which I had examined in the Paris Museum in 1910, and in a short paper on Indian crinoids, and also in my monograph on the crinoids of the Indian Ocean, both published in 1912, I recorded and gave notes on specimens taken in Ceylon and at various places in the Bay of Bengal, described a curiously abnormal specimen, and recorded a molluscan parasite (see page 307). Under the name of Heterometra bengalensis I recorded numerous additional specimens from various localities, all the specimens recorded as bengalensis being in reality reynaudi. The figure, said to represent a typical example of bengalensis in lateral view, was in reality drawn from a specimen of reynaudi. In 1913 I gave notes on the two specimens from Tuticorin in the British Museum, and in 1915 Dr. Hubert Lyman Clark recorded 55 additional specimens from Ceylon. He said that three of these are noticeably larger and have longer arms than the rest. It is possible, or even quite probable, that these large long-armed individuals represent H. amboinae instead of H. reynaudi. In my report upon the unstalked crinoids of the Siboga expedition published in 1918 I included a key to the species of Heterometra in which both reynaudi and bengal- A MONOGRAPH OF THE EXISTING CRINOIDS 311 ensis are given. The characters assigned to bengalensis were drawn up from immature individuals of reynaudi from the east coast of India, which were mistaken for bengalensis. HETEROMETRA ATER (A. H. Clark) PuatE 26, Ficures 109, 110 Alecto savignii (part) J. MtituER, Monatsb. Preuss. / kad. Wiss., 1841, p. 185; Archiv fiir Naturg., 1841, vol. 1, p. 144; Abh. Preuss. Akad. Wiss., 1841 (1843), p. 181. Comatula (Alecto) savignii J. Miituer, Abh. Preuss. Akad. Wiss., 1847 (1849), p. 257 (in part). Antedon ludovici Berl. Mus., MS. A. H. Cuarx, Proc. U. 8. Nat. Mus., vol. 40, 1911, p. 21. Craspedometra ater A. H. Cuarx, Proc. U. S. Nat. Mus., vol. 40, 1911, p- 9 (northeastern Africa), p. 21 (description; Red Sea); vol. 48, 1912, p. 383 (= Alecto savignii of J. Miiller, in part), p. 384 (original reference), p. 393 (synonymy; description; Red Sea); Crinoids of the Indian Ocean, 1912, p. 119 (synonymy; Red Sea). Heterometra ater A. H. Cuarx, Unstalked crinoids of the Siboga-Exped., 1918, p. 78 (in key; range).— Gustin, Kungl. Fysiogr. Sillsk. Handl., new ser.,; vol. 45, No. 11, 1934, pp. 22, 48.—A. H. Cuarx, John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1936, p. 99 (range), p. 104. Diagnostic features—The brachials are distinctly wedge-shaped, with the ends never quite parallel, and are not exceedingly short. The enlarged lower pinnules are smooth, with the earlier segments keeled. The cirri are 30-33 mm. long and are composed of 32-36 subequal segments all of which are half again as broad as long, the outer bearing a small blunt median spine. The 14 arms are about 160 mm. long. P, is half again as long as P; and much larger and stouter, 22 mm. long with 30 seg- ments. PP; is of the same length as P, and is much smaller and slenderer than P>. Description.—The centrodorsal is thick discoidal, with the bare polar area slightly convex, 4 mm. in diameter. The cirrus sockets are arranged in one and a partial second crowded and irregular marginal row. The cirri are XXII, 32-36, from 30 mm. to 33 mm. in length, stout and not tapering distally. The first segment is short, and those following gradually increase in length to the fifth or sixth, which, with the remainder, is about half again as broad as long. On the fifteenth or sixteenth the middle of the distal dorsal edge becomes prominent, this prominence after one or two more segments becoming a prominent rather high rounded carination of the entire mediodorsal line. Distally this carination gradually narrows longitudinally so that on the last four or five segments preceding the penultimate there is only a blunt median spine. The opposing spine is usually rather longer and sharper than the spines on the preceding segments, terminal or sub- terminal, directed obliquely forward, half as long as the width of the penultimate segment. The terminal claw is somewhat longer than the penultimate segment, stout and strongly curved basally but becoming straighter and slenderer in the distal two-thirds. The radials are almost entirely concealed by the centrodorsal in the median line, but their anterolateral portions are visible as very low broad triangles in the interradial angles of the calyx. The IBr, are very short, bandlike, five or six times as broad as long, in close apposition laterally. The [Br (axillaries) are very broadly pentagonal, twice as broad as long. The IIBr series are 4 (8+4) and are well rounded dorsally. The division series as far as Pp are in lateral apposition, though they are not especially flattened. The synarthrial tubercles are obsolete. 312 BULLETIN 82, UNITED STATES NATIONAL MUSEUM The 14 arms of the type specimen are about 160 mm. in length. The first brachials are very short, slightly wedge-shaped, almost entirely united interiorly. The second brachials are larger, more obliquely wedge-shaped. The first syzygial pair (composed of brachials 3+4) is oblong, from two and one-half to three times as broad as long. ‘The next five or six brachials are oblong, three to four times as broad as long, those succeeding becoming wedge-shaped, moderately oblique, about three times as broad as long, soon gradually becoming less and less oblique, and in the outer part of the arms being very short, three or three and one-half times as broad as long, and only slightly wedge-shaped. Pp is 11 mm. long, with 33 segments, moderately stout basally but tapering rapidly and becoming very slender and flagellate in the distal half. The segments at first are nearly three times as broad as long, but gradually increase in length and become nearly or quite as long as broad in the distal third. P, is 13 mm. long, con- siderably stouter than Pp and tapering much less rapidly, with 26 segments, of which the second-eighth are strongly carinate. P, is 22 mm. long, similar in general to P, though much larger and stouter and tapering less rapidly, with 30 segments of which the proximal are broad and those following become about as long as broad on the tenth or twelfth and remain so until near the tip; the second-fifth segments are strongly carinate. P, is 13 mm. long, much smaller and slenderer than P;. P, is somewhat smaller than P;. P,;is 9mm. long, somewhat smaller than P,; the proximal carination occurs on the second-sixth segments. The following pinnules very slowly become slenderer, at the same time gradually increasing in length. The distal carination of the proximal segments, which is very marked on the earlier pinnules, gradually be- comes less and less marked, involving fewer and fewer segments; it is but slightly evident beyond the twentieth. The distal pinnules are 10 mm. long. The color in alcohol is a uniform purplish or violet-black. Locality —Red Sea; Hemprich and Ehrenberg [J. Miiller, 1841, 1843, 1849; A. H. Clark, 1911, 1912, 1918] (1, Berl. M., 1055). History.—The single known specimen of this species was studied by Prof. Johannes Miller and is part of the material upon which he based his Alecto savignii described in 1841 and redescribed under the name of Comatula (Alecto) savignii in 1849. Later someone—probably Dr. P. H. Carpenter—in going over the collection of the Berlin Museum discovered this specimen and separated it out from the type series of Alecto savignii, labeling it Antedon ludovici. It was under this name that I found it in 1910. In 1911 I described it under the name of Craspedometra ater. In 1912 in my memoir on the crinoids of the Berlin Museum I again mentioned it, giving the early references to Miller. In my report on the unstalked crinoids of the Siboga expedition it was included in the key to the species of the genus Heterometra. In this report I referred to H. ater the specimen from Ceylon identified as Craspedometra amboinae by Reichensperger. I am now convinced that Reichensperger’s identification was correct and that the Ceylon specimen really represents amboinae. HETEROMETRA JOUBINI (A. H. Clark) Heterometra joubini A. H. Cuarx, Proc. U. S. Nat. Mus., vol. 40, 1911, p. 8 (southeastern Africa), p. 25 (description; Zanzibar); Bull. Mus. Hist. Nat., Paris, 1911, No. 4, p. 251 (detailed de- scription; Zanzibar); Crinoids of the Indian Ocean, 1912, p. 127 (synonymy; habitat); Un- A MONOGRAPH OF THE EXISTING CRINOIDS 313 stalked crinoids of the Siboga-Exped., 1918, p. 78 (in key; range).—Guisuin, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 22, 48 —A. H. Cuark, John Murray Exped. 1933-84, Sci. Reports, vol. 4, No. 4, 1936, p. 100 (range), p. 104. Diagnostic features.—The brachials are distinctly wedge-shaped, with the ends never quite parallel, and are not exceedingly short. The enlarged lower pinnules are smooth, with the earlier segments keeled. The cirri are 30 mm. long, with 39-43 subequal segments, all of which are about half again as broad as long and of which those in the outer half bear short dorsal spines. The 20 arms are 80 mm. long. P,» is the longest pinnule, half again as long as P, or P;, and is slender, with 21 segments. Description.—The cirri are XXII, 39-43, 30 mm. long. The segments are all subequal, about half again as broad as long. Those in the outer half of the cirri bear short dorsal spines. The cirri as a whole are comparatively long and moderately stout. Ten I1Br series of 4(3-+-4) are present. The arm divisions and the arm structure are as in H. reynaudi. The 20 arms are 80 mm. long. P2 is the longest pinnule, half again as long as P, or P3, slender, with 21 segments, which become about as long as broad on the fourth or fifth, then somewhat longer than broad, and elongate terminally. P; is about aslong as P;. P, and the following pinnules are shorter than P;. The lower segments of the proximal pinnules are moderately carinate. Remarks.—When compared directly with Miiller’s type specimen of H. reynaudi this species is seen to differ in its longer and somewhat more slender cirri, which are composed of slightly longer segments. Locality.— Zanzibar; M. Rousseau, 1841 [A. H. Clark, 1911, 1912, 1918] (1, P. M.). History.—The type specimen of this species was examined by Dr. P. H. Car- penter during a visit to the Paris Museum, and he placed a label with it indicating that it represented a new species. He did not, however, mention it in any of his publications. It was first described in 1911 in a memoir on the crinoids of the coasts of Africa and also in a report on the recent crinoids in the Paris Museum. HETEROMETERA PHILIBERTI (J. Miiller) PuaTE 28, Ficurss 118, 119; Puare 30, Fiaurns 124-126 [See also vol. 1, pt. 1, fig. 258 (centrodorsal), p. 255; fig. 335 (cirrus), p. 283; fig. 473 (centrodorsal), p. 361; pt. 2, figs. 34-36 (radial pentagon), p. 20; figs. 447, 448 (pinnule tip), p. 261.] Comatula philiberti J. Miittumr, Abh. Preuss. Akad. Wiss., 1847 (1849), p. 259 (description; Java).— Dusarpvin and Hup#, Histoire naturelle des zoophytes, Echinodérmes, 1862, p. 205 (synonymy; description; Java) —A. H. Ciark, Bull. Mus. Hist. Nat., Paris, 1911, No. 4, p. 245 (identity); Crinoids of the Indian Ocean, 1912, p. 30 (same). Antedon philiberti P. H. Carrenter, Trans. Linn. Soc. (Zool.), ser. 2, vol. 2, 1879, p. 29 (listed as an Antedon).—BeEt1, Proc. Zool. Soc. London, 1882, p. 533 (listed).—P. H. Carpenter, Journ. Linn. Soe. (Zool.), vol. 16, 1882, p. 513-(compared with Antedon bipartipinna [Craspedometra acuticirra]); Proc. Zool. Soc. London, 1882 (1883), pp. 743 ff. (discussion of Bell’s method of formulation and corrected formula); Challenger Reports, Zoology, vol. 26, pt. 60, 1888, pp. 55, 253, 366, 380 (discussion).—Hartiaus, Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 19 (in key) —Mrunckenrt, Archiv fiir Naturg., Jahrg. 71, 1905, vol. 1, Heft 1, p. 228 (regenera- 208244—40 21 314 BULLETIN 82, UNITED STATES NATIONAL MUSEUM tion).—Ham ann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1582 (listed). —A. H. Cuarx, Crinoids of the Indian Ocean, 1912, pp. 35, 110 (identity of records) ; Smithsonian Misc. Coll., vol. 61, No. 15, 1913, p. 86 (same). Himerometra philiberti A. H. Cuark, Smithsonian Mise. Coll., vol. 50, 1907, p. 356 (listed); Proc. Biol. Soc. Washington, vol. 22, 1909, p. 7 (listed). Amphimetra mortenseni A. H. Cuark, Proc. U. S. Nat. Mus., vol. 36, 1909, p. 635 (description; Port Blair, Andamans); Bull. Mus. Hist. Nat., Paris, 1911, No. 4, p. 250 (synonym of philibertz); Crinoids of the Indian Ocean, 1912, p. 107 (same); Unstalked crinoids of the Siboga-Exped., 1918, p. 77, footnote 2 (same). Amphimetra philiberti A. H. Cuarx, Proc. U. 8. Nat. Mus., vol. 40, 1911, p. 21 (nearest relative of Amphimetra (Heterometra] africana; differential characters); Bull. Mus. Hist. Nat., Paris, 1911, No. 4, p. 245 (=Comatula philiberti J. Miiller), p. 250 (Java; detailed description of the type; includes mortenseni); Smithsonian Misc. Coll., vol. 60, No. 10, 1912, p. 13 (Kwala Cassan, Malay peninsula; description of specimens), pp. 14, 15 (compared with Amphimetra schlegelit) ; Proc. U. 8. Nat. Mus., vol. 43, 1912, p. 393 (compared with Amphimetra [Heterometra] africana) ; Crinoids of the Indian Ocean, 1912, pp. 30, 35 (identity), p. 107 (synonymy; detailed descrip- tion; localities; notes on the type specimen), p. 316 (Andamans), fig. 7, p. 108. Heterometra philiberti A. H. Cuark, Unstalked crinoids of the Siboga-Exped., 1918, p. 77 (in key; range).—GuisLt6n, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, p. 28. Diagnostic features.—The very short discoidal brachials in the outer half of the arms combined with the presence of IIIBr series all, most, or at least some of which are 4(3-++-4), and the more or less swollen elements of the division series and earlier brachials easily distinguish this species. The cirri have 26-45 (usually about 35) segments of which the longest are nearly as long as broad. There are 18-27 arms 80-150 mm. long. Description.—The centrodorsal is thick discoidal, with the bare polar area flat, 4 or 5 mm. in diameter. The cirrus sockets are arranged in two closely crowded alternating marginal rows. The cirri are XVIII-XX, 30-42 (usually about 35), 25 to 30 mm. Jong. The first segment is short, about three times as broad as long, the second and third are about twice as broad as long, and those following gradually increase in length to the ninth or tenth, which is nearly, though never quite, as long as broad. The next five or seven segments are similar; those following gradually decrease in length so that those in almost the whole of the terminal half of the cirri are about half again as broad as long. From the twelfth or fourteenth segment onward sharp median tubercles or small spines are developed on the dorsal side, those on the last few seg- ments occupying a position slightly proximal to median. The opposing spine is much larger than the processes on the preceding segments, triangular, the apex median, arising from very nearly the whole dorsal surface of the penultimate segment, equal to about half the width of that segment in height. The terminal claw is longer than the penultimate segment, moderately stout basally, but gradually becoming slender distally, and is moderately curved. The radials are concealed, or just visible beyond the rim of the centrodorsal. The IBr, are oblong, very short, in close lateral apposition. The IBr, (axillaries) are very broadly pentagonal, almost triangular, the lateral edges not quite so long as those of the IBr,, about two and one-half times as broad as long. The IIBr series are 4(3+4). The IIIBr series are 4(3++-4). The division series and first two brachials are in close lateral apposition and laterally flattened against their neighbors, the dorsal A MONOGRAPH OF THE EXISTING CRINOIDS 315 carination of Pp only being visible exteriorly between them; synarthrial tubercles are usually prominent. The 20-25 arms are 150 mm. long. The first brachial is slightly wedge-shaped, short, about three times as broad as its exterior length, almost entirely united inte- riorly. The second is of about the same size, but more pronouncedly wedge-shaped. The first syzygial pair (composed of brachials 3+4) is oblong, half again as broad aslong. The next five or six brachials are oblong, nearly or quite four times as broad as long, and those following slowly become wedge-shaped and later almost triangular, four times as broad as long, soon becoming wedge-shaped again and, in the distal half of the arm, oblong and very short, though somewhat longer again terminally. The proximal discoidal brachials are somewhat swollen, and most of the brachials have slightly overlapping distal ends. Syzygies occur between brachials 3+4, again from between brachials 13+ 14 to between brachials 31-++32 (usually somewhere between the sixteenth and twenty- fifth, with sometimes an extra one from two to four or five brachials beyond the first), and distally at intervals of 2 to 13 (usually § to 12) muscular articulations. Pp is 7 mm. long, moderately stout basally but tapering rapidly and becoming slender in its distal half, with about 25 segments, which at first are three times as broad as long, becoming twice as broad as long at the sixth, and squarish in the terminal portion; some of the lower segments are bluntly carinate. P, is 10 mm. long with 30 segments, slightly less stout basally than Pp and tapering somewhat less rapidly; the segments are at first about twice as broad as long, becoming as long as broad on about the eighth, and somewhat longer than broad terminally. P, is 15 mm. long, stouter than P,, tapering evenly to a delicate tip, with 30 segments which at first are about half again as broad as long, become squarish on the eighth or ninth, and are about twice as long as broad at the tip. P; is 22 mm. long, stouter than the preceding, with 30 segments, which at first are broader than long, becoming squarish on the tenth, and longer than broad terminally; the pinnule is more or less carinate in its proximal half, and has a moderately developed supplementary ridge on the distal half of the outer side. P, resembles P; but is very slightly longer and proportionately stouter and more carinate. P,;is like P;. P,is 10 mm. long, resembling P; but some- what more strongly carinate proximally. The following pinnules gradually decrease to 7 mm. in length and lose the basal carination, then increase to 12 mm. distally. On some arms P; is small as described for P,, and again P, may also be small, while occasionally P, and P, are similar, and P; is greatly enlarged; sometimes P,, P; and P, are as described for P;, Py and P;. On one or more of the inner arms of each ray P; is often much larger than on the outer, while the adjacent pinnules are reduced in size. Notes.—The preceding description was taken from the specimen from Port Blair, Andaman Islands, which served as the type of Amphimetra mortensent. The six specimens from Kwala Cassan are all rather smaller than those from the Andaman Islands and apparently are not quite fully grown. One has 24 arms 100 mm. long. One of the IIIBr series is 2, all the other division series beyond the IBr series being 4(3+-4). The cirri have 32-33 segments, of which the longest are very nearly as long as broad. 316 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Another specimen has 22 arms 100 mm. long. Four of the IIBr series and two of the IIIBr series are 2, the I1IBr 2 series being both internally developed. The cirri have 33-34 segments. A third specimen has 22 arms 95mm. long. Both the IIBr series, one externally and one internally developed, are 2, all the other division series beyond the IBr series being 4(3+4). The cirri are XXIII, 29-33, 23 mm. long. A fourth specimen has 27 arms 90 mm. long. Of the IBr series one is 2 and the rest are 4(3-+-4). Six of the I//Br series are 2, only one being 4(3+4). The cirri have 26-29 segments. A fifth example has 18 arms 85 mm. long. One postradial series bears two undivided arms. The 8 IJBr series present are 4(3-+4). The cirri have 27-28 segments. One cirrus has the distal half regenerated. The sixth specimen has 21 arms 80 mm. long. One of the IIBr series is 2, the other 9, the single IJIBr series being 4(3+-4). The cirri have 26-30 segments of which the longest are nearly as long as broad. The specimen from Java was described by Prof. Johannes Miller in the following terms: The centrodorsal is flat in the middle and bears about its margin a row of cirri. The cirri have 45 segments which are not longer than broad and of which the greater part, except those at the base, bear a swollen process. There are three radials (that is, there is one radial and the two elements of the [Br series); the radial axillary is without a syzygy (that is, the elements of the [Br series are united by synarthry). The IIBr series are 4(3+-4), the second element resting obliquely on the first. There are sometimes present IIIBr 4(3+4) series. The second element of the II/Br series and the second brachial of the free undivided arms lie obliquely on the ossicle preceding, as in the case of the second element of the I[Br series. There are 26 arms. The brachials following the second are very short. The two first pinnules are small, the two following are larger, and those succeeding become smaller. I examined Miiller’s type specimen at the Paris Museum in 1910 and found it to represent the form I had described in 1909 as Amphimetra mortenseni. The cirri are XXIII, 38-44. There are 25 arms. The 10 IIBr series are 4(3+4). Of the five IlIBr series present, four are 4(3-++4) and one is 2. The lateral borders of the division series are produced into a lateral shelf. The synarthrial tubercles are small, but prominent, as in the type specimen of Amphimetra discoidea from Port Molle. The brachials are very short, with their distal borders slightly produced. Pp is very small and weak. P, is a little larger than Pp, and P, is sometimes a little larger than P, or sometimes very much larger than P, and resembling P;. P; is the largest pinnule, long, rather stout, and composed of 26 segments of which the majority are about as long as broad. P, resembles P; and reaches almost the same size, or is a little shorter than P,. The following pinnules are short. P, is stouter, longer, and more stiffened than is usually the case in related species. Localities.—Investigator; Port Blair, Andaman Islands [A. H. Clark, 1909, 1911, 1912, 1918] (1, I. M.) Investigator; Andaman Islands [A. H. Clark, 1912] (5, U. S. N. M., 35176, 35190, 35203, 35219, 36240, 36241; I.M.). Kwala Cassan, Malay Peninsula; Ed. L. Mayer [A. H. Clark, 1912] (6, H. M.). A MONOGRAPH OF THE EXISTING CRINOIDS 317 Java; M. Philibert [J. Miller, 1849; Dujardin and Hupé, 1862; P. H. Carpenter, 1879, 1882, 1883, 1888; Bell, 1882; Hartlaub, 1891; A. H. Clark, 1907, 1909, 1911, 1912, 1913, 1918] (1, P. M.). Geographical range—From the Andaman Islands to the Malay Peninsula and Java. Bathymetrical range.—Littoral. History.—This species was first described under the name of Comatula philiberti by Prof. Johannes Miiller in 1849 from a specimen from Java collected by M. Philibert which he had personally studied in the Paris Museum. The original description was translated and republished by Dujardin and Hupé in 1862. In 1879 Dr. P. H. Carpenter referred this species to the genus Antedon, and in 1882 he compared it with his new Antedon bipartipinna (=Craspedometra acuticirra) to which he believed it to be closely related. Prof. F. Jeffrey Bell in 1882 published a specific formula for this form, and early in the following year Carpenter published a corrected and emended formula. As this species was not among the comatulids secured by the Challenger, Car- penter did not discuss it in detail in the Challenger report. He merely inserted it in the key to the species of the Savignyi group in which it was paired with Antedon bipartipinna, these two species being distinguished in the key from all the other species of this group by having the IIIBr series 4(3-++-4). Under this common heading he characterized philiberti as having 45 cirrus segments, the later ones short and spiny, whereas in bipartipinna there are nearly 60 cirrus segments of which the later are longer than broad and quite smooth. Dr. Clemens Hartlaub in his key to the species of the Savignyi group published in 1891 put bipartipinna with (Himerometra) crassipinna under a section characterized by having the IIIBr series both 2 and 4(3+-4), while philiberti was placed alone in a section characterized by having the IIIBr series all 4(3+-4). In my first revision of the genus Antedon published in 1907 I assigned philiberti to the new genus Himerometra, and in my revision of the family Himerometridae published in 1909 I retained philiberti in Himerometra, though with a query. Having received for study the extensive collections assembled by the Royal Indian Marine Surveying steamer Investigator, I found among them a specimen from Port Blair in the Andaman Islands that seemed to represent a very distinct new species, quite different from any form with which I was personally acquainted. This I described in 1909 under the name of Amphimetra mortensent. In the following year I visited Paris and there examined Miiller’s type specimen of philiberti. I saw at once that it represented the same species as my recently de- scribed Amphimetra mortenseni. In 1911 I published notes on this specimen, using the name Amphimetra philiberti, and said that my A. mortenseni was a synonym of this species. ; In my monograph on the crinoids of the Indian Ocean published in 1912 I repub- lished my description of Amphimetra mortensent under the heading Amphimetra philiberti, and figured a typical specimen of the former. I recorded four additional specimens from the Andaman Islands in the main portion of this work and another in the appendix, and mentioned Java and Kwala Cassan as the other localities from which it is known. 318 BULLETIN 82, UNITED STATES NATIONAL MUSEUM In the same year I recorded and gave notes upon six specimens from Kwala Cassan in the Malay Peninsula that I had found in the Hamburg Museum. In my report published in 1918 upon the unstalked crinoids collected by the Siboga expedition, philiberti is included in the key to the species of Heterometra, and in a footnote I said that this species includes Amphimetra mortenseni A. H. Clark, 1909. As the species was not collected by the Siboga it is not mentioned further. HETEROMETRA SARAE, 5p. nov. Diagnostic features—There are no IIIBr series. The brachials are exceedingly short, and those in the proximal portion of the arms have the dorsal surface swollen. The cirri are 20-30 mm. long and are composed of 25-33 (usually 30-33) segments of which the longest are about as long as broad and the outer are slightly broader than long and bear a small carinate dorsal spine. The 19 arms are 145 mm. long. The carination on the earlier segments of the enlarged proximal pinnules is almost obsolete. Description.—The centrodorsal is discoidal with the bare dorsal pole flat, 5 mm. in diameter. The cirrus sockets are arranged in one and a partial second marginal row. The cirri are XII, 25-33 (usually 30-33), from 20 to 30 mm. long. The first segment is short and those following slowly increase in length, becoming nearly or quite as long as broad on the eighth; the succeeding segments are subequal, distinctly, but not greatly, broader than long. On the sixth-ninth and one or two following segments there is a small sharp dorsal tubercle situated in the median line just beyond the middle of the segment; this almost immediately becomes a blunt or sharp small carinate spine with the apex subterminal, on the outermost segments moving to median. The opposing spine is conical, erect, in height equal to about half the width of the penultimate segment. The terminal claw is short, scarcely or not at all exceed- ing the penultimate segment in length, stout, and strongly curved. The radials are concealed, or are visible as a thin finely beaded line just above the rim of the centrodorsal. The IBr, are exceedingly short, 8 or 10 times as broad as long, with the proximal and distal edges parallel and the lateral borders in close apposition with those of their neighbors on either side. The IBr, (axillaries) are exceedingly short, three and one-half or four times as broad as long, triangular with the lateral angles acute or slightly truncated and the distal edges almost straight. Of the nine IIBr series present, seven are 4 (3+4) and two are 2. The division series are broad, in lateral apposition and laterally flattened, and are composed of unusually short segments, the terminal syzygial pair being twice as broad as long or even broader. The 19 arms are 145 mm. long. The brachials from the first are extremely short; up to the ninth they have parallel ends, the ends then becoming slightly oblique, and at the end of the proximal quarter of the arm practically parallel again. The earlier brachials have a somewhat swollen dorsal surface, and beyond the basal quarter somewhat prominent, though unmodified, distal ends. P, is 8 mm. long, small and weak, with 26 segments. It tapers rather rapidly in the first six segments and beyond this is delicate and flagellate. The first segment is about two and one-half times as broad as long and those following gradually increase in length to the eighth which, with most of those succeeding, is about as long as broad. P, is 15-18 mm. long, stout, tapering very slowly to the tip, and stiffened, with 25 A MONOGRAPH OF THE EXISTING CRINOIDS 319 segments, which become as long as broad on the sixth and twice as long as broad terminally. The outer surface of the pinnule from about the sixth segment onward is sharpened or gabled in the median line, and there is a slight sharpening of the distal border of the basal segments. P; resembles P, but is longer and noticeably stouter, especially in the outer half. P, may resemble P; or it may be shorter and more rapidly tapering than P,. P; is 11 mm. long, with 22 segments, tapering rather rapidly in the proximal half, with the distal half slender and delicate. The following pinnules resemble P;. The proximal segments of the proximal and middle pinnules are not carinate; at most they are slightly gabled or sharpened on the distal border. Locality Danish Expedition to the Kei Islands; Dr. Th. Mortensen; Amboina; about 2 meters; February 9, 1922 (1, C. M.). Remarks.—This species seems to be very distinct from any previously described, the entirely smooth, enlarged, and much stiffened proximal pinnules, the very slight carination on the basal segments of the lower pinnules, and the exceedingly short elements of the division series and brachials distinguishing it at once. HETEROMETRA PARILIS (A. H. Clark) Puatse 35, Ficures 157-160 [See also vol. 1, pt. 2, fig. 796 (ambulacral deposits), p. 372.] Himerometra bengalensis A. H. Cuarx, Smithsonian Misc. Coll., vol. 52, 1908, p. 214 (Albatross station 5146). Amphimetra parilis A. H. Cuarg, Proc. U. S. Nat. Mus., vol. 37, 1909, p. 32 (description; Albatross station 5147); Crinoids of the Indian Ocean, 1912, p. 112 (synonymy; locality); Unstalked crinoids of the Siboga-Exped., 1918, p. 82 (in key; range), p. 84 (references; localities from which known). Heterometra bengalensis A. H. Cuark, Crinoids of the Indian Ocean, 1912, p. 130 (in part; Philip- pines); Monograph of the existing crinoids, vol. 1, pt. 2, 1921, p. 240, fig. 796, p. 372 (ambulacral deposits); Smithsonian Misc. Coll., vol. 72, No. 7, 1921, pl. 2, fig. 21 (ambulacral deposits; figure from preceding). Diagnostic features.—There are no IIIBr series. The brachials are exceedingly short, and those in the proximal portion of the arm have the proximal end elevated so that the profile of the earlier portion of the arm is regularly serrate with the teeth directed proximally. The profile of the outer two-thirds of the arms is smooth. The cirri are 25-30 mm. long and are composed of 30-33 segments, of which the longest are nearly as long as broad and the outer are about twice as broad as long with promi- nent median dorsal spines. The 10-16 arms are 100-150 mm. long. The carination on the earlier segments of the lower pinnules is strongly marked, though not exag- gerated. Description.—The centrodorsal is thick discoidal with the bare polar area flat 2.5 mm. in diameter. The cirrus sockets are arranged in one and a partial second crowded and irregular marginal rows. The cirri are XII, 30-33, 25 to 30 mm. long, moderately stout. The segments are subequal in length, all broader than long. They are at first very short, slowly increasing in length to the seventh or eighth, which, with the three or four following, is about half again as broad as long, then very slowly decreasing so that the segments in the distal half of the cirri are about twice as broad as long. The twelfth and fol- lowing segments have prominent median dorsal spines which are directed distally. 320 BULLETIN 82, UNITED STATES NATIONAL MUSEUM The opposing spine is longer than the spines on the preceding segments, sharp, tri- angular in profile, rather slender, with the apex median; it reaches a length about equal to half the width of the penultimate segment. The terminal claw is longer than the penultimate segment, slender, more strongly curved proximally than distally. The radials are just visible in the midradial line beyond the rim of the centro- dorsal and form low triangles in the interradial angles of the calyx; their anterolateral corners are slightly swollen. The IBr, are oblong, four times as broad as long, and almost entirely united laterally. The IBr. (axillaries) are almost or quite triangular, somewhat over twice as broad as long, in lateral apposition with their neighbors and laterally flattened. The 10 arms are 150 mm. long and taper rather more gradually than those of A. discoidea. The first two brachials are short, wedge-shaped, the second slightly the longer exteriorly but tapering almost to a point interiorly, the first entirely united interiorly. The first syzygial pair (composed of brachials 3+-4) is oblong, two and one-half or three times as broad as long. The next seven or eight brachials are oblong, very short, about four times as broad as long or even shorter, those succeed- ing becoming wedge-shaped, about the same length, and in the distal half of the arm oblong and very short. In the proximal third of the arm the proximal edge of the brachials is somewhat raised, but the remainder of the arm is perfectly smooth. Syzygies occur between brachials 3-+-4, again between brachials 9+-10 (this some- times omitted) and 14+15 or 15+16, and thence at intervals of 7 to 17 muscular articu- lations up to about the middle of the arm, beyond which point syzygies are rare or entirely lacking. P,; is 9 mm. long, broad basally but tapering rapidly and becoming slender and flagellate in the distal half, composed of 30 segments, of which the first eight are broader than long, very considerably so at first, and the remainder are about as long as broad. P, is 13 mm. long, stout like P, in the basal third but tapering rapidly and becoming slender and flagellate distally, with 30 segments, of which the first six are broader than long and the remainder are about as long as broad. The broad lower segments, like those of P,, are carinate. P; is 19 mm. long, much stouter than P, or P, but of the same general form, stout basally but becoming gradually slender distally with a flagellate tip; it is composed of 30 segments. P, is 20 mm. long, stouter and stiffer than P; with about 25 segments, of which the first 10 are broader than long and those following are about as long as broad becoming slightly longer than broad distally. Like P;, P, is carinate in its basal half and has a low lateral keel in its outer two-thirds. P,; is 10 mm. long, about as stout basally as P, but not tapering so rapidly, with 17 segments, of which the first six are longer than broad and the remainder are about as long as broad; the first seven segments are strongly carinate. P, and the following pinnules are 7 mm. long, with 16 segments, slender, about as stout as P, proximally but not tapering so rapidly, the first five or six seg- ments broader than long and strongly carinate, those following about as long as broad, and finally slightly longer than broad. The distal pinnules are 10 mm. long and moderately slender. The carination of the basal pinnule segments gradually dies away at the end of the proximal fourth of the arm. Notes.—The specimen from Albatross station 5146 has 16 arms 100 mm. long. Of the six IIBr series present, five are 4(3-+-4) and one is 2. The brachials in the A MONOGRAPH OF THE EXISTING CRINOIDS 321 proximal fourth of the arms have produced proximal ends that overlap the distal ends of the preceding brachials; the brachials beyond the proximal fourth of the arms are entirely smooth, joining perfectly end to end. Remarks.—This species is very close to H. sarae from which it differs in lacking the strongly swollen dorsal surface of the ossicles of the division series and brachials, in having much shorter distal cirrus segments, and in having the earlier segments of the lower pinnules conspicuously carinate. Localities.— Albatross station 5147; Sulu (Jolo) Archipelago, in the vicinity of Siasi; Sulade Island (E.) bearing N. 3° E., 8.4 miles distant (lat. 5°41’40’’ N., long. 120°47’10’’ E.); 38 meters; coral sand and shells; February 16, 1908 [A. H. Clark, 1909, 1912, 1918] (i, U.S.N.M., 25515). Albatross station 5146; Sulu (Jolo) Archipelago, in the vicinity of Siasi; Sulade Island (E.) bearing N. 18° W., 3.4 miles distant (lat. 5°46’40’’ N., long. 120°48’50’” E.); 44 meters; coral sand and shells; February 16, 1908 [A. H. Clark, 1908, 1912, 1918] (1, U.S.N.M., 35089). History.—The first known specimen of this species was recorded by me without comment in 1908 under the name of Himerometra bengalensis from Albatross station 5146. In the following year a 10-armed specimen from station 5147 was described as a new species under the name of Amphimetra parilis. The first specimen served as the basis for the inclusion of the Philippines in the range of Heterometra bengalensis in my monograph of the crinoids of the Indian Ocean published in 1912, and the description and figure of the ambulacral deposits of H. bengalensis in Part 2 of the present work were likewise based upon it. But the error in determination was discovered before the report on the unstalked crinoids of the Siboga expedition was published (1918), although at that time the specimen was not redetermined. In the monograph of the crinoids of the Indian Ocean (1912), Amphimetra parilis was simply listed as a distinct species; but in the report on the Siboga crinoids (1918) one of the two specimens included by Grube in his Comatula laevissima was included under A. parilis as a questionable synonym (see pages 349 and 353). HETEROMETRA BENGALENSIS (Hartlaub) PuaATE 30, Figures 127-130 [See also vol. 1, pt. 2, figs. 457-462 (pinnule tips), p. 266.] Antedon bengalensis Hartuaus, Nachr. Ges. Gottingen, May 1890, p. 182 (description; Bay of Bengal); Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 17 (in key), p. 19 (detailed descrip- tion and comparisons; Bay of Bengal), p. 113 (in Gottingen Mus.), pl. 1, fig. 2, pl. 2, fig. 16.— HeErpMaN, Ceylon Pearl Oyster Rep., vol. 5, Suppl. Rep. 41, 1906, p. 449 (redetermination of Chadwick’s A. anceps)—Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1582 (listed). —A. H. Cuarx, Crinoids of the Indian Ocean, 1912, p. 36 (identity). Antedon anceps CHapwick, Rep. Ceylon Pearl Oyster Fisheries, pt. 2, Suppl. Rep. 11, 1904, p. 153 (occurs at Ceylon), p. 157 (stations xxiv and Ly1t)—Herpman, Ceylon Pearl Oyster Rep., vol. 5, Suppl. Rep. 41, 1906, p. 449 (correction of preceding record).—A. H. Cuarx, Crinoids of the Indian Ocean, 1912, p. 40 (same). Himerometra bengalensis A. H. Cuarx, Smithsonian Mise. Coll., vol. 50, 1907, p. 356 (listed). Heterometra bengalensis A. H. Cuarx, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 11 (listed); Die Fauna Siidwest-Australiens, vol. 3, Lief. 18, 1911, p. 440 (in part; Australian specimens are Zygometra punctata); Mem. Australian Mus., vol. 4, 1911, p. 722 (same); Crinoids of the 322 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Indian Ocean, 1912, p. 36 (identity), p. 40 (includes anceps of Chadwick, 1904), p. 1380 (Bay of Bengal [from Hartlaub] and Ceylon [from Chadwick]; Philippines [=H. parilis]; specimens recorded are H. reynaudi; records from Queensland and Holothuria Bank refer to Zygometra punctata), p. 318 (specimens recorded are all H. reynaudi), fig. 12, p. 181, is H. reynaudi; Die Fauna Siidwest-Australiens, vol. 4, Lief. 6, 1913, p. 310 (all Australian records refer to Zygometra punctata), pp. 313, 314 (same).—H. L. Cuarx, Spolia Zeylanica, vol. 10, pt. 37, 1915, p. 93 (occurs at Ceylon).—A. H. Crarx, Unstalked crinoids of the Siboga-Exped., 1918, p. 79 (char- acters given in the key, and ‘eastern coast of India” refer to H. reynaudi); Smithsonian Misc. Coll., vol. 72, No. 7, 1921, pl. 2, fig. 21 (ambulacral deposits).—GisLin, Kungl. Fysiogr. Siillsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 22, 23. Heterometra aspera A. H. Cuarx, Vid. Medd. Naturh. Foren. Kgbenhavn, 1909, p. 162 (description; Singapore), p. 193 (collected at Singapore by Svend Gad); Crinoids of the Indian Ocean, 1912, p. 128 (synonymy; Singapore); Unstalked crinoids of the Siboga-Exped., 1918, p. 79 (in key; range); A. H. Cuarx, Journ. Linn. Soc. (Zool.), vol. 36, No. 249, 1929, p. 638 (Singapore; 7 fathoms); Proc. Biol. Soc. Washington, vol. 47, 1934, p. 11 (Singapore).—Gistin, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, p. 22. Diagnostic features—The brachials beyond the proximal fourth of the arms are exceedingly short, discoidal, with the ends parallel, and the distal ends of the lower brachials are produced and more or less strongly everted. There are no IIIBr series. The cirri are 13-22 mm. long, with 22-31 (rarely so many as 30) segments of which the longest are somewhat longer than broad. The 11-18 arms are 50-100 mm. long. P, and P; are similar, 9 mm. long with 24 segments, or P; is somewhat smaller than P,. Description —The centrodorsal is a rather large convex disk with inwardly sloping sides; the convex bare dorsal pole is sculptured with small pits. The cirri are arranged in one and a partial second marginal rows. The cirri are XVII, 22-24, about 13 mm. long. The fifth-eighth segments are somewhat longer than broad, and the outer are slightly compressed laterally, from the ninth onward bearing a well developed spine. The radials are partially visible. The IBr, are short and broad, laterally either free or partially united. The JBr, (axillaries) are short, broad, and 5-sided. One IIBr 4(3+-4) series is present. The division series have flattened sides from the IBr, to the second brachial. They are bordered by low shelves which are somewhat pro- duced and are sharply set off from the rounded dorsal part of the segments. The 11 arms are apparently 50 mm. long. They have a bluntly serrate dorsal profile. The brachials are very short. The first eight brachials, the first syzygial pair not excepted, are as short as those following. They have, especially the second, alternating slight lateral overlappings at the distal end. The following brachials are wedge-shaped with somewhat produced distal borders, those succeeding later becoming more discoidal and remaining short to the tips of the arms. Syzygies occur between brachials 3+-4, again between brachials 9+-10 or 10+11, and distally at intervals of 3 to 6 (usually 4) muscular articulations. Toward the ends of the arms the intersyzygial interval becomes somewhat larger. In arms arising from a I1Br axillary the second syzygy is about brachials 16+17, and the intersyzygial interval is 8 to 10 muscular articulations. The lower pinnules are rather stiff. Pp, and on the arms arising from a [Br axillary P,, is short and is composed of about 20 segments. P, and P, are of about equal length, 7 mm. long. The following pinnules decrease gradually in length to those of the sixth pair, which are the shortest. From this point they increase rapidly A MONOGRAPH OF THE EXISTING CRINOIDS 323 in size up to their final length of about 5mm. P,, P2, and P; are somewhat longer than the corresponding pinnules (P,, P,, and P,) on the inner side of the arm. The proximal segments of all these pinnules are short and broad, and in the eight or nine first pairs are prominently and sharply keeled. The disk is 8 mm. in diameter and is deeply incised. The sacculi are large and crowded on the pinnules, forming two very irregular rows. They are not found on the disk. The centrodorsal and cirri are white, the arms are flesh colored with chocolate- brown bands between the brachials, and the disk is light gray-brown. The preceding description is adapted from Hartlaub’s description of the type specimen in the Géttingen Museum. Notes.—The two specimens collected during the course of the Ceylon Pearl Oyster Fisheries investigations, one at station Lvi and the other at station xx1v, both have 11 arms, a single IIBr 4(3+-4) series being present in each. The color when living was black. The type specimens of H. aspera from Singapore may be thus described: The centrodorsal is moderately thick, discoidal, with the bare polar area slightly convex, 3 mm. in diameter, and having a more or less pitted surface. The cirri are XVII, 26-31, from 17 mm. to 22 mm. in length, and moderately slender. The first segment is short and those following gradually increase in length to the fifth or seventh, which is about aslong as broad. The succeeding segments are similar, or slightly longer than broad, after about the twelfth or thirteenth, which is a more or less marked transition segment, decreasing in length, soon becoming as long as broad and so continuing to the end. The segments beyond the transition segment have moderate sized blunt dorsal spines. The opposing spine is much larger than the spines on the preceding segments. The radials are almost entirely concealed by the centrodorsal. The IBr, are short and bandlike. The IBr, (axillaries) are almost triangular, two to two and one-half times as broad as long. The IIBr series are 4(3+4). The division series as far as the base of Pp are in close apposition and laterally flattened against each other. The synarthrial tubercles are rather prominent. The arms are 14-18 in number, 60 mm. long. The brachials are very short with, except for the first few discoidal bra- chials, strongly produced distal ends. Pp is 7 mm. long, moderately stout basally but tapering rapidly in its proximal half and becoming slender and flagellate distally. It is composed of 24 segments, which are at first twice as broad as long or even broader but become as long as broad distally. The larger proximal segments are strongly carinate. P, is similar to Pp, of the same length or slightly longer, and of the same size basally but tapering somewhat less rapidly. P, and P; are the same length, 9 mm. long, about as stout as P, basally but tapering much more gradually and hence appearing larger, and somewhat stiffened. They consist of 24 segments which at first are broad, becoming about as long as broad at about the middle of the pinnule, and elongate distally; the proximal segments are strongly carinate. P, is of the same character as the pinnules preceding but is smaller, only 6 mm. in length. P, is slender, small, and weak, 3 mm. long. The following pinnules very slowly increase in length, reaching 5 mm. distally. The color in alcohol is dark brown. 324 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Of 16 young specimens from Singapore one has 17 arms 43 mm. long; two have 14 arms 55 mm. long; four have 13 arms 45 to 50 mm. long; five have 12 arms 45 to 65 mm. long; and four have 11 arms 50 to 60 mm. long. In some of these the distal edge of the radials is bordered with a row of small regular tubercles, and the lateral and proximal edges of the IBr,; may be more or less scalloped. Localities.—Gulf of Manaar; Ceylon Pearl Oyster Fisheries station LVII; outside Dutch Moderagam Paar; 21-66 meters; bottom orbitolites sand, nullipores, and dead corals [Chadwick, 1904; Herdman, 1906; A. H. Clark, 1912, 1918; H. L. Clark, 1915). Off Trincomalee (Trinquemale), Ceylon; Ceylon Pearl Oyster Fisheries station XXIV; 2.5 to 3 miles north of Foul Point; 44-84 meters; bottom hard and rough— probably rock [Chadwick, 1904; Herdman, 1906; A. H. Clark, 1912, 1918; H. L. Clark, 1915]. Bay of Bengal [Hartlaub, 1890, 1891; A. H. Clark, 1907, 1909, 1912, 1918). Singapore; Svend Gad [A. H. Clark, 1909, 1912, 1918] (21, U.S.N.M., E.1073, E.1079; C.M.). Singapore; 13 meters; December 12, 1898 [A. H. Clark, 1929] (2, B. M.). New Harbour, Singapore; 15 meters; July 31, 1899 [A. H. Clark, 1934] (1, Raffles Mus.). Geographical range.—From Ceylon and the Bay of Bengal eastward to Singapore. Bathymetrical range-—From the shoreline down to 44 (?84) meters. History.—This species was first described by Dr. Clemens Hartlaub from a specimen from the Bay of Bengal in the Géttingen Museum in 1890. It was rede- scribed and figured in 1891. Hartlaub said that it is perhaps most nearly related to H. quinduplicava (anceps), which also has only 10-14 arms. He distinguished bengal- ensis from quinduplicava by the following characters: In bengalensis there are only 22-24 cirrus segments as against 25-35 in quinduplicava; the outer cirrus segments are prominently spiny whereas in quinduplicava the outer cirrus segments have no dorsal spines; in guinduplicava more or less distinct synarthrial tubercles are present, whereas there are none in bengalensis; bengalensis is further distinguished by the flattened sides of the division series, and finally by the carination of the pinnule segments in the proximal portion of the arms. In 1904 Herbert C. Chadwick, under the name of Antedon anceps (=quindupli- cava), recorded specimens from Ceylon, where they had been found during the course of the Ceylon Pearl Oyster Fisheries investigations under the direction of the late Sir William A. Herdman In 1906 Professor Herdman published a note reidentifying the specimens recorded by Chadwick as Antedon bengalensis. In 1907 I referred bengalensis to the new genus Himerometra and in 1908 recorded a specimen from Albatross station 5146 under the name of Himerometra bengalensis. This specimen in reality represents Heterometra parilis. In 1909 I referred this species to the new genus Heterometra. In a memoir on the recent crinoids of Australia published in 1911, I recorded, under the name of Heterometra bengalensis, three small specimens from Port Curtis in the Australian Museum and a specimen from Holothuria bank in the British Museum. These four specimens are in reality Zygometra punctata. The same records were in- cluded in a memoir on the crinoids collected by the Hamburg Southwest Australian Expedition, also published in 1911. A MONOGRAPH OF THE EXISTING CRINOIDS 325 In my monograph on the crinoids of the Indian Ocean published in 1912 I recorded and gave notes on specimens identified as Heterometra bengalensis, which had been collected by the Royal Indian Marine Surveying steamer Investigator south of Ceylon and in the Andaman Islands. These specimens really represent H. reynaudi. In the previously known localities given for this species, Queensland and Holothuria Bank refer to Zygometra punctata, and Philippines refers to H. parilis. In an appendix to this monograph additional specimens of H. bengalensis were recorded from off Gopalpore in 25-28 and 30-38 fathoms, and from ‘‘?India.’”’ These specimens all represent H. reynaudi. In a supplement to my memoir on the crinoids of southwestern Australia, published in 1913, I transferred the Australian records for H. bengalensis to Zygometra punctata. In 1915 Dr. Hubert Lyman Clark listed this species from Ceylon, referring to my memoir on the crinoids of the Indian Ocean (1912). In the report on the unstalked crinoids of the Siboga expedition published in 1918, I inserted H. bengalensis in the key to the species of Heterometra and assigned the follow- ing characters to it—brachials longer [than those of H. aspera] with more oblique ends which overlap only very slightly, if at all; 11-15 arms. These characters were taken from small specimens of H. reynaudi which were misidentified as bengalensis. The range is said to be Ceylon and the eastern coast of India. Chadwick’s specimens are the only ones known from Ceylon; all the others recorded from Ceylon, and all those recorded from the eastern coast of India are in reality H. reynaudi. Heterometra aspera, described by me in 1909 from specimens collected at Singapore by the Danish consul at that port, Svend Gad, was admitted as a valid species in the memoirs on the crinoids of the Indian Ocean (1912) and on the Siboga collection (1918) and was again recorded from Singapore in 1929; it ssems not to differ in any tangible features from H. bengalensis. HETEROMETRA AFRICANA (A. H. Clark) Puate 26, Ficure 108; Puars 27, Figures 111-116 Antedon savignyi P. H. CARPENTER, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, pp. 252, 380 (Kurrachee). Amphimetra africana A. H. Cuarx, Proc. U. 8. Nat. Mus., vol. 40, 1911, p. 8 (southeastern Africa), p. 20 (description; Bagamoyo; also Zanzibar and Waxin); vol. 48, 1912, p. 381 (cotype in the U.S. N. M.), p. 384 (original reference), p. 393 (locality); Crinoids of the Indian Ocean, 1912, p. 111 (synonymy; localities); Smithsonian Misc. Coll., vol. 61, No. 15, 1918, p. 24 (published reference to specimens in the B. M.; Zanzibar; Waxin; notes)—Harrmerer, Mitt. Zool. Mus. Berlin, vol. 8, Heft 2, 1916, pp. 235, 237 (Bagamoyo, No. 4616 [type], 6380). Actinometra sp. Brit. Mus., MS. A. H. Cuarx, Proc. U.S. Nat. Mus., vol. 40, 1911, p. 21 (Waxin); Smithsonian Misc. Coll., vol. 61, No. 15, 19138, p. 24 (Waxin). Heterometra savignii A. H. Cuarx, Proc. U. S. Nat. Mus., vol. 40, 1911, p. 24 (Kurrachi); Crinoids of the Indian Ocean, 1912, p. 124 (Kurrachi; Straits of Ormuz; notes); Smithsonian Mise. Coll., vol. 61, No. 15, 1913, p. 27 (specimen from Kurrachi) ; Unstalked crinoids of the Siboga—Exped., 1918, p. 76 (in key; range; Persian Gulf and Kurrachi). Antedon ?milberti Brit. Mus., MS. A. H. Cuarx, Smithsonian Mise. Coll., vol. 61, No. 15, 1913, p. 24 (Zanzibar). Heterometra africana A. H. Cuarxk, Unstalked crinoids of the Siboga—Exped., 1918, p. 77 (in key; range).—Gist&Nn, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 22, 48 (extends to the east coast of Africa) —A. H. Cuarx, John Murray Exped. 1933-34, Sei. Re- ports, vol. 4, No. 4, 1936, p. 100 (range), p. 104.—Guis.in, Kungl. Svenska Vet.-Akad. Handl., ser. 3, vol. 17, No. 2, 1938, p. 12 (relation to var. delagoae). 326 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Diagnostic features.—There are no IIIBr series. The brachials are very short, and those in the proximal portion of the arms are unmodified. The cirri are 20-23 mm. long with usually 30 segments, of which the longest are about as long as broad and the outer are one-third again as broad as long; the tenth or eleventh and following segments bear long and prominent dorsal spines which begin abruptly. The 19 arms are about 125 mm. long. Description.—The centrodorsal is thin discoidal, broad, with the flat polar area 4mm. in diameter. The cirrus sockets are arranged in a single more or less irregular closely crowded marginal row. The cirri are XXIII, 29-32 (usually 30), 20-23 mm. in length. The first seg- ment is short and the following gradually increase in length to the seventh or eighth, which is about as long as broad. The following segments are at first similar, after the eleventh or twelfth very gradually decreasing in length so that the terminal seg- ments are about one-third again as broad as long. The tenth or eleventh (usually the latter) and following segments bear long and prominent dorsal spines which begin abruptly. The cirri do not taper distally. The radials are entirely concealed by the centrodorsal. The IBr, are very short, bandlike, in apposition laterally. The IBr, (axillaries) are short, two and one-half times as broad as long, with the lateral borders, which are not so long as those of the IBr,, in contact. The IIBr series are 4(3-++4). In the type specimen eight IIBr series are present, and the other arms are broken off before the first syzygy. The division series and the first brachials are in close lateral apposition and are sharply flattened against their neighbors. The lateral borders of the elements of the division series are slightly produced. The synarthrial tubercles are obsolete. The arms in the type specimen are 19 (720) in number, apparently about 125 mm. in length. The first brachials are short, slightly wedge-shaped, twice as broad as long exteriorly, interiorly united. The second brachials are slightly larger and more obliquely wedge-shaped. The first syzygial pair (composed of brachials 3-+4) is slightly longer interiorly than exteriorly and twice as broad as long interiorly. The next four or five brachials are oblong, three times as broad as long, those succeed- ing becoming obliquely wedge-shaped, twice as broad as long, then gradually less obliquely wedge-shaped, and after the end of the proximal half of the arms practically oblong and very short. P, is 12.5 mm. long, slender, flagellate distally, with 29 segments, which at first are twice as broad as long, becoming about as long as broad on the eighth and nearly twice as long as broad terminally. The second and following segments bear a sharp carinate ridge with the crest parallel to the axis of the pinnule, which gradually becomes less marked, disappearing on the tenth segment. P, is 15 mm. long, con- siderably stouter than P,; and not tapering so rapidly distally, with 25 segments, which at first are short, becoming about as long as broad on the eighth or tenth, and twice as long as broad terminally. From the fourth segment there runs outward on the outer side of the pinnule just distal to the mediodorsal line a prominent narrow ridge, which continues almost to the tip. There is a similar but less-marked ridge on P,. The second and following segments are rather sharply, but not highly, carinate. P; is similar to P,, of about the same size, or slightly shorter and smaller. P, is 7 mm. long, rather stout, though much smaller than Ps, tapering evenly to the tip, with A MONOGRAPH OF THE EXISTING CRINOIDS 327 15 segments, which become about as long as broad on the eighth. The second-eighth segments are rather strongly carinate. P, is similar but more slender, 6 mm. long, with only the second-sixth segments carinate. The following pinnules gradually lose the basal carination, become more slender, and increase in length, reaching 8 mm. distally. The color in alcohol is brownish purple. Notes.—The preceding description was drawn up from the type specimen from Bagamoyo. The other four specimens from Bagamoyo are very young, with the arms about 20 mm. long and the cirri about 6 mm. in length. Two have 11 and two have 10 arms. A specimen from Zanzibar has the cirri XVI, 25-27. It is smaller than the type specimen from Bagamoyo but is otherwise similar to it. A specimen from Waxin has 23 arms 120 mm. long. The IIBr series are 4(3+-4), and the I1IBr series are 2. The cirri are XX, 27; long sharp dorsal spines are developed from the tenth segment onward. When this specimen is compared directly with a specimen of H. savignii the cirri are seen to be more spiny, the spines first appearing nearer the base, and the proximal pinnules are seen to be larger. Of the seven specimens from Kurrachi in the British Museum, one has 21, two have 20, one has 19, two have 18, and one has 17 arms. The largest has the arms 130 mm. long and the cirri 25 mm. in length. The specimen with 21 arms has three IIBr 2 series; two of these are on the same postradial series, one being followed by a single internal IJIBr 2 series. One of the specimens with 18 arms has a single IIIBr 2 series. In the remaining specimens the IIBr series are invariably 4(3+-4). The specimen from Kurrachi collected by the Investigator is medium sized; there are 26 cirrus segments of which the distal bear strong dorsal spines. In the specimen from Kurrachi in the National Museum the cirri are XXIII, 28-30, about 20 mm. long. Dorsal spines are present from the tenth to thirteenth segment onward; they are rather smaller than usual. There are about 18 arms. In one of the specimens from the Straits of Ormuz the centrodorsal is thin- discoidal with a broad flat finely roughened dorsal pole, the edge of which is abruptly elevated into a smooth rounded arch at the base of each cirrus. The cirrus sockets are arranged in two closely crowded alternating rows. The cirri are XX, 28-34, from 20 to 25 mm. long, and rather slender. The longest proximal cirrus segments, the tenth-eleventh or -thirteenth, are from slightly longer than broad to slightly broader than long, and the distal segments are about half again as broad aslong. Dorsal spines begin from the twelfth to the fifteenth segment. They are high and conspicuous; the first is slightly smaller than the second, which is as large as those following. The nine IIBr series present are all 4(3+-4). The division series are broad, moderately rounded dorsally, in lateral contact with sharply flattened sides, which are produced into a narrow coarsely and irregularly tubercular or rugose border which is abruptly distinct from the dorsal surface. There are 19 arms about 110 mm. long. The brachials in the distal half of the arms are oblong and very short. P, is 7 mm. long, with 24 segments; these at first are twice as broad as long or even broader, gradually increasing in length and becoming about as long as broad on 328 BULLETIN 82, UNITED STATES NATIONAL MUSEUM the seventh, and somewhat longer than broad distally. Most of the segments are about as long as broad. The second-fourth segments bear a high carinate process with the finely spinous crest parallel to the axis of the segment. On the third segment following this crest rapidly decreases in height. P, is 7.5 mm. long, with 24 segments, which at first are much broader than long, becoming about as long as broad on the seventh, and nearly three times as long as broad terminally. The second-fourth seg- ments are carinate, as in P,, though not so strongly. P2 is 12 mm. long, with 23 seg- ments, much stouter than the pinnules preceding, evenly and gradually tapering to a delicate tip. The first two segments are somewhat more than half again as broad as long; those following gradually increase in length becoming about as long as broad on the fifth and about three times as long as broad terminally. The second-fourth seg- ments are rather narrowly carinate, the carinate process being highest on the second, less high on the third, and disappearing at the distal end of the fourth. P, is 6.5 mm. long, with 17 segments; it resembles P, but is proportionately smaller. P, is 5.3 mm. long, with 16 segments, and resembles P;. Ps, is 5 mm. long, with 16 segments, re- sembling P, but less slender terminally and with the carinate processes on the second and third segments much higher. The pinnules following resemble Ps. Another specimen has 18 arms 75 mm. long. A third has 13 arms 115 mm. long, IiBr 4(3+-4) series being developed on three postradial series, and the cirri XIX, 31-33, 20 mm. long. A fourth has 12 arms 80 mm. long, and the cirri XVI, 25-28, 15 mm. long; of the two I1Br series present, one is 4(3-+-4) and the other is 2. A fifth has 12 arms 75 mm. long, and the cirri XX, 33-34, 20 mm. long; both IIBr series are on the same postradial series, and both are 4(3-++4). The sixth specimen has 11 arms 65 mm. long; there is a single IJBr 4(3+-4) series; the cirri are XV, 26-32, 15 mm. long. The last specimen has 10 arms 90 mm. Jong, and the cirri XVI, 26+, 20 mm. long. Remarks.—This species seems to approach most nearly H. philiberti, but it is a smaller and less rugged form with fewer arms and much more spiny cirri. Localities —Bagamoyo (on the mainland of Africa opposite Zanzibar), Tangan- yika Territory [A. H. Clark, 1911, 1912; Hartmeyer, 1916] (5, U.S.N.M., 35218; Berl. M., 6380 [4616]). Zanzibar [A. H. Clark, 1911, 1912, 1913] (1, B.M.). Waxin, south of Mombasa [A. H. Clark, 1911, 1912, 1913] (1, B.M.). Kurrachi, Sind, northwestern India [P. H. Carpenter, 1888; A. H. Clark, 1911, 1912, 1913, all as savignii] (7, B.M.). Kurrachi [A. H. Clark, 1912, as savignii] (1, U.S. N. M., 35189 [original No. 14F.]). Investigator station 291; Straits of Ormuz, entrance to the Persian Gulf (lat. 26°22’ N., long. 56°10’ E.); 88-89 meters; muddy bottom; November 1, 1901 [A. H. Clark, 1912, as savignii] (7, U.S.N.M., 35180 [original No. ZEV 3095-7], 35761; 1.M.). Geographical range—From Bagamoyo and Zanzibar to Waxin and from Kur- rachi to the Persian Gulf. Bathymetrical range—From the shoreline down to 88 (?89) meters. Most of the records are from specimens collected along the shoreline. History.—This species was first mentioned in 1888 by P. H. Carpenter who listed it under the name of Antedon savignyi as occurring at Kurrachee (Kurrachi). A MONOGRAPH OF THE EXISTING CRINOIDS 329 It was first described by me under the name of Amphimetra africana in 1911 from a specimen from Bagamoyo in the Berlin Museum. At the same time two other specimens belonging to the British Museum, one from Zanzibar and one from Waxin (also spelled Wazin and Wasin), were recorded. In a memoir on the crinoids in the Berlin Museum published in 1912 the type specimen was mentioned, and four smaller specimens collected with it were recorded. In the memoir on the crinoids of the Indian Ocean published in 1912 I listed Amphimetra africana and gave as the habitat Bagamoyo, Zanzibar, and Wazin. Under Heterometra savignii I listed one specimen from Kurrachi and seven from the Straits of Ormuz and gave brief notes on them. Recent reexamination of these specimens has shown that they represent not savignii but the very different africana. In my notes on the recent crinoids in the British Museum published in 1913, I briefly described the specimens of Amphimetra africana from Zanzibar and Waxin. Under Heterometra savignii I recorded and gaye notes on seven specimens from Kurrachi. These were the specimens upon which Carpenter had based his record of Kurrachi for this species in the Challenger report. Of these specimens I said, ‘“No differences could be detected between these specimens and those from the Gulf of Suez.” The only specimen from Kurrachi at hand certainly is not savignii, as is shown by the strongly carinate lower pinnules. It appears to represent africana. The brief notes given apply to africana better than they do to savignii, although the chief differential features—the strongly carinate lower pinnules and very short distal brachials of africana—are not mentioned. There is, of course, the possibility that Carpenter and I were correct in our determinations and that both savignii and africana occur at Kurrachi, but this is quite improbable. In 1916 Dr. Robert Hartmeyer corrected the type numbers on the specimens of Amphimetra africana in the Berlin Museum. In my report upon the unstalked crinoids of the Siboga Expedition published in 1918 I transferred africana from Amphimetra, under which genus it had been placed on account of its very short oblong distal brachials, to Heterometra, placing it in a key to the species of the latter genus where the range was given as ““German Southeast Africa, Zanzibar, and British East Africa.’’ In the same key the range of H. savignii was given as “‘Red Sea to the Persian Gulf and Kurrachi.”’ The last two localities should have been placed under africana. HETEROMETRA SCHLEGELI (A. H. Clark) Puate 39, Figures 180, 181 [See also vol. 1, pt. 2, fig. 192 (lateral view), p. 115]. Alecto schlegelii Litkmn, MS. Himerometra schlegelit A. H. Cuark, Proc. Biol. Soc. Washington, vol. 21, 1908, p. 223 (description; Japan [the 10-armed specimen is Amphimetra laevipinna}). Amphimetra schegelii A. H. Cuarx, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 7 (listed). Antedon schlegelii (Liitken, MS.) A. H. Crarx, Vid. Medd. Naturh. Foren. Kgbenhavn, 1909, p. 117 (identity). 208244—40——-22 330 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Amphimetra schlegelit A. H. Cuarx, Vid. Medd. Naturh. Foren. Kgbenhavn, 1909, p. 158 (Japan {one of the specimens is Amphimetra laevipinna]); Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p. 14 (description of a specimen without locality; comparison with philibertz), p. 15 (original description based upon one specimen of schlegelii and one of Amphimetra laevipinna, but the former was selected as the type; northern representative of philiberti); Crinoids of the Indian Ocean, 1912, p. 111 (synonymy; Japan [records from New Guinea, Tonga and Fiji refer to Amphimetra papuensis}); Journ. Washington Acad. Sci., vol. 5, 1915, p. 214 (Malayan species; range and its significance).—GuisL6Nn, Kungl. Svenska Vet.-Akad. Handl., vol. 59, No. 4, 1919, p. 28 (discussion). Heterometra schlegelii A. H. CLarx, Unstalked crinoids of the Siboga-Exped., 1918, p. 77 (in key; range).—GuisL6n, Kungl. Fysiogr. Sillsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 22, 48.— Gistéin, Kungl. Svenska Vet.-Akad. Handl., ser. 3, vol. 17, No. 2, 1938, p. 12 (relation to H. africana var. delagoae). Amphimetra schleegelii Gistin, Kungl. Svenska Vet.-Akad. Handl., vol. 59, No. 4, 1919, p. 28 (discussion). Diagnostic features.—This is a small species without IIIBr series and with exceed- ingly short discoidal brachials in the outer half of the arms. The cirri are about 20 mm. long with usually 30-35 segments of which the longest are about as long as broad and the outer are about one-third again as broad as long; the twelfth and following bear long sharp dorsal spines. The 12-13 arms are 70-85 mm. long. Description —The centrodorsal is thick discoidal, with the polar area thickly covered with small low flattened tubercles. ‘The cirrus sockets are arranged in two closely crowded alternating rows. The cirri are XV, 26-35 (usually 30-35), about 20 mm. long. The first segment is short, the second is about twice as broad as long, the third is slightly longer, and the fourth is about as long as broad. The following to the twelfth or fourteenth are about as long as broad, some of the more proximal being occasionally slightly longer than broad, the length then very gradually diminishing so that the terminal segments are about one-third again as broad as long. From the twelfth segment onward com- paratively long sharp dorsal spines are developed. The opposing spine is considerably longer than the spines on the few preceding segments; it is about equal to the width of the penultimate segment in length, and is rather slender, abruptly curved basally but becoming nearly straight in its distal half. The radials project slightly beyond the rim of the centrodorsal. The IBr, are oblong, three or four times as broad as long, in close lateral apposition, and with the lateral edges swollen into a longitudinally elongate tubercle. The IIBr series are 4(3-+-4); there are three of these in the type specimen. The elements of the division series and the first two brachials are in close lateral apposition with their neighbors and are sharply flattened laterally. The 13 arms in the type specimen are about 70 mm. long. The first brachials are short, slightly longer exteriorly than interiorly, interiorly united for about two- thirds of their length. The second brachials are somewhat larger and irregularly quadrate, rising to a rather prominent tubercle with the first which resembles the tubercle on the articulation between the elements of the IBr series. The first syzygial pair (composed of brachials 3+4) is oblong, about as long as the second brachials, three times as broad as long. The following six brachials are oblong, about three times as broad as long, those succeeding becoming wedge-shaped or almost triangular, A MONOGRAPH OF THE EXISTING CRINOIDS 331 then gradually oblong again and decreasing in length so that the brachials in the distal half of the arm are extremely short and discoidal, with projecting distal borders. Syzygies occur between brachials 3-+-4 (in one case between brachials 4-++5), again from between brachials 11+-12 to between brachials 17+18, with occasionally an additional syzygy between brachials 9+ 10, and distally at intervals of 11-25 muscular articulations. P, is about 5 mm. long, comparatively slender and tapering evenly from the base to the delicate tip, with about 18 segments, all of which are about as long as broad. P, is about 7.5 mm. in length, slightly stouter than P, but becoming slender and delicate distally, with about 18 segments of which the first two are not quite so long as broad, the next two are about as long as broad, and the following gradually become elongated and about twice as long as broad distally. P, is about 8 mm. long, stouter and stiffer than P., though the component segments are in the same number and of the same proportions. P, resembles P; but is shorter and somewhat less stout with about 12 segments of which the first two or three are not quite so long as broad and those following are about as long as broad, becoming rather longer than broad distally. Distally the pinnules slowly increase in length and become more slender. The distal pinnules are 6 mm. long, moderately slender, with about 20 segments of which the first is not so long as broad, the second is about as long as broad, and those following become gradually longer, reaching a length of about twice the breadth distally. The color in alcohol is dull flesh color, with the perisome brown. Notes.—A specimen without locality in the Hamburg Museum may be described as follows: The cirri are XV, 29-35, from 20 to 23 mm. long. The longer proximal segments are about as long as, or slightly longer than, broad. The distal segments are nearly twice as broad as their median length. The twelfth, thirteenth, or four- teenth and following segments bear a long and prominent, rather broad, dorsal spine. The segments that bear the dorsal spines have the dorsal surface flattened so that there is a rather conspicuous rounded ridge along the boundary between the lateral and dorsal surfaces. The radials have a finely beaded distal edge as is frequently the case in H. philiberti. The 12 arms are 85 mm. long. One postradial series bears a single IIBr 4(3-++-4) series, which carries internally a IIBr 2 series. P, is small and slender, 6 mm. long, with 24 segments, which become about as long as broad on the fifth, the remainder being slightly longer than broad. The second-eighth bear a high broad finely spinous carinate process, which rapidly de- creases in height after the third. P, is 10 mm. long, with 25 segments, of which the fifth-seventh are about as long as broad and the distal are twice as long as broad. The second-ninth are diminishingly carinate, the outer border of the carination being convex on the second but parallel to the longitudinal axis of the segment on the remainder. The pinnule is evenly tapering, and is proportionately larger than P). P; is 13 mm. long, with 24 segments. It resembles P2, but the supplementary ridge is strongly indicated, and it becomes more distinctly prismatic distally. P, is 7 mm. long, with 18 segments, the second-fifth with carinate processes which are propor- tionately higher than those on the corresponding segments of the preceding pinnules. P; is 4.5 mm. long, with 14 segments, of which the second-fifth are very strongly and the sixth is somewhat carinate. The following pinnules resemble P;. The distal 332 BULLETIN 82, UNITED STATES NATIONAL MUSEUM pinnules are slender, 7 mm. long, with 19 segments of which the distal are about twice as long as broad. Localities—Japan [A. H. Clark, 1908, 1909, 1912, 1915, 1918; Gislén, 1919, 1924] (1, C. M.). No locality; C. Eberstein [A. H. Clark, 1912, 1915, 1918) (1, H. M.). History.—This species, which had long before received the manuscript name of Alecto schlegelii from Dr. Christian F. Liitken, was first described by me in 1908 under the name of Himerometra schlegelii. The description was based upon two specimens from Japan in the Copenhagen Museum that had been labeled Alecto schlegelii by Dr. Liitken. On the establishment of the genus Amphimetra in 1909 schlegelii was transferred to it. Later in the same year the two original specimens were again mentioned. At that time I wrote that this species is most closely related to Amphimetra ensifer and, like that form, has the disk completely covered with a pavement of small plates. It may be distinguished at a glance, however, by the longer and much more prominent dorsal spines on the cirrus segments, and by the comparatively small size of the synarthrial tubercles. In 1912 I deseribed another specimen, which I had examined at the Hamburg Museum in 1910. I wrote that this species is closely related to Amphimetra philiberti, but it may easily be distinguished by the longer proximal cirrus segments, the longer and more prominent spines on the distal cirrus segments, and by the smaller number of arms. It is a considerably smaller and more delicate species than A. philiberti. I said further that at the time I described schlegelii I had before me two specimens, one with 10 and the other with 13 arms, which were of the same size and general appearance. Such differences as I found between them I considered as probably the result of individual variation. At Hamburg there were also two specimens, one with 10 and the other with 12 arms, which resembled the pair in the Copenhagen Museum. The presence in the collection of the Hamburg Museum of the type specimen of laevipinna, and of six specimens of philiberti, a species I had not seen when I described schlegelii, showed clearly that the 10-armed specimen at Copenhagen and the one at Hamburg are specifically identical with Carpenter’s laevipinna, while the 12- and 13- armed specimens represent a very distinct form, easily differentiated by the strong carination of the pinnules, related to philiberti. I remarked further that the specimen I selected as the type of schlegelii when I wrote the original description happened to be the one with 13 arms. Thus while the specimen with 10 arms, at first considered as identical with it, must be referred to laevipinna, the name schlegelii is available for the 13-armed specimen, and for the 12-armed example at Hamburg, which represents the same species. In conclusion I said that schlegelii appears to be the northern representative of philiberti, just as laevipinna is the northern representative of the discoidea group of species—that is, of Amphimetra as now understood. In my memoir on the crinoids of the Indian Ocean published in 1912 I listed Am- phimetra schlegelii and gave as the range Japan and southward to New Guinea, Tonga, and Fiji. The last three localities refer to Amphimetra papuensis and are based on specimens from Port Moresby and Hood lagoon, New Guinea, in the Australian Museum, which were recorded in 1911 as Amphimetra discoidea, and others from A MONOGRAPH OF THE EXISTING CRINOIDS 333 Tonga and Fiji in the British Museum, which were recorded in 1913 as A. papuensis. In 1918 I transferred schlegelii from the genus Amphimetra to Heterometra, insert- ing it in a key to the species of the latter genus in the Siboga report. Dr. Torsten Gislén said in 1919 that Amphimetra schlegelii, though nearly related to A. milberti, is probably a good species because of the large number of pinnulars combined with small arms. He had evidently overlooked my resolution of the species as originally described into its two components in 1912. HETEROMETRA FLORA (A. H. Clark) Antedon laevissima Bri, in Gardiner, Fauna and Geography of the Maldive and Laccadive Archi- pelagoes, vol. 1, pt. 3, 1902, p. 224 (Mulan, Maldives). Amphimetra flora A. H. Cuark, Smithsonian Mise. Coll., vol. 61, No. 15, 1913, p. 23 (published reference to specimens in the B, M.; Mulan, Maldives; characters). Heterometra flora A. H. Cuark, Unstalked crinoids of the Siboga-Exped., 1918, p. 77 (in key; range).— GisLbN, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, p. 22—A. H. Cuiarx, John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1936, p. 99 (range), p. 103.—GisLin, Kungl. Svenska Vet.-Akad. Handl., ser. 3, vol. 17, No. 2, 1938, p. 12 (relation to. H. africana var. delagoae). Diagnostic features.—There are no I1IBr series. The brachials are very short and those in the proximal portion of the arms are unmodified. The cirri are stout and strongly curved, about 20 mm. in length, and are composed of 29-30 segments, all of which are much broader than long, the longest being about one-third again as broad as long. The 10-13 arms are 170-180 mm. long. Description.—The cirri are about XV, 29-30, stout and strongly curved, about 20 mm. in length. All the segments are subequal, all broader than long, the longest being about one-third again as broad as long. The tenth or eleventh and following bear prominent and sharp dorsal spines. The 10-13 arms are 170 to 180 mm. in length. The longest proximal pinnules are about 20 mm. long and rather slender. The distal ends of their component segments are perfectly smooth. The basal segments are strongly carinate. Notes.—In the type specimen the cirri are XV, 30. The 12 arms are 170 mm. long. Another specimen similar to the preceding has 10 arms 180 mm. long. , Ps, and P, are elongated, but those of the succeeding pinnules are shorter. Hartlaub said that corresponding interrelationships of the lower pinnules are shown by a much smaller specimen from Atjeh, Sumatra (=A. molleri), which through his intervention the Géttingen Museum received from the Leyden Museum. This specimen differs strikingly from that from the Sunda Straits in its light color. The color of the latter confirms Carpenter’s statement that the color (of milberti) is dark reddish brown bleaching to white. The color of the dorsal surface of the arms from the arm bases outward passes over from the darkest red-brown very gradually into white. Hartlaub said it was remarkable that Carpenter did not identify this specimen as milberti and did not include the locality Sunda Straits in the Challenger report. Hart- laub noted that in many places the axillaries and first brachials have a tendency toward wallsidedness. The photographs published by Hartlaub show that this specimen was a very typical example of this species. Of the 17 specimens from the Danish Expedition to the Kei Islands station 67, one has 11 arms, a single [Br 2 series being present. All the other specimens have 10 arms, which in the smallest are only 25 mm. in length. Of the two specimens recorded by Professor Koehler from Biliton the larger has XX cirri and the smaller XVII only. The dorsal spines are transverse. The larger specimen is very dark in color, purple violet. The smaller has the arms quite colorless and gray, the pinnules alone showing a purple-violet color. The largest specimen from Siboga station 33 has the arms 135 mm. long. The centrodorsal is thick discoidal, with the dorsal pole broad and convex, 7 mm. in A MONOGRAPH OF THE EXISTING CRINOIDS 365 diameter. The cirri are XXIV, 44-47 (usually 46-47), 45 to 50 mm. long. The cirri are rather stout basally but taper very appreciably distally. Rather prominent dorsal spines are developed from the thirteenth or fourteenth segment onward. The cirrus segments are subequal, the longest (in the proximal portion) being slightly broad- erthanlong. The ends of the segments in the earlier part of the cirri are slightly swol- len and prominent. The synarthrial tubercles are small but sharp and prominent. The disk is almost completely covered with a pavement of rather small plates. Another specimen has the arms 130 mm. long and the cirri XIII, 38-40, from 35 to 40 mm. long. The longest cirrus segments are half again as broad as long. Dor- sal spines begin on from the fifteenth to the seventeenth segments. A third specimen has the arms 115 mm. long and the cirri XIV, 34-41, about 30 mm. long. The longest cirrus segments are half again as broad as long. Dorsal spines begin on from the tenth to the thirteenth segments. On several of the cirri the two to eight spines following the first three or four have a longitudinally elongate chisellike crest, or are longitudinally paired. The color is whitish, becoming purplish toward the ends of the cirri and arms. The fourth specimen has the arms 115 mm. long and the cirri XVII, 41-43, from 35 mm. to 40 mm. long. The longest cirrus segments are half again as broad as long. Dorsal spines are developed from the twelfth segment onward. The color is purple. These specimens, like the type specimen from Ceram, have the cirri much less curved than the others I have examined. They were killed by immersion in fresh water, which may account for this condition. The longest proximal cirrus segments have the most produced distal ends. The specimen recorded by Prof. René Koehler from the Bay of Amboina was of very large size with the arms 200 mm. long. The cirri were XX XI, 40-42. The first brachials were broader and more swollen than usual. The largest of the three specimens collected by Professor Strubell at Amboina is, according to Dr. Reichensperger, perhaps the largest known specimen. The 10 arms are 300 mm. long. The cirri are about XXX, 28-43, from 35 to 40 mm. long, partly of uniform lateral width, and partly more slender proximally. From the fifteenth to the eighteenth segment onward the segments have prominent dorsal spines which are mostly large, more rarely less well developed. The first segment is short, up to twice as broad aslong. From thefourth or fifth onward the segments are more squarish, but soon again become broader, and distally are about half again as broad as long, not counting the spine. The last segment is about as long as broad, or longer than broad. The radials are almost entirely concealed. Very strong synarthrial tubercles are developed on the articulation between the first two brachials. P, is 13 mm. long, with 18-20 segments, of which the middle ones are somewhat longer than broad. P, is up to 16mm. long, with 22 segments, of which the middle ones, as in P,, are somewhat longer than broad. P; is 14 mm. long, with about 18 segments, which from the fourth onward are almost regularly longer than broad. There is some variation in the rela- tive size of the lower pinnules, and P; is not always the longest. The form of the brachials and to a lesser extent the relative development of the synarthrial tubercles also vary. The disk is 18 mm. in diameter and is very finely granulated. In the intermediate specimen the arms are 200 to 220 mm. long. The cirri are XLIII, 35-42, and resemble those of the preceding. The synarthrial tubercles are 366 BULLETIN 82, UNITED STATES NATIONAL MUSEUM somewhat less strongly developed. The pinnules are somewhat shorter, with fewer (14-18) segments. Reichensperger said that, except for the somewhat less robust cirri, this specimen might have served for the figure given by Carpenter in the Challen- ger report. In the smallest specimen the arms are 130 mm. long. The cirri are XVI, about 30. The synarthrial tubercles are less strongly developed than in that just preceding. P, is about 8 mm. long, with 16-18 segments. P, is 11 mm. long, with 18-20 segments. P, is equal to P,, or is slightly shorter. The brachials of this specimen are noticeable for their marked shortness. All three specimens have some of the lower pinnules with a trace of carination, usually only on the lowest segments. The color of the three specimens in alcohol is blackish brown. The specimen collected at Amboina by the Danish Expedition to the Kei Islands is large; the cirri have 36-39 segments. The specimen collected by the Willebrord Snellius at Amboina has 11 arms 220 mm. long. One IIBr 2 series is present. The cirri are 40 mm. long, with 43-47 segments. Miiller described Comatula jacquinoti as follows: There are 10 arms. The centro- dorsal is moderately convex and appears to be entirely covered with cirri. The cirri are XX, 35. Toward the end of the cirri, or much earlier, there is developed dorsally on the segments a spine that—certainly on the forward part of the segment—is directed distally. The cirrus segments are broader than long. The radials are very short. The brachials are short. The intersyzygial interval is 4-7 muscular articulations. P, to P, or P, are stouter than the other pinnules. The color is blackish brown. The expanse approaches 2 feet (which would give an arm length of about 300 mm.). Habitat Ceram. My notes on the type specimen, which I examined at the Paris Museum, are as follows: The cirri are large and stout, becoming laterally compressed distally. The first segment is very short, and those following progressively increase in length to the eighth or ninth, which is half as long as broad. All the cirrus segments are approxi- mately equal in size. Rather prominent dorsal spines are developed from the twelfth or thirteenth segments onward. There are 10 stout arms. The radials are concealed. The IBr, are very short and are almost entirely united with their neighbors. The IBr, (axillaries) are almost triangular, twice as broad aslong. They are not quite in lateral contact. The synarthrial tubercles are only slightly developed. The brachials are extremely short, with their distal ends slightly overlapping. The pinnules resemble those of the A. milberti described by P. H. Carpenter (=A. mollert). The color is brownish black. In the two specimens from Siboga station 164 the arms and cirri are slightly stouter than in the specimens from Siboga station 33, and the cirri are relatively shorter, with a less-marked production of the distal ends of the segments. In one the arms are about 140 mm. long, and the cirri are XVII, 34, 27 mm. long. The longest cirrus segments are from half again to twice as broad as long. Dorsal spines are developed from the tenth or eleventh segments onward, and the earlier spines are double as described for the specimens from station 33. The color is violet, with the cirri yellow, becoming violet distally in one specimen. A MONOGRAPH OF THE EXISTING CRINOIDS 367 These two specimens closely resemble others at hand from Western Australia. The cirri are proportionately shorter than in those from Siboga station 33, and the dorsal and ventral profiles of the individual segments are much less concaye, so that they appear much smoother, as in the Western Australian variety. The largest specimen from the Danish Expedition to the Kei Islands station 11 has 12 arms 205 mm. long. Two IIBr 2 series are developed, one on each of two adjacent rays. The cirri are XIX, 37-46, 40 mm. long. The first segment is very short and those following gradually increase in length, becoming about twice as broad as long on the sixth and half again as broad as long on the last six or eight. The last 23-28 segments have a sharp dorsal spine, which arises from the entire dorsal surface of the segment and is directed obliquely forward, becoming median and almost erect on the two segments preceding the penultimate. The two other specimens have the arms 150 mm. long. In one the cirri are XVIII, 29-33, 25 mm. long. The specimen from the Danish Expedition to the Kei Islands station 19bis has the arms 170mm. long. The cirri are XXV, 30-33, from 30 to 35mm. long. The dorsal pole of the centrodorsal is gently convex. Of the specimens from the Danish Expedition to the Kei Islands station 19, one has 12 arms, two IIBr 2 series being present. Another has 11 arms about 170 mm. long, a single IIBr 2 series being present; the cirri are XIV (with numerous others partially developed), 39-40, 35 mm. long. A third has 11 arms, with one IIBr 3 series; an extra division series of two ossicles is developed on a first brachial. All the other specimens have 10 arms. In three of them the arms are about 170 mm. long. Two others have the arms 80 mm. and 65 mm. long. One is very small, with the arms 25 mm. long. Dr. H. L. Clark said that this stiff and rather ungainly comatulid is by no means common at Broome, but he met with it several times while dredging, and one speci- men was found at extremely low water in September 1929. The individuals were all adult, with the 10 arms 110-130 mm. long. The cirri are XV-XXX, 28-36. He secured 10 specimens. As described by Dr. Torsten Gislén the specimen from Mjéberg’s station 6 has the cirri XXIX, 28-35. The peripheral cirri have the following numbers of segments: 35, with the dorsal spines beginning on the twelfth; 33, with spines from the twelfth; 33 (a young cirrus) ; 32, 30, with spines from the eleventh; and 30 (a very young cirrus 3 mm. long); an intermediate cirrus has 30 segments with spines beginning on the twelfth. The cirri of the apical row have the following numbers of segments: 30 (a young regenerating cirrus); 28, with spines from the nineteenth; 31, with spines from the twelfth; and 28 (a very young cirrus without dorsal spines). The 10 arms are 120 mm. long. P, is 7.5 mm. long, with 15 segments. P, is 9 mm. long, with 16 segments. P; is 9 mm. long, with 16 segments. P, is 7.5 mm. long, with 15 segments. P,,; is 7 mm. long. P, is 6 mm. long, with 14 segments. P, is 8 mm. long, with 13 segments. P, is 8 mm. long, with 15 segments. The distal pinnules are 9 mm. long, with 20 segments. This is a pale-colored specimen. wet The specimen from Mjéberg’s station 8 has the cirri XXIII, 26-35. The periph- eral cirri have the following numbers of segments: 35, with spines from the twenty- first; 31, with spines from the nineteenth; 31, with spines from the sixteenth; 29, with 368 BULLETIN 82, UNITED STATES NATIONAL MUSEUM spines from the eighteenth; 29, with spines from the seventeenth; and 27 (a very young regenerating cirrus). The cirri of the second (apical) row have the following numbers of segments: 26, with spines from the thirteenth; 27, with spines from the seventeenth; 27, with spines from the thirteenth; 27, with spines from the seventeenth; 28, with spines from the eleventh; 28, with spines from the thirteenth; and 29 with spines from the fourteenth. The 10 arms are 75 mm. long. The brachials are smooth, the proximal ones a little rugged. The first four brachials are a little flattened laterally. The IBr, are like the first brachials but are partly fused. P, is 5.5 mm. long, with 13 segments. P, is 7 mm. long, with 13 or 14 segments. P, is of about the same length, with 13 segments. P, is up to 6 mm. long, with 12-14 segments. P, is 4.5 mm. long, with 11 segments. P, is 6.5 mm. long, with 12 segments. P, is 8 mm. long, with 13 segments. The distal pinnules are 6.5 mm. long, with 16 or 17 segments. The segments of the pinnules are short. The basal segments are never longer than broad, and are coarse and thick; the distal segments are somewhat longer than broad and, especially in P,, are slender and thin. The specimen from Mjéberg’s station 9 was thus described by Gislén. The centrodorsal is thick discoidal, with the bare dorsal pole 3 mm. in diameter. The borders of the dorsal pole are somewhat swollen toward the cirrus bases. The cirri are XXITI-X XIX, 26-38, from 20 to 28 mm. long, and are arranged in two rows on the centrodorsal. The dorsal spines may begin anywhere from the ninth to the twenty- first segment. The dorsal spines are short, in height never reaching more than one- third the width of the segments, and arise from the distal portion of the segments. The opposing spine is twice as large as the preceding dorsal spines, is erect, and arises from almost the whole of the dorsal surface of the penultimate segment. The terminal claw is half again as long as the penultimate segment, usually rather slender with a very long point, and is strongly curved. The radials are smooth and very short, longest in the interradial angles. The IBr, are laterally united. The elements of the 1Br series bear a synarthrial tubercle, as in Heterometra crenulata, but arising more abruptly than in that species. The 10 arms are 140 mm. long and are basally closely appressed, though not at all or only very slightly flattened laterally. The first brachials are interiorly united, and there is a synarthrial tubercle on the articulation between the first two brachials. The brachials to about the tenth are irregularly discoidal with dorsolateral swellings, after the tenth becoming triangular. The arms are smooth until about the fiftieth or sixtieth brachial, after which the distal portion of the brachials becomes somewhat produced. The syzygies on two postradial series are as follows: on one arm between brachials 3+4, 10+11, 18+19, 25+26, 34+-35, 48+44, and 55+-56, and on the other arm between brachials 3+-4, 11+12, 18+19, 29+30, 42+43, and 57+-58; on one arm between brachials 3+4, 9+10, 21+-22, 32+38, and 49+50, and on the other arm between brachials 3+4, 9+10, 16+17, 29+30, 37+38, and 50+51. P, is 5.5-8 mm. long, with 13 (station 8) to 16 (station 11) segments, which are smooth and, except for the distalmost, a little longer than broad. The nine proximal segments are triangular in cross section. The outermost segments have the distal borders somewhat produced. Similar conditions are seen in P, and P; also. Ps is 7-10.5 mm. long, with 13 (station 8) to 19 (station 11) segments, of which the first is short and squarish and the third and fourth are the longest; the second-sixth seg- A MONOGRAPH OF THE EXISTING CRINOIDS 369 ments are triangular in cross section. Except as noted, P, resembles P;. P, is 7-9 mm. long, with 13-17 segments. P,) is 5.5 mm. long, with 16 segments. Pz» is 7.5 mm. long, with 18 segments. P,; is 9 mm. long, with 23 segments. The first two segments of the distal pinnules are broadened and swollen. From the ninth onward the segments are long and slender. P, is 7 mm. long, with 15 segments, corresponding to P;. P, is 9 mm. long, with 17 segments. The color in life was dark red, or bluish violet with the dorsal side of the arms and pinnules and in the outer half the ventral side also pale yellow, the yellow representing regenerated portions. Although it is not so stated, this description seems to have been based mainly on the specimen from station 9, a few notes being added from the specimens from the other stations. In the specimen from Mjéberg’s station 11 the cirri are XXIV, 27-38. The cirri in the peripheral row have the following numbers of segments: 38; 37, with spines from the sixteenth; 36, with spines from the fifteenth—an old cirrus; 32, with a spine only on the thirty-first—a young cirrus; 32 (a young cirrus) and 32. A very young intermediate cirrus has 28 segments. ‘The cirri in the lower or apical row have the following numbers of segments: 34, with spines from the twelfth; 31, with spines from the eleventh; 31, with spines from the eleventh; 28, with spines from the eleventh; and 27, with spines from the tenth. In the specimen presumably from the vicinity of Perth, Western Australia, the cirri have 42-47 segments and reach a length of 40 mm. The longest cirrus segments are slightly less than twice as broad aslong. The twentieth or twenty-first and follow- ing segments bear dorsal spines. The synarthrial tubercles are prominent and conical with the apex sharp, though they are not especially produced. The six specimens dredged by the Endeavour between Fremantle and Geraldton are all large and well developed, with the arms 185 to 195 mm. long. The centro- dorsal is 6 to 8 mm. in diameter and is very broad, with a flat or more or less convex dorsal pole, in the center of which there is sometimes to be seen a small pit. The cirri are XX—X XXII, the longest with 37-44 segments, and are 35 to 40 mm. in length. They are moderately stout and are composed of approximately subequal segments of which the longest (in the proximal portion) are half again to twice as broad as long, the distal being slightly shorter. Miiller described Comatula tessellata in the following terms: There are 10 arms. The cirri are XX—XV, 45; the cirrus segments are scarcely so long as broad, and the last 24 have small dorsal spines. The radials are very short. The intersyzygial interval is 8-11, rarely as many as 15, muscular articulations. The brachials are very short and are discoidal and imbricating, without a keel. Ps, P3, and perhaps P, are the largest pinnules. The skin of the disk bears small scattered calcareous plates. The color throughout is violet. The size is 1 to 1% feet (which would mean an arm length of 150-225 mm.). Habitat India. The type specimen was in the Bamberg Museum to which it had been given by Schénlein. Miller described Comatula (Alecto) milberti as follows: There are 10 arms. The centrodorsal is convex. The cirri are XXV—XXX, 35, the segments in the distal half with a dorsal spine in the middle, which is placed transversely. The radials are ex- tremely short. The brachials are short. The intersyzygial interval is 9-10 muscular 370 BULLETIN 82, UNITED STATES NATIONAL MUSEUM articulations. 2, Ps, and P, are the largest. The ventral surface of the disk is soft. The color is blackish brown. The expanse is nearly 2 feet (which would give an arm length of about 300 mm.). Habitat North America. The type specimen was an alcoholic example in the Paris Museum, which had been received from M. Milbert, of New York. Localities—Nha’trang Bay, Annam; littoral; Dr. C. Dawydoff [Gislén, 1936). Sunda Straits; Capt. G. W. Boot; L. Agassiz, 1859 [Hartlaub, 1912]. Danish Expedition to the Kei Islands; Dr. Th. Mortensen; station 67; Java Sea (lat. 5°48’ S., long. 106°12’ E.); 38 meters; sand and shells; July 27, 1922 (17). Biliton (north of Java), Sunda Islands; M. Korotnev, 1885 [Koehler, 1895; A. H. Clark, 1918]. Siboga station 33; Bay of Pidjot, Lombok; down to 22 meters; mud, coral, and coral sand; March 24-26, 1899 [A. H. Clark, 1918] (4, U.S. N. M., E. 479; Amsterdam Mus.). Bay of Amboina; MM. Bedot and Pictet [Koehler, 1895; A. H. Clark, 1912, 1918). Amboina; Professor Strubell [Reichensperger, 1913]. Danish Expedition to the Kei Islands; Dr. Th. Mortensen; Amboina; breakwater; about 1 meter; February 28, 1922 (1). Amboina; pier; 0-2 meters; Willebrord Snellius, May 6, 1930 (1, L. M.). Ceram; MM. Hombron and Jacquinot [J. Miiller, 1846, 1849; Dujardin and Hupé, 1862; P. H. Carpenter, 1879, 1883, 1888; Bell, 1882; A. H. Clark, 1911, 1912, 1918] (1, P. M.). Siboga station 164; east of Misool (lat. 1°42’30’’ S., long. 130°47’30’’ E.); 32 meters; sand, small stones, and shells; August 20, 1899 [A. H. Clark, 1918] (2, Amsterdam Mus.). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; station 11; 20 meters; sand; April 9, 1922 (3). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; station 19 bis; about 20 meters; sand; April 18, 1922 (1). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; station 19; 20 meters; sand; April 14, 1922 (10). Port Molle, Queensland; Alert [Bell, 1884; A. H. Clark, 1911, 1913] (1, B. M.). Broome, Western Australia; about 1 mile off jetty; H. L. Clark, August 27 and September 5, 1929 [H. L. Clark, 1938]. Broome; off Gantheaume Point; 4-7 meters; H. L. Clark, August 30, 1929 {H. L. Clark, 1938]. Broome; 9-15 meters; H. L. Clark, June 1932 [H. L. Clark, 1938). Mjoberg’s station 6; Cape Jaubert (south of Broome), Western Australia, 45 miles westsouthwest; 18 meters; July 8, 1911 [Gislén, 1919]. Mjéberg’s station 8; Cape Jaubert 45 miles westsouthwest; 18 meters; July 15, 1911 [Gislén, 1919]. Mjoberg’s station 9; Cape Jaubert 45 miles westsouthwest; 20 meters; July 15, 1911 [Gislén, 1919]. Mjoberg’s station 11; Cape Jaubert 45 miles westsouthwest; 22 meters; July 17, 1911 [Gislén, 1919]. A MONOGRAPH OF THE EXISTING CRINOIDS 371 ?Vicinity of Perth, Western Australia; Hamburg Southwest Australian Expedi- tion, 1905 [A. H. Clark, 1911, 1912; Hartmeyer, 1916] (1, H. M.). Endeavour; between Fremantle and Geraldton, Western Australia; 109-183 meters [A. H. Clark, 1914] (6, U.S. N. M., 35110; W. A. M.). Doubtful and erroneous localities —India [J. Miiller, 1841, 1843, 1849; Dujardin and Hupé, 1862; Wyville Thomson, 1865; P. H. Carpenter, 1879, 1888; A. H. Clark, 1907, 1912, 1918). North America (probably Western Australia) [J. Miiller, 1846, 1849; Dujardin and Hupé, 1862; Verrill, 1866, 1867; Pourtalés, 1869; P. H. Carpenter, 1879, 1883, 1888; Bell, 1882, 1884; A. H. Clark, 1907, 1911, 1912, 1918]. Geographical range-—From Annam, Misool, and Ceram southward to Port Molle, Queensland, and to between Fremantle and Geraldton, Western Australia, and westward to Billiton and the Sunda Straits. Bathymetrical range-—From the shoreline down to 109 (?183) meters; most of the records are from depths of not more than 20 meters. History.—There can be no doubt but that the Alecto tessellata described by Prof. Johannes Miiller in 1841 is the species under consideration, since the description fits this form perfectly and does not apply to any other. The type specimen of Alecto tessellata was presented to the Bamberg Museum by Johann Lucas Schénlein, who was a celebrated physician born in Bamberg. It is labeled as having come from India, which in those days was a very indefinite geographical term. Schénlein never visited India or the East Indies, and just how he acquired the specimen is not clear. The type specimen of Alecto tessellata has not been reexamined since it was originally described. In the Challenger report (1888) Dr. P. H. Carpenter especially mentioned that it was the only comatulid given in the list of species included in that report that he had not personally examined. In 1846 Miiller described Comatula (Alecto) milberti from a specimen presented to the Paris Museum by M. Milbert, of New York, and labeled as having come from North America. Jacques Gerard Milbert was a French painter and naturalist. In 1800 he accompanied, as chief artist, the French expedition in the corvettes Geographe and Naturaliste sent out under Capt. Nicolas Baudin to explore the Australian seas. Frangois Péron was also a member of this expedition. Captain Baudin died at Mauri- tius, and the Freycinet brothers succeeded him in command. In 1815 M. Milbert came to America, where he spent seven years in investigations in natural history, later returning to Paris. One of our common northern butterflies, Vanessa milberti, is named for him. It is most probable that Milbert collected the type specimen of Comatula (Alecto) milberti on the coast of Western Australia and brought it with him to New York. Therefore it is fairly safe to regard Western Australia as the type locality of this species. On the same page and immediately following the description of Comatula (Alecto) milberti, Miller described Comatula jacquinoti, which was based upon a specimen in the Paris Museum from Ceram labeled as having been collected by Hombron and Jacquinot, though Miller mentioned only Jacquinot. Honoré Jacquinot was a French physician and a brother of Admiral Charles Hector Jacquinot. In the capacity of ship’s surgeon he had taken part in Dumont-d’Urville’s expedition around the world in the Zélée in 1837-1840. 372 BULLETIN 82, UNITED STATES NATIONAL MUSEUM In 1857 Prof. Chr. F. Liitken mentioned Alecto milberti as an American species in his account of the echinoderms of Greenland. Dujardin and Hupé in 1862, under the headings Comatula tessellata, Comatula milberti, and Comatula jacquinoti, published translations of Miiller’s original descriptions. Wyville Thomson in 1865, under Antedon tessellatus, mentioned the plating of the disk as described by Miiller. In 1866 Prof. Addison E. Verrill said that he had not seen Antedon milberti, which had been described from ‘North America.’ Verrill again mentioned the species, as Antedon milbertii, in 1867, and in 1869 Count Pourtalés noted that it had been described from North America. In 1879 Dr. P. H. Carpenter referred milberti and jacquinoti to the genus Antedon as that genus was understood by him, but he was unable to place Comatula tessellata, as he knew of no description from which it is possible to obtain any information regard- ing the position of the mouth or the character of the oral pinnules. In October 1882 Prof. F. Jeffrey Bell published specific formulas for Antedon milberti and A. jacquinoti, and in April 1883 Dr. P. H. Carpenter published emended formulas for the same species. In the report upon the collections of H. M.S. Alert published in 1884 Professor Bell recorded Antedon milberti from Port Molle, Port Denison, the Prince of Wales Channel], and Torres Strait and said that the rich supply of this species in the Alert collection amply justified the doubts that P. H. Carpenter had expressed to him as to the exactness of the locality (North America) ascribed by Miiller to this species. He remarked also that this species was well represented in a collection of E. P. Ramsay’s, of the Australian Museum, Sydney. Only a single specimen from Port Molle, how- ever, represents the present species, all of the others being examples of the closely allied Amphimetra discoidea. In the Challenger report on the comatulids published in 1888 Carpenter made Antedon milberti the type of a special group that he called the Milberti group, to which he assigned a heterogeneous assemblage of 14 ten-armed species. In the key to these species he inserted milberti and tessellata; milberti was said to possess cirri with 25-40 segments, while fessellata has cirri with 45 segments and an intersyzygial interval of 8-11 muscular articulations. He remarked in a footnote that he had not seen the type specimen of tessellata. In the synonymy of Antedon milberti he placed jacquinoti and one of the two speci- mens upon which Grube’s Comatula laevissima was based (see page 353). The Antedon milberti of the Challenger report includes the following forms herein regarded as dis- tinct: Amphimetra tessellata (Ceram; Alert, Port Molle [part]); Amphimetra molleri (North Borneo; Challenger stations 203 and 212; Padan [=Padau] Bay); and Amphi- metra discoidea (Alert, Port Molle [part]; Port Denison; Prince of Wales Channel; Torres Strait). Carpenter said that under the name milberti he had united the two species that were found by Miiller in the Paris Museum with the manuscript names Comatula milberti and Comatula jacquinoti, which had been given them by Valenciennes. They are each based upon single specimens, which he had carefully examined in 1876 and again in 1880, and the subsequent study of a considerable amount of material ae A MONOGRAPH OF THE EXISTING CRINOIDS 373 obtained by the Challenger and the Alert, and also by Dr. John Anderson, of the Cal- cutta Museum, had convinced him that the two types are really identical. He pointed out that Miller hardly ever made any comparison of his species one with another but simply contented himself with descriptions, leaving his readers to determine the real points of difference between his various species. He showed that, as described by Miiller, the number of cirrus segments, the characters of the radials and of the brachi- als, the color, and even the size are the same in the two types; Comatula milberti has XXV-XXX cirri, with the spines transverse, while in C. jacquinoti there are XXII cirri, with the spines directed forward. In Comatula milberti the intersyzygial interval is 9-10 muscular articulations, and P,, Ps, and P, are the largest, while in C. jacquinoti the intersyzygial interval is 4-7 muscular articulations, and the first three or four pinnules are stouter than the others. Carpenter said that no one of these characters, or even the combinations of them, can be regarded as of specific value, especially when we remember that each of Miiller’s species was based upon a single specimen. He recalled that the type specimen of Comatula jacquinoti had been obtained at Ceram in 1841 by the expedition of d’Urville in the Zélée, while the form Miller described under the specific name of milberti had previously received the name from Valenciennes in honor of M. Milbert, of New York, who had given it to the Paris Museum, and it was possibly for this reason that the type was labeled as having come from North America. Carpenter said that under these circumstances Valenciennes, and after him Miller, were perhaps a little predisposed to regard it as distinct from the Comatula jacquinoti of Ceram, which Miiller described along with it in such nearly identical terms. Car- penter felt quite confident that Milbert’s specimen (see page 371) was not obtained any- where on the Atlantic coast of North America. He had seen nothing like it among the West Indian comatulids dredged by the Blake, while the only species of “‘Antedon”’ found on the Atlantic coast were Hathrometra tenella and perhaps Heliometra gla- cialis. All the characters of milberti, Carpenter pointed out, are those of the species of “Antedon’”’ that inhabit the eastern seas, where more or less similar individuals have been obtained at various localities from the Mergui Archipelago to eastern Australia, and he had little doubt that Milbert’s specimen had been brought to America from somewhere within this region. He noted that Verrill referred it to the Caribbean fauna, but with a query, while Dujardin and Hupé, who must have seen it for them- selves in the Paris Museum, refer to it as having come from North America. He said that we know nothing respecting any comatulids on the Pacific coast of Central and North America, and he strongly suspected that Milbert’s specimen must have been wrongly labeled. In 1891 Dr. Clemens Hartlaub added Carpenter’s Antedon laevipinna (= Amphi- metra laevipinna), which Carpenter had considered as valid in the Challenger report, to the synonymy of milberti, and said that Carpenter had already recognized the fact that the two were identical, as he had informed him by letter. In 1895 Prof. René Koehler recorded this species from Biliton in the Sunda Islands, and in another paper from the Bay of Amboina, giving notes on the specimens. In a first revision of the old genus Antedon published in 1907 I referred milberti and tessellata (accepting the interpretation given them in the Challenger report) to the new genus Himerometra. 374 BULLETIN 82, UNITED STATES NATIONAL MUSEUM In 1908, under the name of Himerometra milberti, I recorded a specimen from Albatross station 5100 in the Philippine Islands. This specimen has been responsible for much of the confusion in which the genus Amphimetra has been involved. It was identified on the basis of the description and figures given by Carpenter in the Chal- lenger report, and for a long time I regarded it as a typical example of milberti, basing all my ideas of that species upon it. It is described herein as the type specimen of Amphimetra spectabilis. In a revision of the family Himerometridae published in 1909 milberti and tessel- lata were transferred to the new genus Amphimetra. Later in 1909 I compared my new species molleri with milberti, my concept of the latter being based upon the specimen from Albatross station 5100. In 1910 I examined the type specimen of Comatula jacquinoti at the Paris Museum and in 1911 published a description of it, referring it to Amphimetra milberti. I failed to see the type specimen of milberti in the same museum and placed jacquinoti in the synonymy of milberti on the basis of the comparison made by Carpenter in the Challenger report. In a memoir on the crinoids collected by the Hamburg Southwest Australian Expedition, which was published in 1911, I gave notes under the name Amphimetra discoidea, on a specimen presumably from the vicinity of Perth. I said that in this specimen the cirri are stouter than usual, approaching the condition found in A. milberti (that is, the specimen of A. spectabilis from Albatross station 5100), although the individual should undoubtedly be referred to A. discoidea. All the additional locali- ties given refer to A. discoidea except Port Moresby and Hood Lagoon, New Guinea, which refer to A. papuensis. In a memoir on the recent crinoids of Australia, also published in 1911, I said that the Alert collected a specimen of Amphimetra milberti (=tessellata) at Port Molle, all the other specimens listed in the Alert report from Port Molle belonging to A. discoidea. I said that milberti ranged from the Mergui Archipelago to Borneo and the Philippine Islands and southward to northern Australia. The locality Mergui Archipelago was based upon Carpenter’s record (=A. molleri); Borneo was based upon Grube’s Comatula laevissima (=A. molleri); and Philippine Islands was based upon the specimen from Albatross station 5100 (=A. spectabilis). In a paper on the crinoids of the Berlin Museum, published in 1912, I again mentioned the specimen presumably from the vicinity of Perth under the name of Amphimetra discoidea. In my memoir on the crinoids of the Indian Ocean, published in 1912, I listed Amphimetra milberti (including jacquinoti), giving the synonymy and a list of the localities from which it is known. The species represented at the localities given are as follows: Amphimetra laevipinna (Canton); Amphimetra discoidea (Prince of Wales Channel, Torres Strait; Port Molle); Amphimetra tessellata (Amboina; Ceram; Port Molle); Amphimetra molleri (Panay and Zamboanga [Challenger stations 203, 212]; Padan {[=Padau] Bay, Mergui Archipelago; North Borneo; Ceylon); Amphimetra spectabilis (Philippine Islands). Alecto tessellata was given in the unidentifiable list, with a translation of the original description. In his memoir on the comatulids collected by the United States Coast Survey steamer Blake, published in 1912, Dr. Clemens Hartlaub described in considerable A MONOGRAPH OF THE EXISTING CRINOIDS 375 detail and figured a specimen, under the name Antedon milberti, from the Sunda Strait that had been collected by Capt. G. W. Boot and given to the Museum of Compara- tive Zoology by Prof. Louis Agassiz in 1859. In a paper on the crinoids of the British Museum published in 1913 I recorded one specimen of Amphimetra milberti from Port Molle, where it had been collected by the Alert, one from Challenger station 212, and one from Challenger station 203. The first represents the present species, but the last two represent A. molleri; they were referred to milberti because of their resemblance to the specimen from Albatross station 5100 (=A. spectabilis). Dr. August Reichensperger in 1913 recorded and described three specimens collected at Amboina by Professor Strubell and commented on the relation between this form and A. discoidea. Under the name Amphimetra discoidea, in 1914, I recorded and gave notes upon six specimens dredged by the Endeavour between Fremantle and Geraldton, Western Australia. I said that the specimens appeared undoubtedly to be exceptionally large and well-developed examples of discoidea but that typically discoidea has more slender and more tapering cirri in which the longest proximal segments are very nearly or quite as long as broad, but the distal are shorter, being broader than long. I remarked that in typical milberti the cirri are much stouter than they are in these specimens, the segments all being approximately of equal length, though the outer are a trifle longer proportionately, about four times as broad as long. The specimen assumed to represent milberti upon which the preceding comparison was based was the specimen from Albatross station 5100 subsequently described as the type of spectabilis. Dr. Robert Hartmeyer in 1916 published a note saying that the specimen pre- sumably from Perth is in the Hamburg and not in the Berlin Museum. In the report upon the unstalked crinoids of the Siboga expedition published in 1918, I attempted a complete revision of the species of this genus. The specimen from Albatross station 5100 was finally recognized as a distinct form and called spectabilis, although it was not formally described. The present species, under the name of Amphimetra jacquinoti, was recorded from Siboga stations 33 and 164, and notes on the specimens were given. The specimens previously called discoidea from ?the vicinity of Perth and from between Fremantle and Geraldton were transferred to this species. Amphimetra milberti and A. tessellata were included in a list of doubtful species of Amphimetra on the basis of their type specimens only, and the former was discussed at considerable length. In 1919 Dr. Torsten Gislén recorded and gave notes upon four specimens collected by Dr. Eric Mjéberg on the northwest coast of Australia and also discussed the relationships of the species at some length. In 1921 Dr. Hubert Lyman Clark discussed the range of this form on the coasts of Australia, and in 1924 Dr. Gislén described various structural features on the basis of the material from Mjéberg’s collection. In 1936 Dr. Gislén recorded and gave notes on a specimen from Annam. Dr. H. L. Clark in 1938 recorded and gaves notes on 10 specimens collected by himself in 1929 and 1932 in the vicinity of Broome, Western Australia. 376 BULLETIN 82, UNITED STATES NATIONAL MUSEUM AMPHIMETRA TESSELLATA DISCOIDEA (A. H. Clark) Puate 39, Ficures 178, 182; Pirate 40, Ficure 188 [See also vol. 1, pt. 1, figs. 183, 184 (centrodorsal), p. 235; fig. 336 (cirrus), p. 285; pt. 2, fig. 259 (arm), p. 205; fig. 309 (proximal pinnules), p. 223; fig. 349 (distal pinnules), p. 229; figs. 463, 464 (pin- nule tip), p. 266; fig. 791 (ambulacral deposits), p. 372.] Comatula carinala Goutrin-M6NEvVILLE, Iconographie du régne animal de G. Cuvier, 1828-1837, pl. 1, figs. 2, 2a.—Gnrirrirn, Cuvier’s Animal Kingdom, XII, Mollusca and Radiata, 1834, pl. 8, fig. 2. Comatula dibrachiata Dusarpin and Hup#, Histoire naturelle des zoophytes, Echinodérmes, 1862, p. 208 (nomen nudum).—A. H. Crark, Bull. Mus. Hist. Nat., Paris, 1911, p. 245 (=A. discoidea). Antedon milberti P. H. Carpenter, Trans. Linn. Soc. (Zool.), ser. 2, vol. 2, 1879, p. 29 (in part).— Bett, Proc. Zool. Soe. London, 1882, pp. 533, 534 (in part)—P. H. Carrrnver, Proc. Zool. Soc. London, 1882 (1883), p. 746 (in part) Bru, Rep. Zool. Coll. H. M.S. Alert, 1884, p. 155 (specific formula), p. 156 (Port Molle; Port Denison; Prince of Wales Channel; Torres Strait) ; Proce. Linn. Soc. New South Wales, vol. 9, 1884 (1885), p. 497 (Port Denison; Port Molle).— P. H. Carpenter, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 194 (specimens from the Australian localities given; Challenger specimens are mollert).—BE.1, Proc. Zool. Soc. London, 1894, p. 394 (northwestern Australia, 8-15 fms.)—Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1580 (listed; in part). Himerometra discoidea A. H. Cuarx, Proc. Biol. Soc. Washington, vol. 21, 1908, p. 225 (nomen nudum; compared with H. ensifer); Smithsonian Misc. Coll., vol. 52, 1908, p. 215 (description; Port Denison, Queensland; also Albatross Sta. 5138). Amphimetra discoidea A. H. Cuark, Proc. U. S. Nat. Mus., vol. 37, 1909, p. 32 (arms compared with those of A. parilis); Vid. Medd. Naturh. Forening Kgbenhayn, 1909, p. 158 (compared with A. formosa); Proc. U.S. Nat. Mus., vol. 40, 1911, p. 2 (possibly the species represented by Guérin-Méneville’s fig. 2a of Comatula carinata, which was possibly drawn from a specimen in the Paris Museum labeled Antedon milberti var. dibrachiata); Notes Leyden Mus., vol. 33, 1911, p. 182 (synarthrial tubercles compared with those of Amphimetra pinniformis); Bull. Mus. Hist. Nat., Paris, 1911, p. 245 (=Comatula dibrachiata Dujardin and Hupé), p. 250 (Australia; description of the specimen; possibly served Guérin-Méneville for his figure of Comatula cari- nata); Die Fauna Siidwest-Australiens, vol. 3, Lief. 13, 1911, p. 436 (northwest Australia, 8-15 fathoms); p. 440 (East Indian species occurring south to Perth and Port Molle), p. 443 (range on the east coast), p. 444 (range on the west coast), p. 459 (localities [New Guinea records are papuensis)), p. 466 (association with other species); Mem. Australian Mus., vol. 4, 1911, p. 717 (known to P. H. Carpenter from Australia), p. 721 (occurs south to Port Molle), p. 723 (north- west Australia), p. 734 (in key), p. 766 (synonymy; characters; Port Denison; Port Molle; also Singapore and Philippines; summary of previous records; range; specimen labeled dibrachiata in the P. M.); Crinoids of the Indian Ocean, 1912, p. 31 (=Antedon milberti, part, Bell, 1884), p. 34 (=Antedon milberti, part, P. H. Carpenter, 1888), p. 38 (=Antedon milberti, part, Bell, 1894), p. 112 (synonymy; range); Smjthsonian Misc. Coll., vol. 61, No. 15, 1913, p. 24 (pub- lished references to specimens in the B. M.; localities of specimens in the B. M.).—RerrcHen- SPERGER, Abh. Senck. naturf. Ges., vol. 35, Heft 1, 1913, p. 82 (Aru Islands), p. 93 (details of localities; abnormal specimen; comparisons), fig. 7, p. 95.—A. H. Ciark, Internat. Revue gesamt. Hydrobiol. und Hydrogr., 1915, pp. 223 ff. (part; detailed account of the distribution in Aus- tralia; Unstalked crinoids of the Siboga-Exped., 1918, p. 83 (in key range), p. 87 (references; notes; stations 99, 273; additional localities), pp. 272, 275 (listed), pl. 6 (colored), pl. 18, fig. 35.— GisL6£Nn, Kungl. Svenska Vet.-Akad. Handl., vol. 59, No. 4, 1919, p. 4 (synonym of milberti).— A. H. Crark, Smithsonian Mise. Coll., vol. 72, No. 7, 1921, pl. 2, fig. 22 (ambulacral deposits).— H. L. Crank, Echinoderm fauna of Torres Strait, 1921, p. 6 (history), p. 21 (range).—GisL6Nn Zool. Bidrag Uppsala, vol. 9, 1924, p. 11( abnormal pinnulation; from Reichensperger).—A. H. Crark, Proc. Biol. Soc. Washington, vol. 47, 1934, p. 11 (Singapore). -(mphimetra formosa A. H. Cuark, Proc. U. 8S. Nat. Mus., vol. 37, 1909, p. 32 (nomen nudum; arms compared with those of A. parilis); Vid. Medd. Naturh. Foren. Kgbenhavn, 1909, p. 156 (compared with A. mélleri), p. 157 (Singapore; short description), p. 193 (collected at Singapore A MONOGRAPH OF THE EXISTING CRINOIDS Shai by Svend Gad); Mem. Australian Mus., vol. 4, 1911, p. 766 (synonym of discoidea); Crinoids of the Indian Ocean, 1912, p. 112 (same).—ReIcHENSPERGER, Abh. Senck. naturf. Ges., vol. 35, Heft 1, 1913, p. 94 (characters discussed). Antedon milberti var. dibrachiata A. H. Ciarx, Proc. U. S. Nat. Mus., vol. 40, 1911, p. 2 (specimen so labeled in the Paris Mus. possibly the original of Guérin-Méneville’s figure). Comatula (Antedon) milberti var. dibrachiata A. H. Cuarx, Bull. Mus. Hist. Nat., Paris., 1911, p. 251 (MS. name by P. H. Carpenter found with a specimen of A. discoidea from Australia); Mem. Australian Mus., vol. 4, 1911, p. 767 (comparison with type of discoidea; appears to be the origi- nal of Guérin-Méneville’s figure). Actinometra brachiolata (B. M., MS.) A. H. Ciarx, Smithsonian Misc. Coll., vol. 61, No. 15, 1913, p. 75 (name found with a specimen of this species in the B. M.). Diagnostic features—A medium-sized or rather large form with the cirri mod- erately stout and straight proximally, becoming more slender and curved distally; the cirri have 34-51 segments of which the longest are about as long as broad and the distal bear short dorsal spines or pointed tubercles; the arms are 110-200 mm. long; prominent sharp-pointed synarthrial tubercles are present. Description.—The centrodorsal is large, hemispherical or somewhat columnar, with a large convex polar area. The cirrus sockets are arranged in two crowded alternating marginal rows. The cirri are XVII, 37-45 (usually about 40), 30 mm. long, decreasing very gradually in thickness for the first 8 or 10 segments, then remaining uniform. The first segment is short, about twice as broad as long or rather shorter, and those follow- ing gradually increase in length, becoming about as long as broad after the twelfth or sixteenth, and about cne-third broader than long in the terminal portion of the cirri. From about the eighteenth segment onward prominent, though small, dorsal spines are developed, which are subterminal in position, becoming terminal on the last two or three segments. The opposing spine is small, median in position, not rising to more than one-third the width of the penultimate segment in height. The terminal claw is somewhat longer than the penultimate segment and is moderately curved. The radials project slightly beyond the rim of the centrodorsal; their dorsal surface is parallel to the dorsoventral axis of the animal. The IBr, are short, oblong, about three times as broad as long. The IBr, (axillaries) are rhombic, about twice as broad as long, proximally rising to a low conical tubercle with the IBr,. The 10 arms are about 130 mm. in length. The first brachials are wedge-shaped, about two and one-half times as broad as the exterior length, almost entirely united interiorly. The second brachials are irregularly quadrate, rather larger than the first. The first syzygial pair (composed of brachials 3-+-4) is oblong, two and one-half times as broad as long. The following brachials to the tenth are slightly wedge-shaped, about three times as broad as long, and those succeeding become more obliquely wedge-shaped, somewhat over twice as broad as their greatest length, then gradually shorter and less and less obliquely wedge-shaped, and very short and discoidal after about the proximal third of the arm. P; is small and comparatively slender, 7 mm. long, with 15 segments, all of which are somewhat longer than broad, the first two and the terminal three or four being not quite so long as the others. PP, is 11 mm. long, stouter than P,, with 17 segments, of which the first two are about as long as broad and the remainder are slightly longer 2082444095 378 BULLETIN 82, UNITED STATES NATIONAL MUSEUM than broad. P, is 10 mm. long, resembling P2, with 15 segments. P, and the follow- ing pinnules are 7 mm. long, with 14 segments, P, and those succeeding being about as stout as P;. The distal pinnules are 10 mm. long. Notes.—The specimen from Siboga station 99 is typical, and resembles others at hand from Australia. The color is yellow brown with the cirri and pinnules purple. The specimen collected by Dr. Mortensen off Jolo has the arms about 150 mm. long. “The specimen from Singapore is slightly smaller than the specimens from the Philippine Islands but agrees perfectly with them. The 10 arms are 110 mm. long, and the cirri are XV, 31-34, from 20 to 25 mm. long. The color is a beautiful deep violet, more or less blotched dorsally with pinkish flesh color. The specimens collected by the Albatross in the Philippimes and the one from Singapore I originally considered as representing a distinct species, which I called Amphimetra formosa. This species was supposed to differ from A. discoidea from Port Denison, Queensland, in having the lower pinnules with shorter segments of which the basal are usually slightly carinate, and the distal have slightly thickened edges vaguely suggesting an approach to the conditions found in Heterometra crenu- lata; the synarthrial tubercles in formosa were said to arise more abruptly than they do in discoidea, and to be somewhat higher and more prominent. The study of additional material showed that these supposed differences do not hold. The three specimens collected by the Danish Expedition to the Kei Islands at Amboina are all large. One has the arms 200 mm. long and the cirri XV, 39-42, about 45 mm. long. Another has the arms 195 mm. long and the cirri XVII, 37-39, 40 mm. long. The third specimen is similar to the other two. Reichensperger recorded seven specimens from Dr. Merton’s station 16 in the Aru Islands. The centrodorsal varies from flat to hemispherical; if it is flat there is sometimes a small tubercle in the center. The cirri are XV—XXVIII (the last in the largest specimen), 32-42, from 30 to 35 mm. long. From the fifteenth-eighteenth segment onward a prominent dorsal spine is developed. The earliest segments are very short and broad, but from the third onward the relative length increases up to the fifteenth-eighteenth segments, which are almost squarish, though always some- what broader than long. The relative length of the segments then decreases, and usually toward the end of the cirri again increases so that the antepenultimate segment is approximately squarish while the last is visibly longer than broad. The radials are slightly visible. The IBr, are at least three times as broad as long. The IBr, (axillaries) are at the most twice as broad as long, rhombic, rising on the articulation with the IBr, to a very stout sharp tubercle. An almost equally stout tubercle is found on the articulation between the first two brachials. There are in all cases 10 arms, which are 120 to 150 mm. long. The form of the brachials is as described and figured by Carpenter in the Challenger report (=A. molleri). P, is about 8-9 mm. long and is composed of 15 or 16 segments of which the two first are almost squarish and the remainder are longer than broad. P, is about 11-13 mm. long, with 17-20 segments, and is somewhat stouter than P, and P3. P; is 9-10 mm. long, with 16 segments. P, and P; have the two lowest segments almost squarish as in P,, and the remainder longer than broad. The pinnules following decrease slowly in length and in the number of their segments (14-10) and again increase in the distal portion of A MONOGRAPH OF THE EXISTING CRINOIDS 379 the arms. The lower pinnules are to a greater or lesser extent keeled. The diameter of the disk is 10-12 mm. When dried the disk is usually prominently and often rather coarsely granulated. The anal tube is less conspicuously granulated. The color in life of the dorsal side was white. The cirri more or less extensively, and the ambulacral grooves, were violet, this color sometimes extending onto the sides of individual arms. In alcohol the color is grayish white and dark violet. One of the seven specimens from this station is very small with the arms only about 35 mm. long. It was found attached to one of the larger individuals. The cirri are XVII, 17-20, up to 10 mm. long. The fourth segment is already approxi- mately squarish, and those following are almost without exception markedly longer than broad. P, has 10 and P, has 12 or 13 segments. The relative length of the individual segments is not quite the same as it is in the pinnules of the larger speci- mens. The synarthrial tubercles are much less developed than in the latter. But there can be no doubt that all seven specimens represent the same species. A specimen from Merton’s station 10 has the arms 120 mm. long, but the cirri relatively short, not over 20 mm. long, with about 35 segments. The synarthrial tubercles are moderately developed. The pinnules are relatively stout. P, is com- posed of about 14 segments and is up to 8 mm. in length. P, is composed of 16 segments and is 10mm. long. P; has 16 segments and is 8mm. long. In none of the other specimens of this species did Reichensperger find such varying proportions in the individual cirrus segments. For instance, there are two P,; in which both of the first two segments are much longer than broad, while the third segment is squarish. On the other hand, there are P; with squarish basal segments, the second segment somewhat longer than broad, the third much longer, the fourth abruptly much broader than long, and the fifth and following again markedly longer than broad. There is no trace of regeneration in these pinnules. The color in life of this specimen, as also of two others from the same station, was white, with the ambulacral grooves and the whole of the cirri violet. A specimen from between Batu Kapal and Meriri was in life colored as the preceding. The arms are 70 mm. long. Synarthrial tubercles are only weakly developed. The cirri are only XII, arranged in a single row. Most of the lower pinnules and the anal tube are in process of regeneration. The disk is very promi- nently and coarsely granulated. A large specimen from Merton’s station 12 has the arms about 140 mm. long, and the cirri up to 35 mm. long with 35-43 segments, which in the relation of length to breadth are almost half way between discoidea and milberti (=tessellata); they can in no way be described as “‘slender.’’ In the relation to body size and to arm length they are at least as stout as the cirri of similar specimens of true milberti (=tessellata). P, is relatively shorter than in the other specimens, and P, is not over 9-10 mm. in length. The color is the same as in the others. A typical example from Siboga station 273 has the arms 115 mm. long and the curi XV, 32-34, 22 mm. long. The longest cirrus segments are slightly broader than long. The disk is almost completely covered with plates. From this station there are 12 other similar specimens, and two small and immature examples. These speci- mens exactly resemble others at hand from Queensland and from northwestern Australia. The cirri are comparatively slender and taper distally. 380 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Of the eight specimens from Torres Strait in the British Museum seven are brownish yellow, with the cirri deep purple narrowly banded with white at the articulations, and the last is entirely purple. Of the 12 specimens from Port Denison in the collection of the Australian Museum one has the cirri XXII, 46, 30 mm. long; in another the cirri are XX, 36-49, 30 mm. long; in a third they are XXIII, 43, 40 mm. long; and in a fourth they are XXII, 42-45, 30 mm. long. These and five others closely resemble the type specimen, which also came from Port Denison, and six additional specimens, three of which are large and beautiful examples of the species, in the British Museum. They show no tend- ency toward the curiously abrupt type of synarthrial tubercle or toward the short- segmented proximal pinnules characteristic of the form from the Philippine Islands and Singapore that I have called formosa. Three other large and well-developed specimens from Port Denison in the Australian Museum have the cirri about XX, 46-51, from 30 to 37 mm. long. These three all show the abrupt synarthrial tubercles and the short-segmented proximal pinnules of formosa in more or less perfected form. One of the Alert specimens from Port Molle in the British Museum is a fine example of the species. The two specimens from Port Molle in the Australian Museum have the cirri about XX, 39-42, 30 mm. long. Both of these specimens have syn- arthrial tubercles resembling those in the type specimen of formosa. The two specimens from northwestern Australia in the British Museum are slaty gray, purplish ventrally. The three specimens from Western Australia are typical. The specimen in the Paris Museum labeled Comatula (Antedon) milberti var. dibrachiata exactly resembles the type specimen in the United States National Museum. The cirri have 26 or 27 segments. The specimen from the Danish Expedition to the Kei Islands station 101 has the arms about 100 mm. long. It is flesh colored, with the pinnules and ventral surface deep purple. The seven small specimens from the Danish Expedition to the Kei Islands station 75 appear to be young examples of this form. The dorsal spines on the cirri of some individuals begin as early as the fourth or fifth segment. Reichensperger said that in discoidea the carination on the earlier segments of the lower pinnules is in general more marked than it is in tessellata, although there is individual variation in this feature. The general appearance of the pinnules in cross section is thus somewhat more slender and more rounded in tessellata than it is in discoidea. Abnormal specimens.—A 4-rayed individual was secured by the Siboga at station 273. The missing ray appears to be the anterior. Except for the absence of one of the rays this example seems to be quite similar to the others from this station. In a medium-sized specimen from the Aru Islands described by Reichensperger two pinnules instead of a new arm have regenerated on a stump formed by the breaking off of an arm between the fifth and sixth brachials. Reichensperger regards this as proof of equivalent developmental potentiality in arm and pinnule buds. He noted that this peculiarity has already been noticed in Comatella nigra, which is a slip for C. stelligera (see Part 3, p. 106). A MONOGRAPH OF THE EXISTING CRINOIDS 381 Remarks.—Whether discoidea should be recognized as a species distinct from tessellata, or as a geographical race, or perhaps merely as a form, is at the present time simply a matter of personal opinion. Although when they are typically developed tessellata and discoidea are very different in appearance, they certainly intergrade. The ranges of the two types overlap, though generally speaking tessellata is character- istic of the region from the Sunda Islands to the Moluccas and southward along the coast of Western Australia, while discoidea is equally characteristic of the region from the Philippines to Singapore and eastern and northern Australia as far as the Aru Islands. Dr. August Reichensperger said in 1913 that the similarities between typical tessellata and the Australian discoidea are so great, while on the other hand the differ- ences, which have to do only with the relative width of the cirrus segments and a few cirrus segments, are so slight and so variable, that one can scarcely regard the two forms as species; at most they can be considered as local varieties. He said that he kept discoidea distinct from tessellata only to avoid increasing the present (1913) confusion in the taxonomy of the comatulids. He remarked that the small specimen from Merton’s station 16 seemed to him to cast doubt on the possibility of establishing new species on the basis of differential characters furnished by the cirri. His view is that as the cirrus segments change the relation of length to breadth during growth the cirri afford specific characters of only very limited value. He also noted that the number of pinnule segments is not fixed, and synarthrial tubercles are much less strongly developed in small than in large individuals. He believed that, in view of the detailed descriptions given of their specimens by Carpenter and by Hartlaub, supplemented by his study of the large specimen of tessellata from Amboina, the most reasonable assumption is that the various so-called species really represent a kind of developmental series. In the present state of our knowledge the identification of comatulids in very many cases is practicable only if adults are available for study. This is particularly true in the case of species with more than 10 arms. But the fact that a small individual of discoidea differs widely in cirrus and pin- nule characters and in the relative development of the synarthrial tubercles from an adult is no indication that it would not be distinguishable from a small individual of tessellata in the same stage of development. In many different animal types growth changes are as far reaching as they are in the crinoids, and in not a few groups that are much better known than the comatulids accurate identifications are possible only if adults are available. Dr. Torsten Gislén (1919) entirely agreed with Dr. Reichensperger that it is incorrect to maintain discoidea and tessellata as distinct species on the ground of the somewhat different cirrus segments. Localities —Takao, Formosa (Taiwan); Dr. Fred. Baker, December 3-4, 1914 [A. H. Clark, 1912, 1918] (1, U.'S.N.M., 34497). Port Galera, Mindoro, Philippines; Dr. Lawrence E. Griffin (2, M.C.Z., 680). Albatross station 5432; in the vicinity of eastern Palawan; Corandagos Island (N. W.) bearing N. 30° E., 5. 7 miles distant (lat. 10°37’50’’ N., long. 120°12’00”’ E.); 93 meters; sand; April 8, 1909 (1, U.S.N.M., 36010). 382 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Albatross station 5131; Sulu (Jolo) Sea, off western Mindanao; island off Pana- butan Point bearing N. 20° E., 0.4 mile distant; 49 meters; green mud and coral sand; February 6, 1908 (1, U.S.N.M., 36029). Albatross station 5132; Sulu Sea, off western Mindanao; island off Panabutan Point bearing N. 15° W., 0.3 mile distant; 47 meters; green mud and sand; February 6, 1908 (1, U.S.N.M., 35246). Albatross station 5138; in the vicinty of Jolo (Sulu); Jolo light bearing S. 19° E., 2.5 miles distant (lat. 6°06’00’’ N., long. 120°58’50’’ E.); 35 meters; sand and coral; February 14, 1908 [A. H. Clark, 1908, 1909, 1911, 1918] (1, U.S.N.M., 35201). Philippine Islands [A. H. Clark, 1912, 1918]. This refers to the 4 preceding Albatross stations. Dr. Th. Mortensen’s Pacific Expedition, 1914-16; off Jolo; about 22 meters; sand and coral; March 17, 1914 (1); about 27 meters; March 21, 1914 (1). Siboga station 99; anchorage off North Ubian (lat. 6°07'30’’ N., long. 120°26’00’’ E.); 16-23 meters; lithothamnion bottom; June 28-30, 1899 [A. H. Clark, 1918] (1, Amsterdam Mus.). Singapore; Svend Gad [A. H. Clark, 1909, 1911, 1912, 1918] (1, C.M.). Singapore; 1899 [A. H. Clark, 1934] (1, Raffles Mus.). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; Amboina; about 2 meters; stony bottom; February 9, 1922 (3). Aru Islands; Dr. H. Merton’s station 10; north of Penambulai; 8 meters; stony bottom; April 2, 1908 [Reichensperger, 1913; A. H. Clark, 1918]. Aru Islands; Dr. H. Merton’s station 12; near Mimien; 15 meters; coarse sand; April 8, 1908 [Reichensperger, 1913; A. H. Clark, 1918]. Aru Islands; Dr. H. Merton’s station 16; near Udjir; 10-14 meters; coral rock and sand; April 16, 1908 [Reichensperger, 1913; A. H. Clark, 1918]. Aru Islands; Dr. H. Merton; between Batu Kapal and Meriri; 10 meters; March 30, 1908 [Reichensperger, 1913; A. H. Clark, 1918]. Siboga station 273; anchorage off Pulu Jedan, eastern coast of the Aru Islands; pearl banks; 13 meters; sand and shells; December 23-26, 1899 [A. H. Clark, 1918] (16, U.S.N.M., E. 472; Amsterdam Mus.). Prince of Wales Channel; 13-16 meters; Alert; sand [Bell, 1884; P. H. Carpenter, 1888; A. H. Clark, 1911, 1913, 1918] (14, B. M.). Torres Strait; 18 meters; sand; Alert [Bell, 1884; P. H. Carpenter, 1888; A. H. Clark, 1911, 1918] (8, B. M.). Port Denison, Queensland [Bell, 1885; P. H. Carpenter, 1888; A. H. Clark, 1911, 1918]. Same [A. H. Clark, 1911, 1918]; (12, Austr. M.; U.S.N.M., 35182, 36161). Same [A. H. Clark, 1908, 1911, 1918] (1, U.S.N.M., 25453). Same; Alert [Bell, 1884; P. H. Carpenter, 1888; A. H. Clark, 1911, 1913, 1918] (6. B. M.). Same; Alert, 5 meters [A. H. Clark, 1911, 1918]. Same; Alert; 5-7 meters [Bell, 1884; P. H. Carpenter, 1888; A. H. Clark, 1911, 1918]. Port Molle, Queensland [Bell, 1885; P. H. Carpenter, 1888; A. H. Clark, 1911, 1918]. Same [A. H. Clark, 1911, 1918] (2, U.S.N.M., 35080; Austr. M.). Same; Alert {Bell, 1884; P. H. Carpenter, 1888; A. H. Clark, 1911, 1913, 1918] (3, B. M.). Same; Alert; 22-36 meters [Bell, 1884; P. H. Carpenter, 1888; A. H. Clark, 1911, 1913, 1918] (1, B. M.). A MONOGRAPH OF THE EXISTING CRINOIDS 383 Holothuria Bank; 27 meters [Bell, 1894; A. H. Clark, 1911, 1918]. Same; 27 meters [A. H. Clark, 1913, 1918] (1, B. M.). au Australia; 15-27 meters [Bell, 1894; A. H. Clark, 1911, 1913, 1918] 2, B. M.). een Australia [A. H. Clark, 1911 (also as West Australia), 1913, 1918] (3, B. M.). Australia; MM. Péron and Lesueur [Guérin-M éneville, 1828-1837 ; Griffith, 1834; Dujardin and Hupé, 1862; A. H. Clark, 1911, 1918] (1, P. M.). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; station 101; Java sea; 49 meters; sand, stones, and sponges; August 5, 1922 (1). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; station 75; Sunda Straits (lat. 6°10’ S., long. 105°44’ E.); 40 meters; sand and shells; July 29, 1922 (7). No locality [A. H. Clark, 1913] (2, B.M.). Geographical range.—From Formosa (Taiwan) and the Philippines to Singapore, and southward to Port Molle, Queensland, and Holothuria Bank, northwestern Australia. Bathymetrical range-—From the shoreline down to 93 meters; nearly all the records are from depths of less than 30 meters. History.—The figure of Comatula carinata published by Guérin-Méneville in his “Tconographie du Régne Animal” (1828-1837) and republished by Griffith in 1834 undoubtedly represents this form. In 1862 Dujardin and Hupé gave a list of 12 names that they found with speci- mens in the Paris Museum. One of these names was Comatula dibrachiata. Dr. P. H. Carpenter did not distinguish this form from milberti. It was included by him in milberti when he assigned that species to the genus Antedon in 1879. Bell’s specific formula for milberti published in 1882 and Carpenter’s emendation of it pub- lished in 1883 both included discoidea. In the Alert report published in 1884 Bell recorded Antedon milberti from Port Molle, Port Denison, the Prince of Wales Channel, and Torres Strait. Except for a single specimen from Port Molle, all Bell’s specimens represented the present form. In the Challenger report on the comatulids published by Carpenter in 1888 the Alert localities just given that were included under Antedon milberti refer to discoidea, with the single exception noted. In 1894 Prof. F. Jeffrey Bell published a list of the echinoderms collected by P. W. Bassett-Smith, Surgeon, R. N., while attached to H. M. S. Penguin during a surveying trip in northwestern Australia. Professor Bell gave no specific localities for the specimens, saying merely that the chief localities were Holothuria Bank, Magnetic Shoal, Cossack Island, and Baudin Island (lat. 14°08’ S., long. 125°36’ E.). He said that Antedon milberti was obtained in 8-15 fathoms. The present species was secured only on Holothuria Bank. In a paper published on December 10, 1908, I described Himerometra ensifer (=Amphimetra ensifer) and compared its arms with those of H. discoidea, the latter name appearing as a nomen nudum. In another paper published on December 23, 1908, Himerometra discoidea was described in detail from a specimen from Port Denison in the United States National Museum, which had many years before been secured from the Australian Museum, and was recorded also from Albatross station 384 BULLETIN 82, UNITED STATES NATIONAL MUSEUM 5138 in the Philippines. It was remarked that this is probably the species that has been recorded from Port Denison as Antedon milberti, but the relatively slender cirri with comparatively long segments contrast sharply with the very stout cirri of milberti (in reality spectabilis, mistaken for milberti=tessellata), which have exceed- ingly short segments. In a revision of the family Himerometridae published in 1909 discoidea was referred to the genus Amphimetra, and later in the same year it was compared with A. formosa. In 1911 I wrote, in a paper on the crinoids of the African coasts and also in a paper on the crinoids of the Paris Museum, that one of the figures (2a) of Comatula carinata published by Guérin-Méneville appears to be a species of Amphimetra, and that there is a specimen of A. discoidea (labeled by P. H. Carpenter Antedon milberti var. dibrachiata) in the Paris Museum from which I suspected it had been drawn. In a paper on the crinoids of the Leyden Museum published in 1911 the synarthrial tubercles of Carpenter’s Antedon (=Amphimetra) pinniformis were compared with those of discoidea. In a memoir on the crinoids collected by the Hamburg Southwest Australian Expedition in 1905, published in 1911, a specimen of discoidea (in reality tessellata, see page 360) was recorded from the ?vicinity of Perth, and the known localities for discoidea were given (see page 369). In a memoir on the crinoids of Australia published in the same year specimens of discoidea were recorded from Port Denison and Port Molle, and notes on them were given. I mentioned Amphimetra formosa from Singapore and the Philippine Islands, comparing it with the specimens of discoidea from Port Denison, and gave the Australian localities from which discoidea is known, and also the range of the species (see page 380). Guérin-Méneville’s figure and the specimen upon which it was probably based were again mentioned. In my memoir on the crinoids of the Indian Ocean published in 1912 I gave the synonymy and range of this form. In the synonymy Amphimetra formosa was included without comment, and in the range Formosa (Taiwan) was mentioned on the basis of a specimen recently received by the United States National Museum. Dr. August Reichensperger in 1913 recorded and gave notes on a number of speci- mens collected by Dr. H. Merton in the Aru Islands and discussed in detail the status of this form, which he regarded as at best but a local variety of milberti (=tessellata). In the same year I recorded 12 lots of specimens in the British Museum, 7 from the Alert collection previously recorded by Bell (1884), 2 from the collection of the Penguin, also previously recorded by Bell (1894), 1 from Western Australia, and 2 without locality, one of which was labeled Actinometra brachiolata. In my memoir on the unstalked crinoids of the Siboga expedition published in 1918, I recorded specimens from Siboga stations 99 and 273 and gave a complete synonymy, together with a complete list of the localities from which this form is known. In the synonymy the references to Grube’s Jaevissima should have been placed under molleri. None of the specimens included under Himerometra anceps (1908) are really this form. The specimen presumably from the vicinity of Perth recorded as discoidea in 1911 was transferred to jacquinoti (=tessellata), as were those from between Fremantle and Geraldton recorded as discoidea in 1914. All the A MONOGRAPH OF THE EXISTING CRINOIDS 385 localities given are correct with the single exception of North Borneo, which refers to Grube’s Comatula laevissima (=molleri). In 1919 Dr. Torsten Gislén said that he entirely agreed with Reichensperger’s opinion that it is incorrect to maintain milberti (=tessellata) and discoidea as species on the basis of the somewhat different cirrals. AMPHIMETRA TESSELLATA PAPUENSIS A. H. Clark [See vol. 1, pt. 2, figs. 37, 38 (radial pentagon), p. 20.] Amphimetra discoidea (part) A. H. Cuarx, Die Fauna Siidwest-Australiens, vol. 3, Lief. 13, 1911, p- 549 (specimens from Port Moresby and Hood lagoon, New Guinea). Amphimetra schlegelit (part) A. H. Cuarx, Crinoids of the Indian Ocean, 1912, p. 111 (New Guinea; Tonga; Fiji). Amphimetra papuensis A. H. Cuarx, Smithsonian Misc. Coll., vol. 61, No. 15, 1913, p. 25 (Tonga and Fiji; characters of the cirri; also Hood lagoon and Port Moresby, New Guinea); Unstalked crinoids of the Siboga-Exped., 1918, p. 83 (in key; range), p. 89 (references; localities from which known).—GisLin, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, p. 48 (ex- tends to Tonga). Diagnostic features —A medium-sized or rather small form with the cirri moderately stout and more or less straight basally but becoming slender and recurved distally, with 30-34 segments of which the longest are usually about as long as broad and the short outer bear prominent long, slender, and sharp dorsal spines; the arms are up to 110 mm. in length; synarthrial tubercles are obsolete; the proximal pinnules are slender, with 17 or 18 segments. Description.—The centrodorsal is discoidal, rather thin, with the small irregular dorsal pole flat and finely pitted. The cirri are arranged in two closely crowded and somewhat irregularly alternating rows. The cirri are XVII, 30-34, about 30 mm. long. The first segment is very short, three to four times as broad as long, and those following slowly increase in length to from the fifth-eight to the ninth-eleventh segments, which are usually about as long as broad or slightly broader than long but vary from half again as broad as long to slightly longer than broad. The distal segments are between half again and twice as broad as long, the terminal gradually lengthening again so that the last two or three are about as long as broad. The cirri are moderately stout basally, tapering very slowly in the proximal half but more rapidly in the distal half, so that the width of the terminal segments is not much more than half that of the basal segments. From the ninth-thirteenth segment onward long, sharp, and conspicuous dorsal spines are developed. The radials are rather long, twice as long laterally as in the median line, with the distal border strongly and regularly concave. The sides of the radial pentagon are parallel. The IBr, are about three times as long as the radials in the median line, rather short, between five and six times as broad as long, with the distal border slightly and very broadly concave in the middle and at the ends straight or very slightly and broadly curved downward. The lateral portions of the dorsal surface are flattened or slightly and broadly swollen, and the lateral edges are throughout in lateral apposition and are rather broadly flattened. The IBr, (axillaries) are low, between two and one-half and three times as broad as long, about twice as long as the IBr,, triangular with the lateral angles rather broadly truncated and the corners 386 BULLETIN 82, UNITED STATES NATIONAL MUSEUM broadly rounded. The extreme lateral portions of the dorsal surface are slightly flattened or slightly and very broadly swollen. The sides of the IBr, and of the first brachials and of the second brachials as far as the base of P, are flattened. Syn- arthrial tubercles are very slight or absent. The 10 arms are about 110 mm. long. The first brachials are wedge-shaped, nearly twice as long exteriorly as interiorly, with the interior sides united for from two-thirds to nearly the whole of their length. The second brachials are slightly larger and somewhat more obliquely wedge-shaped, more than twice as long exteriorly as interiorly. The first syzygial pair (composed of brachials 3+-4) is oblong, short, rather more than three times as broad as long. The next six brachials are practically oblong, about four times as broad as long. Those following are very obliquely wedge- shaped, almost triangular, more than twice as broad as the longer side, soon becoming less and less obliquely wedge-shaped and in the outer half of the arm very short and oblong with the distal ends very slightly produced. Syzygies occur between brachials 3-++4 and 9+-10, again from between brachials 14+-15 to between brachials 17-+-18, and distally at intervals of usually 8 muscular articulations. P; is 9 mm, long, with 18 segments, slender and tapering evenly to the tip. The first segment is about twice as broad as long, the second is nearly as long as broad, the third is longer than broad, and those following increase in length so that the distal are half again as long as broad and the last two or three are twice as long as broad. The second-fourth or -fifth have a slight carinate ridge armed with numerous minute spines on the side toward the arm tip. The sixth and following have the half of the distal edge toward the arm tip slightly produced and everted, and dentate; on the segments succeeding this dentate production of the distal edge soon involves all of the distal edge, though always remaining most developed on the side toward the arm tip. P, is 9.5 mm. long, with 18 segments, and is stouter than P;, with proportionately very slightly shorter segments, though otherwise similar to it. P; is about as long as P, or slightly shorter and is similar to it or very slightly stouter. It has 17 or 18 segments. P, is 6 mm. long, with 15 segments, resembling P; but very slightly more slender basally and tapering more rapidly. The distal pinnules are slender, 6.5 mm. long, with 17 segments, most of which in lateral view are from half again to twice as long as broad. The color in alcohol is purple, darkest on the ventral perisome and on the cirri. Notes.—The specimen described above is one of those from Hood Lagoon, New Guinea. The specimens from this locality are more or less intermediate between typical papuensis and discoidea, though nearer the former. No examples of what should be regarded as typical papuensis are at present available. In the small specimen in the British Museum labeled Tonga and Fiji the cirri have 25 segments of which the outer are more spiny than usual. It resembles very closely the specimens from Hood Lagoon and Port Moresby. Remarks.—This form is simply an extreme development of Amphimetra tessellata discoidea from which it differs in being of more slender and delicate build with more slender cirri, which have longer and more prominent dorsal spines, and in having the A MONOGRAPH OF THE EXISTING CRINOIDS 387 synarthrial tubercles only very slightly developed. It appears to intergrade with discoidea just as discoidea intergrades with tessellata. Localities.—Port Moresby, New Guinea [A. H. Clark, 1911, 1912, 1918] (1, Austr. M.). Hood Lagoon, New Guinea [A. H. Clark, 1911, 1912, 1918] (11, U.S.N.M., 35244). Tonga and Fiji [A. H. Clark, 1912, 1913, 1918] (1, B. M.). Geographical range—From Tonga and Fiji to southeastern New Guinea. Bathymetrical range.—Littoral. Mistory.—In a list of the localities at which Amphimetra discoidea has been found, which I published in 1911, the specimens on the strength of which Port Moresby and Hood Lagoon are included really represent this form. In my memoir on the crinoids of the Indian Ocean (1912) these two localities, and in addition Tonga and Fiji, are given under Amphimetra schlegelii with which, because of its slender cirri with long dorsal spines, this form was at the time supposed to be synonymous. In a paper on the crinoids of the British Museum published in 1913 a specimen labeled Tonga and Fiji was recorded under the name of Amphimetra papuensis. This name here appeared for the first time, but the only characters given were those men- tioned above under “‘Notes.”’ In the Siboga report on the unstalked crinoids (1918) papuensis was included in the key to the species of Amphimetra, and the synonymy and known localities were ora AMPHIMETRA ENSIFER (A. H. Clark) Puate 38, Fiaures 172-174; Puats 39, Figure 179 [See also vol. 1, pt. 1, fig. 86 (lateral view), p. 141; fig. 256 (centrodorsal), p. 255; fig. 337 (cirrus), p. 285; fig. 474 (radial pentagon), p. 361; pt. 2, figs. 39, 40 (radial pentagon), p. 26.] Himerometra ensifer A. H. Cuarx, Proc. Biol. Soc. Washington, vol. 21, 1908, p. 225 (description; Singapore). Amphimetra ensiformis A. H. Cuark, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 7 (listed; nomen nudum). Amphimetra ensifer A. H. Cuarx, Vid. Medd. Naturh. Foren. Kgbenhayn, 1909, p. 158 (Singapore; specimens described), p. 159 (compared with A. schlegeliz), p. 193 (collected at Singapore by Svend Gad); Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p. 17 (compared with A. laevipinna) ; Crinoids of the Indian Ocean, 1912, p. 113 (synonymy; Singapore); Unstalked crinoids of the Siboga-Exped., 1918, p. 83 (in key; range), p. 88 (references; locality); Journ. Linn. Soc. (Zool.), vol. 36, 1929, p. 640 (Singapore; 7 fathoms).—GuisLtin, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1984, p. 45. Diagnostic features —A medium-sized or rather small form with the synarthrial tubercles extravagantly developed and produced; the cirri are 23-30 mm. long, with usually 30-35 segments of which the longest are usually slightly longer than broad; the arms are 80-120 mm. long. 5 Description.—The centrodorsal is thick discoidal. The cirrus sockets are arranged in two closely crowded and irregular more or less alternating rows. The cirri are XV-XX, 30, from 20 to 25 mm. in length. The first segment is very short, and those following gradually increase in length to the sixth, which is about as long as broad. The next five or six segments are slightly longer than broad, 388 BULLETIN 82, UNITED STATES NATIONAL MUSEUM and in those succeeding the length very gradually decreases so that the distal seg- ments are about one-third again as broad as long. The tenth and following segments have well-developed dorsal spines. The disk is completely covered with a pavement of small plates. The synarthrial tubercles between the elements of the IBr series and the first two brachials are extravagantly developed, as in Perometra diomedeae, giving the animal a very characteristic appearance. The 10 arms are 80 mm. long and resemble those of A. discoidea. The segments of all the proximal pinnules are about as long as broad. P, is 7 mm. long, very slender, with 20 segments. P, is 7 mm. long, stouter than P;, with 17 segments. P, is similar to P, but is only 5 mm. long, with 16 segments. P, and the following pinnules are less stout than P, and P; though stouter than P,, 5 mm. long, with 13 segments. The color in alcohol varies from nearly white to dark brown, with the perisome darker. The cirri of the lighter specimens are usually purplish. Notes.—Of the four specimens recorded in 1909 one has the arms 90 mm. long and the cirri XVII, 30-33, from 20 to 23 mm. long. It is white, with the perisome brownish purple and the cirri becoming light purplish distally. Another is similar but is yellow-brown. A third has the arms 80 mm. long and the cirri XIV, 30-33, from 23 to 27 mm. long; in colorit is brownish white. The fourth is smaller and is light bluish gray. Other specimens not previously recorded show the following characters: (1) The arms are about 95 mm. long, and the cirri are XTX, 33-36, from 22 to 24 mm. long; the dorsal pole of the centrodorsal is convex, 3.5 mm. in diameter. (2) The arms are 95 mm. long, and the cirri are XVI, 28-34, from 22 to 24 mm. long; the dorsal pole of the centrodorsal is flat, 4 mm. in diameter. (8) The arms are 115 mm. long and the cirri are XVI, 29-33, 25 mm. long; the dorsal pole of the centro- dorsal is convex, 4.0 to 4.5 mm. in diameter. (4) The arms are 120 mm. long, and the cirri are XVI, 34-39, 25 mm. long; the dorsal pole of the centrodorsal is convex, 3.5 mm. in diameter. (5) The arms are 95 mm. long, and the cirri are XIV, 38-40, from 25 to 30 mm. long; the dorsal pole of the centrodorsal is slightly convex, 3.5 mm. in diameter. (6) The arms are 80 mm. long. (7) The arms are 75 mm. long. (8) The arms are 40 mm. long. Locality —Singapore; Svend Gad [A. H. Clark, 1908, 1909, 1912, 1918] (63, U.S.N.M., 36197, 36260, 36261, E. 1083; C.M.). Same, 13 meters; December 12, 1898 [A. H. Clark, 1929] (7, B. M.). History.—This species was first described as Himerometra ensifer in 1908 from a specimen from Singapore in the collection of the Copenhagen Museum that had been collected by the Danish consul at that port, Svend Gad. In a revision of the family Himerometridae published in 1909 this species was referred to the new genus Amphimetra, being listed through error as Amphimetra ensiformis. Later in 1909 four additional specimens were recorded from Singapore, and notes on them were given. It was listed in my memoir on the crinoids of the Indian Ocean (1912) and in my report upon the unstalked crinoids of the Siboga expedition (1918), and in both A MONOGRAPH OF THE EXISTING CRINOIDS 389 volumes the synonymy and habitat were given. In the Siboga report it was included in a key to the species of Amphimetra. Seven additional specimens from Singapore belonging to the British Museum were recorded in 1929. AMPHIMETRA LAEVIPINNA (P. H. Carpenter) Puate 38, Figures 175, 176 Antedon laevipinna P. H. Carpenter, Journ. Linn. Soc. (Zool.), vol. 16, 1882, p. 502 (description; Canton, China); Proc. Zool. Soc. London, 1882 (1883), p. 746; Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 206 (in key).—Hanrtiaus, Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 81 (discussion; considered a synonym of milberti)—Hamann, Bronns Klassen und Ordnungen des Tier- Reichs, vol. 2, Abt. 3, 1907, p. 1580 (listed) —A. H. Cuark, Proc. U. S. Nat. Mus., vol. 34, 1908, p. 479 (listed); Smithsonian Misc. Coll., vol. 60, No. 10, 1912, p. 2 (=Amphimetra laevipinna); Crinoids of the Indian Ocean, 1912, p. 34 (of P. H. Carpenter, 1888= Amphimetra milberti). Himerometra schlegelii A. H. Cuarx, Proc. Biol. Soc. Washington, vol. 21, 1908, p. 223 (10-armed specimen). Amphimetra milberti (part) A. H. Cuarx, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 7 (listed); Crinoids of the Indian Ocean, 1912, p. 111 (record from Canton, China). Amphimetra schegelii (part) A. H. Cuarx, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 7 (listed). Amphimetra schlegelii (part) A. H. Cuarx, Vid. Medd. Naturh. Foren. Kgbenhayn, 1909, p. 158 (10-armed specimen) ; Crinoids of the Indian Ocean, 1912, p. 111. Amphimetra laevipinna A. H. Cuarx, Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p. 2 (= Antedon laevipinna P. H. Carpenter), p. 15 (compared with A. schlegelii and with A. philiberti; northern representative of the discoidea group), p. 17 (=milberti, in part, of P. H. Carpenter, 1888, and Hartlaub, 1891;=schlegelii, in part, of A. H. Clark, 1908, 1909; Canton, China; detailed descrip- tion of the type and of another from an unknown locality); Proc. Biol. Soc. Washington, vol. 26, 1913, p. 179 (range in east Asia); Journ. Washington Acad. Sci., vol. 5, 1915, p. 214 (Malayan species; range and its significance); Unstalked crinoids of the Siboga-Exped., 1918, p. 83 (in key; range), p. 88 (references; localities from which known).—Morrensen, Hong Kong Nat., Suppl., No. 3, 1934, p. 5 (Canton).—Gistin, Kungl. Fysiogr. Sillsk. Handl., new ser., vol. 45, No. 11, 1934, p. 48 (extends to southern Japan). Amphimetra schleegelii GisLiN, Kungl. Svenska Vet.-Akad. Handl., vol. 59, No. 4, 1919, p. 28 (comparison with milberti [=tessellata]). Diagnostic features —A small species with the arms 70-95 mm. long and the cirri XI-XVII, 25-35, 23 mm. long, with the longest segments about as long as broad. The synarthrial tubercles are prominent, though broad and blunt. This species is closely related to A. ensifer from which it differs most obviously in its broad and blunt, instead of produced, synarthrial tubercles. Description —The centrodorsal is a thick convex disk, bearing marginal cirri. The cirri are XVII, 25-35. The segments are thick, the first five or six short and broad and their immediate successors are about as long as broad. A faint dorsal spine appears on about the tenth segment and increases in size rather rapidly, the segments also shortening somewhat. In the later segments the spine is slightly smaller and projects forward rather less strongly. The opposing spine is larger than any of the preceding spines. The radials are visible and are longer at the sides than in the median line. The IBr, are broadly hexagonal and are partly united laterally. The IBr, (axillaries) are almost triangular and are but little longer than the IBr,. The elements of the [Br series rise toward the middle of the articulation between them into an elevation which stands up as a prominent tubercle. 390 BULLETIN 82, UNITED STATES NATIONAL MUSEUM The 10 arms are 75mm. long. They are in close apposition with their neighbors on each side. The first brachials are broadly pentagonal and are partially united by their inner sides, which are somewhat shorter than the outer. The second brachials are irregularly quadrate with their inner sides projecting beyond the edges of the pre- ceding segments. The outer sides of both the first and second brachials are somewhat flattened laterally, and there is a synarthrial tubercle on the articulation between them just as in the case of the elements of the IBr series. The first syzygial pair (composed of brachials 3+4) is short and oblong. The next few brachials are short, strongly rounded dorsally, and nearly oblong, with slight backward projections alter- nately from the inner and outer sides of their proximal edges. The following brachials are short and sharply wedge-shaped; they are at first considerably broader than long but narrow rather quickly and also commence to overlap so that the middle and later brachials are almost saucer-shaped. Syzygies occur between brachials 3+4 and 9+10, and distally at intervals of 7-13 (usually 10 or 11) muscular articulations. None of the pinnules are specially distinguished. PP, is slender and consists of about 15 smooth cylindrical segments. P, is shorter and slenderer. P, is about as long as P, but it is stouter, composed of thicker segments, which are all very smooth. The basal segments of the next two pairs of pinnules are still rather thick, after which they decrease in size and the pinnules increase very slowly in length, never becoming especially long and consisting of smooth cylindrical segments. The disk is about 6 mm. in diameter, and the anal tube is plated. Sacculi are closely set along the pinnule ambulacra. The color is ight brown, with purplish bands. Notes.—The preceding description is adapted from Carpenter’s original description of the type specimen in the Hamburg Museum. I examined this specimen in 1910 and made the following notes upon it: The 10 arms are about 95 mm. long. The cirri are XI, 27-33, 23 mm. long, moderately slender. The longest proximal segments are from slightly broader than long to about as long as broad, and the short outer segments are about twice as broad as long. Long, sharp, and prominent dorsal spines are devel- oped from the tenth or eleventh segment onward. The cirri have a very slight distal taper. The synarthrial tubercles are prominent, but broad and blunt and not pro- duced as in A. ensifer. I was able to compare the type specimen directly with the 10-armed specimen from Japan in the Copenhagen Museum, which had been included in the original description of Himerometra (Heterometra) schlegelii in 1908, and found that in size, as in other characters, there is the closest agreement. A specimen without locality in the Hamburg Museum closely resembles the type specimen in every particular, though it is slightly smaller, the arms being 70 mm. long. Localities—Canton, China; Werner [P. H. Carpenter, 1881, 1883, 1888; Hartlaub, 1891; A. H. Clark, 1912, 1913, 1915, 1918; Gislén, 1919; Mortensen, 1934] (1, H. M.). “Japan” [A. H. Clark, 1908, 1909, 1912, 1913, 1915, 1918] (1, C. M.). No locality; C. Eberstein [A. H. Clark, 1912, 1913, 1915, 1918] (1, H. M.). Geographical range—The habitat of this species is somewhat uncertain, but is probably southern Japan or somewhere in the area covered by the southerly exten- sion of the southern Japanese fauna. The specimen without locality carried the same A MONOGRAPH OF THE EXISTING CRINOIDS 391 data as a specimen of Eudiocrinus variegatus (see page 157), a species known only from southern Japan. As a locality Canton means little, since it is merely a center for the assembling and distributing of curios from over a wide area, though of course mainly from the more or less immediate vicinity. The type specimens of Heterometra varii- pinna (see page 278) and of Comantheria grandicalyx (see Part 3, p. 515) were also labeled Canton. History.—This species was originally described by Dr. P. H. Carpenter in 1881 from a specimen in the Hamburg Museum labeled Canton. In 1883 he published a specific formula for it. In the Challenger report on the comatulids (1888) he inserted it in a key to the six species of ““Antedon” that did not seem to fit into any of the groups he had established for 10-armed species. In this key it was paired with (Oligometrides) adeonae, both possessing P; and P, and having the lower pinnules tolerably equal; according to Carpenter laevipinna has 25-35 spiny cirrus segments and a distal inter- syzygial interval of 10 or 11 muscular articulations, while adeonae has 20 cirrus seg- ments without spines and a distal intersyzygial interval of 4 or 5 muscular articulations. Hartlaub wrote in 1891 that he had examined two specimens of Antedon milberti from Atjeh (in reality molleri) in the Leyden Museum and also the type specimen of laevipinna in the Hamburg Museum and that laevipinna is identical with milberti. He said further that Carpenter had also realized this and had had the kindness briefly to confirm his conclusion in a letter. He remarked that the Hamburg specimen, according to Carpenter, is noteworthy in having the segments of the pinnules of the second and third pairs not so thick nor so long as usual. After comparing the type specimen of laevipinna directly with one of the specimens from Atjeh he found that this observation was entirely justified. That Antedon laevipinna proves to be a syn- onym of a previously known species Hartlaub said was the more acceptable since it is among those species that Carpenter was not able to place in any one of his groups. J at first accepted Hartlaub’s dispostion of laevipinna, but in 1910 I examined the type specimen in the Hamburg Museum and was convinced that it could not be specifically identical with the specimen from Atjeh (=molleri), which I saw in the Leyden Museum. In 1912 I published some notes on the type of laevipinna and recorded a second specimen without locality in the Hamburg Museum. At the same time under the heading Amphimetra schlegelw I redetermined the 10-armed specimen included in the original description of schlegelii in 1908 as Amphimetra laevipinna. In my memoir on the crinoids of the Indian Ocean (1912) Antedon laevipinna is given as a synonym of Amphimetra milberti, following Hartlaub. In my memoir on the unstalked crinoids of the Siboga expedition (1918) the synonymy and localities of Amphimetra laevipinna are given, and the species is in- cluded in the key to the species of Amphimetra. Dr. Torsten Gislén in 1919 discussed the relationship between milberti (tessellata as understood herein) and schlegelii as represented by the 10-armed specimen, arriving at the conclusion that the latter is probably a good species because of the relatively large number of pinnule segments combined with the small arms. Family MARIAMETRIDAE A. H. Clark Palmata group (part) P. H. Carpenter, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 223; Journ. Linn. Soc. (Zool.), vol. 21, 1889, p. 305 (species collected at Mergui).—Harriavus, Nova 392 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Acta Acad. German., vol. 58, No. 1, 1891, p. 13 (diagnosis; range; 30 species), p. 35 (diagnosis; discussion; key to the species); Bull. Mus. Comp. Zool., vol. 27, 1895, p. 137 (structural position) ; Mem. Mus. Comp.}Zool., vol. 27, 1912, p. 409.—A. H. Cuan, Proc. U. S. Nat. Mus., vol. 33, 1907, pp. 127, 130 (new species); Smithsonian Mise. Coll., vol. 50, 1907, p. 344 (characters differentiating this group not diagnostic), p. 355 (Himerometra, gen. nov., includes part of species), p. 357 (Cyllometra, gen. nov., includes part of species). Himerometridae (part) A. H. Ciarx, Proc. Biol. Soc. Washington, vol. 21, 1908, p. 135; Proc. U. S. Nat. Mus., vol. 34, 1908, p. 211 (includes Oligometra, Cyllometra, Himerometra, and Pontiometra). Mariametrinae A. H. Cuarx, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 176 (includes Mariametra and Dichrometra); Mem. Australian Mus., vol. 4, 1911, p. 728 (raised to family rank). Stephanometrinae A. H. Ciarx, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 176 (includes Ozy- metra and Stephanometra); Mem. Australian Mus., vol. 4, 1911, p. 728 (raised to family rank); Journ. Washington Acad. Sci., vol. 5, 1915, pp. 126-134 (bathymetrical range; phylogenetic and paleontological significance). Mariametridae A. H. Cuark, Proc. U. S. Nat. Mus., vol. 40, 1911, p. 649 (referred to the Oligoph- reata); Die Fauna Siidwest-Australiens, vol. 3, Lief. 18, 1911, p. 438 (1 genus and 4 species in Australia); Mem. Australian Mus., vol. 4, 1911, p. 717 (proportion of species of this family in the Australian fauna as known to P. H. Carpenter), p. 720 (proportion of species of this family in the Australian fauna), p. 725 (Mariametra and Selenemetra absent from Australia), p. 730 (in key), p. 731 (1 genus in Australia) ; Crinoids of the Indian Ocean, 1912, p. 6 (exclusively confined to the East Indian region; number of genera and species), p. 11 (represented in the Ceylon region by Dichrometra and Selenemetra), p. 12 (represented in the Red Sea by Dichro- metra; in the southeast African region by the same), p. 22 (distribution in detail; down to 59 fathoms), pp. 47, 52 (in keys), p. 57 (key to the included genera); Proc. Biol. Soc. Washington, vol. 26, 1913, p. 141 (revised; considered as including the genera of Pontiometridae, except Epimetra, which is assigned to the Colobometridae; 2 new genera, Liparometra and Lamprometra, described); Bull. Inst. Océanogr., Monaco, No. 294, 1914, pp. 7-8 (relation to temperature).— ALEXANDER, Rec. Western Australian Mus., vol. 1, pt. 3, 1914, p. 108.—A. H. Cuarx, Journ. Washington Acad. Sci., vol. 4, 1914, pp. 559-563 (correlation of geographical and bathymetrical ranges), p. 582 (relation to temperature of habitat); vol. 5, 1915, pp. 126-134 (bathymetrical range; phylogenetic and paleontological significance); American Journ. Sci., vol. 40, 1915, p. 67 (detailed philosophical discussion of bathymetrical range); Internat. Revue gesamt. Hydrobiol. und Hydrogr., 1915, p. 223 ff. (detailed account of the distribution of Australian species) ; Smithsonian Mise. Coll., vol. 65, No. 10, 1915, p. 45 ff. (phylogenetic study) ; Unstalked crinoids of the Siboga-Exped., 1918, p. 97 (key to the included genera); Univ. Iowa Stud. Nat. Hist., vol. 9, No. 5, 1921, p. 14 (mot represented in the West Indies); Smithsonian Misc. Coll., vol. 72, No. 7, 1921, p. 3— Gustin, Nova Acta Reg. Soc. Sci. Upsaliensis, ser. 4, vol. 5, No. 6, 1922, pp. 9, 74; Zool. Bidrag Uppsala, vol. 9, 1924, pp. 19, 42, 55, 78, 84, 89, 99, 213, 230, 231, 235, 240, 280, 287; Vid. Medd. Dansk. Naturh. Foren., vol. 88, 1927, pp. 22, 23.—H. L. Cuarx, Rec. South Australian Mus., vol. 3, 1928, p. 368.—A. H. Cuarx, Journ. Linn. Soc. (Zool.), vol. 36, 1929, p. 640; Rec. Indian Mus., vol. 34, pt. 4, 1932, p. 557; Treubia, vol. 14, livr. 2, 1933, p. 213.— Gistin, Kungl. Fysiogr. Sillsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 20, 25; Temminckia, vol. 1, 1936, p. 302.—GisLGn, Kungl. Fysiogr. Sallsk. Lund Férh., vol. 7, No. 1, 1936, p. 12.—A. H. Crarx, John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1936, pp. 88, 100, 103.—H. L. Clark, Mem. Mus. Comp. Zool., vol. 55, 1938, p. 35.—Boong, Bull. Vanderbilt Marine Mus., vol. 7, 1938, p. 120. Stephanometridae A. H. Cuarx, Proc. U. S. Nat. Mus., vol. 40, 1911, p. 649 (referred to the Oligo- phreata); Die Fauna Siidwest-Australiens, vol. 3, Lief. 18, 1911, p. 438 (1 genus and 1 species in Australia); Mem. Australian Mus., vol. 4, 1911, p. 720 (proportion of species of this family in the Australian fauna), p. 725 (Oxymetra absent from Australia), p. 730 (in key), p. 731 (1 genus in Australia); Crinoids of the Indian Ocean, 1912, p. 6 (number of genera in the East Indian region; number also found in the Atlantic; number represented by closely allied genera in the Atlantic; number exclusively East Indian; number of East Indian species), p. 9 (absent from Australia), p. 10 (absent from Japan), p. 11 (represented in Ceylon by Stephanometra), p. 12 (represented in the southeast African region by the same), p. 13 (East Indian and corre- A MONOGRAPH OF THE EXISTING CRINOIDS 393 sponding West Indian genera), p. 22 (distribution in detail; down to 35 fathoms; all the large species confined to the Philippine Islands, the Moluccas, and the Solomons; represented in the Atlaatie by Analcidometra [in reality belongs to the Colobometridae]), pp. 47, 51 (in keys), p. 57 (key to the included genera); Journ. Washington Acad. Sci., vol. 4, 1914, pp. 559-563 (correlation of geographic and bathymetrical ranges), p. 582 (relation to temperature of habitat) ; American Journ. Sci., vol. 40, 1915, p- 67 (detailed philosophical discussion of bathymetrical range) ; Internat. Revue gesamt. Hydrobiol. und Hydrogr., 1915, p. 223 and ff. (detailed account of the distribution in Australia); Unstalked crinoids of the Stboga-Exped., 1918, p. 93; Univ. Iowa Stud. Nat. Hist., vol. 9, No. 5, 1921, p. 14 (not represented in the West Indies); Smith- sonian Mise. Coll., vol. 72, No. 7, 1921, p. 3—Gustin, Zool. Bidrag Uppsala, vol. 9, 1924 p. 10 (P2), pp. 19, 42, 55, 78, 84, 89, 99, 231, 280, 287; Arkiv fér Zoologi, vol. 19, No. 32, 1928, p. 6; Kungl. Fysiogr. Sillsk. Handl., new ser., vol. 45, No. 11, 1934, pp. 20, 25; Kungl. Fysiogr. Sallsk. Lund Férh., vol. 7, No. 1, 1936, p. 11. Stéphanométres A. H. Crarx, Bull. Mus. Hist. Nat., Paris, 1911, p. 252. Mariamétres A. H. Crark, Bull. Mus. Hist. Nat., Paris, 1911, p. 253. Diagnosis.—A family of the superfamily Mariametrida in which the elements of the [Br series are united by synarthry, and the arms are always more than 10 in num- ber, the I1Br series being always2. The perisome of the disk ismost commonly naked, but sometimes there are more or less conspicuous scattered thick plates or concre- tions, which may be more or less contiguous, especially along the ambulacral grooves, or the disk may be almost or quite completely plated (see Part 2, p. 228). The lateral perisome of the pinnules usually contains simple, more rarely forked or multiradiate, spicules, between which and the edge of the pinnulars there may be small straight spicules or small rods; but calcareous deposits are frequently wholly absent (see Part 2, pp. 241 [Stephanometra] to 244). Geographical range.—Southern Japan from the Oki Islands in the Sea of Japan to Tokyo Bay, the Bonin Islands, the Philippine, Pelew, Caroline, Marshall, and Hawaiian Islands, Samoa, Fiji, the Tonga Islands, New Caledonia, northern Australia south to Cape Hillsborough, Queensland, and the Abrolhos Islands and possibly Perth, Western Australia, and the Lesser Sunda Islands, and westward to Madagascar and the east coast of Africa from the Red Sea south to Durnford Point, Zululand. Bathymetrical range-——From the shore line down to 245 (?421) meters. Remarks.—The 27 forms included in the family Mariametridae fall into seven genera—Ozrymetra, Stephanometra, Pelometra, Liparometra, Lamprometra, Dichro- metra, and Mariametra. The structural range within the Mariametridae is very small when compared with that within the Zygometridae and Himerometridae, and especially within the Colobometridae. The genus Oxrymetra, with its long many-jointed cirri and characteristic lower pinnules, is rather sharply separated from the other genera of the family, and bears a considerable superficial resemblance to the genus Pontiometra of the Colobometridae. The other six genera are all closely related, and young or immature individuals of the included species are sometimes not easily determined generically. The species of the genus Lamprometra is exceedingly variable in regard to the proximal pinnules and the cirri, though these always keep within well-defined limits, and there is never any difficulty in distinguishing it from species in other genera. 20824440 —26 394 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Gislén said that it seemed to him rather unfortunate to characterize the genera within the family Mariametridae on the basis of the relationships between the relative lengths of the proximal pinnules, P;, P2, and Ps, as has been done in the case of Liparo- metra, Lamprometra, and Dichrometra, which surely are very closely allied. It cannot be denied that these three genera are very closely related and that certain individuals are not always readily placed generically merely by reference to the proximal pinnules. Dr. Hubert Lyman Clark also wrote that he was somewhat inclined to question the desirability of recognizing these three very closely allied genera. Yet Liparometra, Lamprometra, and Dichrometra appear to represent definite generic types, which are most obviously and readily differentiated from each other by the interrelationships between the proximal pinnules when these are fully and typically developed. The species of Mariametridae are especially characteristic of the region from southern Japan to Polynesia, northern Australia, and east Africa. Two genera, Stephanometra and Lamprometra, occur from the Philippines and Polynesia to north- ern Australia and east Africa; one, Dichrometra, ranges from southern Japan to the Admiralty Islands and New Guinea and westward to east Africa; one, Liparometra, is found from southern Japan to the Tonga Islands and the Lesser Sunda Islands; one, Mariametra, occurs from southern Japan to Amboina and the Lesser Sunda Islands; one, Oxymetra, occurs from the Philippines to New Britain and westward to Ceylon, and one, Pelometra, is known only from Amboina. Like the species of the family Himerometridae, the species of Mariametridae are especially characteristic of the littoral and sublittoral regions. Of the seven included genera, five occur along the shore line. The two that have not as yet been taken in shore collecting are Mariametra, occurring in 40-153 meters, and Pelometra, known from only a single specimen dredged in 91 meters. Four genera, Oxymetra, Pelo- metra, Liparometra, and Lamprometra, are not known from a greater depth than 95 meters, though three, Stephanometra, Mariametra, and Dichrometra, extend down to more than 150 meters, the last having been brought up from 245 meters. History—The family Himerometridae, which I established in 1908, was on June 25, 1909, restricted by the removal of Cenometra and Cyllometra to the new family Colobometridae, which was not at the time defined. On September 14, 1909, it was further restricted by the establishment of the family Pontiometridae for the reception of the single genus Pontiometra, and at the same time was divided into three subfami- lies, Himerometrinae, Stephanometrinae, and Mariametrinae. In my paper on the systematic position of the genus Marsupites published on June 24, 1911, the families Stephanometridae, Pontiometridae, and Mariametridae were included in the families belonging to the suborder Oligophreata, the two last being here mentioned as families for the first time. In a paper on the recent crinoids of Australia published on August 17, 1911, I formally divided the family Himerometridae as previously understood into three families, the Himerometridae, Stephanometridae, and Mariametridae (see p. 184). In my memoir on the crinoids of the Indian Ocean published in 1912 I recognized the family Stephanometridae, with the genera Orymetra and Stephanometra, the family Pontiometridae, with the genera Pontiometra and Epimetra, and the family Mariame- tridae, with the genera Selenemetra, Mariametra, and Dichrometra. A MONOGRAPH OF THE EXISTING CRINOIDS 395 In a revision of the family Mariametridae published in 1913 I suppressed the family Pontiometridae, transferring the genus Epimetra to the Colobometridae and Pontiometra to the Mariametridae. As here constituted the family Mariametridae included the genera Pontiometra, Oxymetra, Liparometra, gen. nov., Lamprometra, gen. nov., Dichrometra, and Mariametra. In the memoir on the unstalked crinoids of the Siboga expedition published in 1918 I recognized the family Stephanometridae, with the single genus Stephanometra, and the family Mariametridae, with the genera just given. In the present work the family Mariametridae is considered as including, together with the genera assigned to it in 1913 (except Pontiometra), the genus Stephanometra, the family Stephanometridae being suppressed, and also the new genus Pelometra. The genus Pontiometra is herein placed in the family Colobometridae. In his report upon the comatulids collected by the Challenger expedition Car- penter placed all the species of the family Mariametridae known to him in the Palmata group, which in addition included Antedon manca, A. disciformis, and A. clemens, the first two of which belong to the family Colobometridae, falling in the genus Cyllometra, while the third belongs to the family Himerometridae, falling in the genus Heterometra. KEY TO THE GENERA OF THE FAMILY MARIAMETRIDAE a‘, Cirri large and long, composed of more than 40 (usually of more than 50) segments (from the Philippines southward to New Britain and the Kei Islands, and westward to Ceylon; 0-82 INOLETS) eee et nS ee Ss a ee os eg Oxymetra (p. 396) a, Cirri of moderate length or short, composed of less than 40 segments. bt. One or more of the oral pinnules enlarged, greatly stiffened, sharp pointed, and spinelike; division series well separated, the component ossicles with rounded ventrolateral exten- sions (from Macclesfield Bank and Philippine Islands to Pelew, Caroline, and Marshall Islands, Samoa, Fiji, the Tonga Islands, New Caledonia, Torres Strait, and Lesser Sunda Islands, and westward to Madagascar; 0-245 meters)_-___-___- Stephanometra (p. 407) b?. Enlarged lower pinnules tapering distally to a fine and usually (but not always) flexible point, without spiniform tips; division series usually in lateral contact, the component ossicles never with ventrolateral extensions. c!, Genital pinnules with a narrow expansion on lower edges of segments (Amboina; about 91 MOGULS) eee ee 8 es eee eee ene ee eee Pelometra (p. 459) c?. No expansion of lower edge of segments of genital pinnules. d', P, and P; slender, similar, and of same length (southern Japan from Hirado Strait to Sagami Bay, Bonin Islands, and southward to Tonga Islands, Port Molle, Queensland Moluccas, and Lesser Sunda Islands; 0-69 meters)----------- Liparometra (p. 460) d?, P, and P; unequal in length or stoutness, usually in both. e!, P. longer and stouter than P; (from Hongkong and Philippines to Caroline, Marshall, and Hawaiian Islands, Fiji, Tonga Islands, New Caledonia, the Solomon Islands, north- ern Australia south to Cape Hillsborough, Queensland, and Abrolhos Islands and possibly Perth, Western Australia, and westward to east coast of Africa from Red Sea south to Dar-es-Salaam, Tanganyika Territory; 0-51 [?421] meters) Lamprometra (p. 472) é. P; longer and stouter than P». f'. Lateral portions of dorsal surface of the division series smooth (from southern Japan, from Korean Straits to Sagami Bay, to Hongkong, Philippines, Pelew and Ad- miralty Islands, New Guinea, and Amboina, and westward to Madagascar and east coast of Africa from Lamu, Kenya, southward to Durnford Point, Zululand; G=164jmeters)2ose-t SUES Sean Sere Se ees Dichrometra (p. 536) 396 BULLETIN 82, UNITED STATES NATIONAL MUSEUM f?. Lateral portions of dorsal surface of division series with closely crowded small spines, tubercles, or granulations (southern Japan from Oki Islands in Sea of Japan to Tokyo Bay, and southward to Amboina, Kei Islands, Timor, Flores, and Sapeh Strait, and westward to Mergui Archipelago; 40-153 meters) Mariametra (p. 566) Genus OXYMETRA A. H. Clark Antedon (part) Harriaus, Nachr. Ges. Géttingen, May 1890, p. 176, and following authors, Himerometra (part) A. H. Cuarx, Smithsonian Mise. Coll., vol. 50, 1907, p. 356. Dichrometra (part) A. H. Cuarx, Proc. Biol. Soc. Washington, vol. 22, 1909, pp. 13, 76. Oxymetra A. H. Cuarx, Proce. Biol. Soc. Washington, vol. 22, 1909, p. 13 (characters; genotype Antedon erinacea Hartlaub, 1890), p. 176 (referred to the Stephanometrinae); Mem. Australian Mus., vol. 4, pt. 15, 1911, p. 725 (absent from Australia); Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p. 22 (includes Selenemetra) ; Crinoids of the Indian Ocean, 1912, p. 57 (in key), p. 131 (original reference; type); Proc. Biol. Soc. Washington, vol. 26, 1913, p. 141 (includes Selene- metra; referred to the Mariametridae), p. 143 (synonymy; genotype; range, depth; included species), p. 142 (in key); Unstalked crinoids of the Siboga-Exped., 1918, p. 97 (in key; range), p. 98 (key to the included species). Selenemetra A. H. Cuark, Proc. U. 8. Nat. Mus., vol. 39, 1911, p. 541 (diagnosis; genotype Antedon finschii Hartlaub, 1890; referred to the Mariametrinae) ; Memoirs of the Australian Mus., vol. 4, pt. 15, 1911, p. 725 (absent from Australia); Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p. 22 (synonym of Ozymetra); Crinoids of the Indian Ocean, 1912, p. 9 (absent from Australia), p. 10 (absent from Japan; reason), p. 11 (represented in Ceylon region), p. 22 (distribution in detail), p. 57 (in key), p. 139 (original reference; type); Proc. Biol. Soc. Washington, vol. 26, 1913, p. 141 (synonym of Oxymetra). Diagnosis.—A genus of Mariametridae in which the cirri are long and stout with more than 40 (rarely less than 50) segments, of which the distal bear dorsal spines; several pairs of the proximal pinnules are moderately enlarged and moderately elon- gated, and more or less stiffened, at least in the basal half; and the arms are 40 or more in number. Geographical range-—From the Philippine Islands southward to New Britain and the Kei Islands and westward to Ceylon. Bathymetrical range.—From the shoreline down to 82 meters. Remarks.—The genus Oxymetra as herein understood includes three closely related species that may eventually prove to be simply more or less distinct forms of the same species. Although Oxymetra erinacea with its stiff, enlarged, and spinelike proximal pinnules seems at first sight to be very different from O. finschii with its more slender proximal pinnules of which more or less of the distal portion is flexible, this difference is really of little importance, for in many single species in genera of the Zygometridae, Himerometridae, Mariametridae, and Colobometridae the proximal pinnules vary greatly in length and stiffness. Ozxymetra tenwicirra is simply a form of OQ. erinacea with elongated cirrus segments. FTistory.—The two first-known species of this genus, finschii and erinacea, were described by Dr. Clemens Hartlaub in the genus Antedon in 1890. When the new genus Himerometra was established by me in 1907 both of these species were included in it. In 1908 I described a new species under the name of Himerometra gracilipes (=finschii). In the revision of the family Himerometridae published on January 9, 1909, I referred Himerometra gracilipes to the new genus Dichrometra. I said that I had A MONOGRAPH OF THE EXISTING CRINOIDS 397 not been able satisfactorily to place Antedon finschii and A. erinacea because of a lack of material for comparison. I remarked that the first (finschii) appears to be most closely related to Pontiometra andersoni, and the diagnosis of the genus Pontio- metra may have to be altered for its reception; the second (erinacea) appears to repre- sent a distinct generic type for which the name Oxymetra would be appropriate. In a paper published on April 17, 1909, I described a new species, Dichrometra aranea, in the family Himerometridae. In a paper published on September 14, 1909, a new subfamily of the family Himerometridae, the Stephanometrinae, was established, which included the genera Oxymetra and Stephanometra; the genera Mariametra and Dichrometra were placed in another new subfamily, the Mariametrinae. In 1911 I established the new genus Selenemetra in the subfamily Mariametrinae, with Antedon finschii as the genotype. At the same time I described a new species in this genus, S. viridis. In a paper on the crinoids of the Hamburg Museum published in 1912, after a personal examination of Hartlaub’s type specimen of Antedon erinacea, I said that this species has nothing to do with the species of Stephanometra but is closely related to the species that I had grouped in the genus Selenemetra, especially to S. tenuicirra. Pointing out that the generic name Ozxymetra with Antedon erinacea as the genotype has precedence over Selenemetra with Antedon finschii as the genotype, I said that all the species heretofore assigned to Selenemetra must be transferred to Oxymetra, of which Selenemetra is a synonym. In my memoir on the crinoids of the Indian Ocean, published 13 days later than the preceding paper, though written some time before it, I placed Ozymetra, with the single species erinacea, in the family Stephanometridae, and Selenemetra, with the species finschii, aranea, gracilipes, and viridis, in the family Mariametridae. In my revision of the family Mariametridae published in 1913 the genus Oxymetra is given with Selenemetra as a synonym, and the species aranea, erinacea, finschii, gracilipes, tenuicirra, and viridis are listed as belonging to it. In my report on the unstalked crinoids of the Siboga expedition I included a key to the species of Oxymetra. The species recognized were tenwicirra, erinacea, finschit (including gracilipes), and aranea (including viridis). KEY TO THE SPECIES IN THE GENUS OXYMETRA a, Cirrus segments long, the longest about twice as long as broad and the distal about as long as broad; 69-78 cirrus segments; 49 arms 150 mm. long (between Borneo and eastern Java; S210 GLETS) ese ee tenuicirra (p. 401) a?. Cirrus segments shorter, the longest but little longer than broad and the distal twice as broad as long or even broader. b!. First four pairs of pinnules very stiff and spinelike; 36-51 (usually more than 40) arms 90-115 mm. long; cirri with 48-63 (usually 50-60) segments (Philippines to Kei Islands, and westward to Sunda Strait; 0-50 meters) _.-..-_-------------_------ erinacea (p. 398) b?. Proximal pinnules, though somewhat stiffened, are not spinelike, but become flexible distally and end in a delicate and more or less flagellate tip; 41-80 cirrus segments; 40-44 arms 90-140 mm. long (Philippines to New Britain and westward to northeastern Ceylon; 0-77 EN GETS) Ee ae ne ee finschii (p. 402) 398 BULLETIN 82, UNITED STATES NATIONAL MUSEUM OXYMETRA ERINACEA (Hartlaub) Pratrp 41, Freurps 189, 190; Puatn 42, Ficure 194 Antedon erinacea Hartiaus, Nachr. Ges. Gottingen, May 1890, p. 177 (description; Cebu); Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 38 (in key), p. 52 (detailed description and com- parisons; Cebu), p. 58 (in first section of species of the Palmata group with spiny lower pin- nules), pl. 3, fig. 29.—Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1581 (listed).—A. H. Cuarx, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 13 (type of Oxymetra); Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p. 2 (identity) ; Crinoids of the Indian Ocean, 1912, p. 37 (identity.) Himerometra erinacea A. H. Cuarx, Smithsonian Misc. Coll., vol. 50, 1907, p. 356 (listed). Oxymetra erinacea A. H. Cuarx, Smithsonian Misc. Coll., vol. 60, No. 10, 1912, p. 2 (identity), p. 21 (Cebu; detailed description of the type), p. 22 (comparison with finschit and tenuicirra); Crin- oids of the Indian Ocean, 1912, p. 37 (identity); p. 182 (synonymy; Cebu); Unstalked crinoids of the Siboga-Exped., 1918, p. 99 (in key; range).—Gis.tin, Kungl. I'ysiogr. Sallsk. Handl., new ser., vol. 45, No, 11, 1934, p. 33. Diagnostic features—The longest cirrus segments are but little longer than broad, and the distal cirrus segments are twice as broad as long or even broader; the pinnules of the first four pairs are enlarged, moderately elongated, much stiffened, spinelike, and more or less erect; there are 36-51 (usually more than 40) arms 90-115 mm. long; and the cirri have 48-63 (usually 50-60) segments. Description.—Vhe centrodorsal is large and hemispherical and is almost entirely covered by cirrus sockets. The dorsal pole is very small, irregular in shape, and slightly concave. The cirri are XXX—XXXV (about XXV according to Hartlaub), 50-60, about 40 mm. long. They are rather scattered, rather thin and slender, and are distally somewhat compressed laterally. The longer proximal segments are about as long as broad. After the middle of the cirri the segments gradually decrease in length so that those in the outer portion of the cirri are twice as broad as long. The shorter segments in the distal half of the cirri have the distal dorsal edge thickened and everted and very finely spinous, this eversion gradually becoming more and more triangular in end view, the spinosity concurrently becoming gradually restricted to the lateral part of the eversion, and on the terminal segments becoming a single smooth sharp spine. The opposing spine is well developed. The radials are partially visible in the interradial angles of the calyx. The IBr, are laterally united. The IBr, (axillaries) are pentagonal. The postradial series divide four times, but the IVBr series arise only on the outer side of the IIJBr axil- laries. In other words, the arms are arranged in 3, 2, 2, 3 order so that the normal number of arms is (3-+2+2+3)5=50. The division series are 2 and are rather slender and elongated. They are somewhat compressed laterally with a strongly convex dorsal surface and perfectly straight sides. Small synarthrial tubercles are present. The division series are rather widely separated, and the outermost arms are rather rarely in contact. The two ossicles following each axillary, with the exception of the first brachials, are always only partially united. The 51 arms in the type specimen are rather slender and relatively short and are 105-115 mm. in length. Their dorsal surface is entirely smooth. The first brachials are strongly rounded dorsally, somewhat thickened dorsoventrally, and as long as broad. The second brachials are slightly longer exteriorly than interiorly. The first syzygial pair (composed of brachials 3+4) is markedly longer than broad. A MONOGRAPH OF THE EXISTING CRINOIDS 399 About eight or nine of the following brachials are oblong, and those succeeding are shorter and rather sharply wedge-shaped. Toward the ends of the arms the brachials become more bluntly wedge-shaped, and finally more squarish. Syzygies occur between brachials 3+4, again somewhere between the forty-first and fifty-first brachials, and distally at intervals of usually nine or ten, but occa- sionally only four, muscular articulations. The disk is surrounded by a thick circlet of spiny pinnules. The pinnules of the first four pairs are very stiff and sharply spiny. Those of the first and second pairs are about 14 mm. long and are composed of about 25 short segments. The pinnules of the third and fourth pairs decrease in length. Ps on the outer arms is also stiff- ened and is about 5 mm. long. Ps and P; are very small. The following pinnules gradually increase in length, distally reaching about 8 mm. The spiny pinnules are equally stout on all the arms, and the pinnules on the inner side of the arms are as long as those on the outer. The segments of the enlarged lower pinnules are all short, about as long as broad, and of uniform size. These pinnules are stiffened, though scarcely so much so as those of the species of Stephanometra, and the tip, though sharp, ends in the normal manner. The pinnules are more slender than the enlarged and spiny pinnules of Stephanometra and are evenly tapering. The disk of the type specimen has been lost. Sacculi are thickly set on the pinnules. The color in alcohol is uniform light brown. Notes.—The preceding description is adapted from Hartlaub’s original descrip- tion, with supplementary information supplied by my examination of the type speci- men at Hamburg in 1910. The specimen from the Danish expedition to the Kei Islands station 20 has 35 arms 110 mm. long. One of the specimens from the Danish expedition to the Kei Islands station 90 has 41 arms 95 mm. long. A single IVBr series is present, on the outer side of one of the inner IIIBr series—that is, next to the midradial line. The cirri are XXXVI, 54-63, from 40 to 50 mm. long. The longest cirrus segments are one-third again as long as broad, and the short distal segments are from half again to twice as long as broad. P, is much stiffened, 9 mm. long with 20 segments of which the first is very short, the fourth is nearly as long as broad, and the tenth and following are twice aslong as broad. The distal side of the segments (toward the arm tip) is slightly concave. P, is cylindrical, 10 mm. long, with 19 segments; it is about as stout basally as P, but tapers more gradually and is less slender distally. P ; is 6 mm. long with 13 segments, of which the last three or four are about three times aslongasbroad. P,is 5.5 mm. long, with 12 segments. Another specimen has 49 arms 105 mm. long; the extra axillary is next to the mid- radial line. The cirri have 50-52 segments and are up to 40 mm. in length. P, is 11.5 mm. long, with 22 segments. P, is 12 mm. long, with 22 segments, more stiffened than P,, tapering very slowly and therefore stouter in the distal half. P;is 9 mm. long, with 17 segments, and resembles P;. P,is 6.5 mm. long, with 14 segments, and resem- bles P3. 400 BULLETIN 82, UNITED STATES NATIONAL MUSEUM A third specimen has 44 arms 110 mm. long. The extra axillaries are usually next the midradial line. The cirri have 51-55 segments and are from 40 to 45 mm. long. A fourth example has 44 arms 110 mm. long. The cirri have 48-57 segments and are 40 to 43 mm. long. Another specimen has 36 arms 90 mm. long. Other specimens have 38 (1), 40 (3), 41 (1), 42 (2), 43 (3), 44 (1), 45 (4), and 46 (3) arms, which are usually from 100 to 110 mm. long. Abnormal specimen.—In a specimen from the Danish expedition to the Kei Islands station 90 with 46 arms 90 mm. long one of the arms, the outermost arm on one of the postradial series, has the tenth brachial bearing both the continuation of the arm and on the outer side another arm about two-thirds as large, which is directed obliquely distally at an angle of 45°. Localities —Cebu, Philippines; Captain Ringe [Hartlaub, 1890, 1891; A. H. Clark, 1907, 1909, 1912, 1918] (1, H. M.). Danish expedition to the Kei Islands; Dr. Th. Mortensen; station 20; about 50 meters; sand and shells; April 14, 1922 (1). Danish expedition to the Kei Islands; Dr. Th. Mortensen; station 90; Sebesi Strait, between Java and Sumatra; 36 meters; hard bottom; August 1, 1922 (24). Geographical range.—From the Philippines southward to the Kei Islands and westward to the Sunda Strait, between Java and Sumatra. Bathymetrical range-—From the shoreline down to about 50 meters. History —This species was described from a specimen from Cebu collected by Captain Ringe under the name of Antedon erinacea by Dr. Clemens Hartlaub in 1890 and was described and figured in the following year. Hartlaub said that Antedon erinacea is an undoubted new species, immediately distinguished from all related species by the large number of spinelike pinnules that surround the disk and also by the relatively short and numerous segments of these pinnules. He said that the single species that in respect to the shortness and number of segments in the proximal pinnules resembles erinacea is Antedon (Stephanometra) indica, but that any confusion with this species is prevented because this form, in contrast to erinacea, possesses only a small number of elongated and styliform pinnules. He noted that erinacea differs from all similar species in the following points: It has relatively very long cirri, which are composed of 50-60 segments; it possesses IVBr series, and therefore a large number of arms; and lastly—a feature that is very charac- teristic—the second syzygy is separated by an unusually long distance from the first. In my first revision of the old genus Antedon published in 1907 this species was referred to the new genus Himerometra. In the revision of the family Himerometridae published in 1909 I did not place Antedon erinacea definitely in any genus. I said that it appears to represent a distinct generic type for which the name Orymetra would be appropriate. In a paper on the crinoids of the Hamburg Museum published in 1912 I gave notes on the type specimen of Antedon erinacea, which I had examined in 1910, and said that this species is most nearly related to that which I had recently described under the name Selenemetra (= Ozxymetra) tenuicirra, which has similarly stiffened and enlarged proximal pinnules; but in tenwicirra the pinnules are stouter proximally A MONOGRAPH OF THE EXISTING CRINOIDS 401 and more delicate distally so that they taper less regularly, and they are considerably shorter, while the cirri are very much longer with more numerous segments of which the distal are much longer. I said that from the published description and figure I had assumed that this form was allied to the species of Stephanometra with which it was associated by Hartlaub and, as it differed materially from all the forms which I had grouped in that genus, I had suggested the generic name Oxymetra for it. Examina- tion of the specimen, however, showed that it has nothing to do with the species of Stephanometra but instead is closely related to the species which I had grouped in the genus Selenemetra, especially to S. tenuicirra. In my memoir on the crinoids of the Indian Ocean published in 1912 Oxymetra erinacea was placed in the family Stephanometridae, Selenemetra and the species included in it (finschii, aranea, gracilipes, and viridis) being assigned to the family Mariametridae. In my report on the unstalked crinoids of the Siboga expedition, erinacea was included in the key to the species of the genus Orymetra but was not further considered, as it was not secured by the Siboga. OXYMETRA TENUICIRRA (A. H. Clark). Puate 43, Figures 195, 196 Selenemetra tenuicirra A. H. CLarx, Ann. Mag. Nat. Hist., ser. 8, vol. 10, 1912, p. 32 (description; Siboga station 320); Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p. 22 (nearest to Oxymetra erinacea; comparison). Oxymetra tenuicirra A. H. Cuarx, Unstalked crinoids of the Siboga-Exped., 1918, p. vu (discovery by the Siboga and its significance), p. 98 (in key; range), p. 99 (detailed description; station 320), p. 276 (listed), pl. 19, figs. 37, 38.—Gisuén, Kungl. Fysiogr. Sillsk. Handl., new ser., vol. 45, No. 11, 1934, p. 33. Diagnostic features.—Resembling O. erinacea but with the cirrus segments longer, the longest about twice as long as broad and the distal about as long as broad; there are 49 arms 150 mm. long and the cirri have 69-78 segments. Description.—The cirri are XX XV, 69-78, from 55 to 70 mm. long, much slenderer than those of O. finschii, with much longer distal segments, which are nearly or quite as long as broad, and also longer proximal segments, the longest of which are about twice as long as broad. Each IIBr series bears two IIIBr series of which the inner bears an internal [VBr series so that there are normally 10 arms to each ray arranged in 2, 3, 3, 2 order. The 49 arms are 150 mm. long. P, is 11 mm. long, with 20 or 21 segments, which become about as long as broad on the fifth or sixth, and twice as long as broad distally. P, is 12 mm. long, with 22 segments, resembling P;. P; is 10 mm. long, with 18 segments. P, is 7.5 mm. long, with 13 segments and with the distal taper more marked than in P;. P; is 7mm. long, with 13 segments, slightly slenderer than Py, especially in the distal portion. The distal pinnules are 8 mm. long, with 17 segments. Remarks.—This species is very closely related to Orymetra erinacea, from which it differs chiefly in having slenderer cirri with much longer distal segments, which are nearly or quite as long as broad instead of twice as broad as long or even broader as in O. erinacea, and also longer proximal segments, the longest of which are about twice as long as broad instead of only slightly, if at all, longer than broad. 402 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Locality.—Siboga station 320; between Borneo and eastern Java (lat. 6°05’00”’ S., long. 114°07/00’’ E.); 82 meters; fine gray mud; February 23, 1900 [A. H. Clark, 1912, 1918] (1, Amsterdam Mus.). History.—This species was first described by the present author under the name of Selenemetra tenuicirra in 1912 from a single specimen dredged by the Siboga at station 320. In the original description it was said to be closely related to Oxymetra finschii. In another paper published in the same year it was compared with 0. erinacea to which it was said to be most closely related. It was redescribed in detail and figured in 1918. OXYMETRA FINSCHT (Hartlaub) Prats 37, Ficure 168; Puatp 42, Ficures 191-193; Puan 44, Ficures 197-202; Prats 45, Figures 203, 204 [See also vol. 1, pt. 2, fig. 195 (lateral view), p. 122; figs. 476, 477 (pinnule tip), p. 269.] Antedon finschii Hartiaus, Nachr. Ges. Géttingen, May 1890, p. 176 (description; New Britain) ; Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 37 (in key), p. 47 (detailed description and comparisons), p. 113 (in Géttingen Mus.); pl. 3, fig. 32—Kornier, Mém. Soc. Zool. France, vol. 8, 1895, p. 419 (Sunda Islands, Biliton; notes).—Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1581 (listed).—A. H. Cuarx, Proc. Biol. Soc. Washing- ton, vol. 22, 1909, p. 13 (related to Pontiometra andersoni); Smithsonian Misc. Coll., vol. 60, No. 10, 1912, p. 22 (referred to Orymetra); Proc. U. S. Nat. Mus., vol. 43, 1912, p. 383 (identity) ; Crinoids of the Indian Ocean, 1912, p. 37 (same). Himerometra finschii A. H. Cuarx, Smithsonian Mise. Coll., vol. 50, 1907, p. 356 (listed). Himerometra gracilipes A. H. CuarKx, Smithsonian Misc. Coll., vol. 52, 1908, p. 219 (description; Albatross station 5163); Proc. U. S. Nat. Mus., vol. 39, 1911, p. 541, footnote (referred to Selene- metra); Unstalked crinoids of the Siboga-Exped., 1918, p. 99, footnote 1 (=Ozymetra finschit). Dichrometra aranea A. H. Cxiarx, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 76 (description; lat. 8°51'30’’ N., long. 81°11’52’’ E.; 28 fathoms). [Dichrometra] finschii A. H. Ciarx, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 76 (compared with D. aranea). Dichrometra gracilipes A. H. CLark, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 13 (listed). Selenemetra viridis A. H. Cuarx, Proc. U. 8S. Nat. Mus., vol. 39, 1911, p. 541 (description; Albatross Sta. 5413); Crinoids of the Indian Ocean, 1912, p. 139 (synonymy; locality); Unstalked crinoids of the Siboga-Exped., 1918, p. 99, footnote 2 (synonym of Oxymetra aranea). Selenemetra gracilipes A. H. Cuark, Proc. U. 8. Nat. Mus., vol. 39, 1911, p. 541 (calyx and arm bases compared with those of S. viridis); Crinoids of the Indian Ocean, 1912, p. 139 (synonymy; locality). Selenemetra finschit A. H. Cuarx, Bull. Mus. Hist. Nat., Paris, 1911, p. 253 (compared with Dichro- metra articulata); Ann. Mag. Nat. Hist., ser. 8, vol. 10, 1912, p. 832 (compared with S. tenuicirra) ; Smithsonian Misc. Coll., vol. 60, No. 10, 1912, p. 22 (comparison with Oxymetra erinacea; con- generic with that species; Orymetra has priority over Selenemetra so that this species becomes Ozymetra finschii); Proc. U. S. Nat. Mus., vol. 48, 1912, p. 382 (cotype from New Britain in U.S.N.M.), p. 383 (identity), p. 397 (New Britain); Crinoids of the Indian Ocean, 1912, p. 37 (identity), p. 1389 (synonymy; New Britain)—Hartmeyer, Mitt. zool. Mus. Berlin, vol. 8, Heft 2, 1916, p. 235 (Neu Pommern, Nos. 2602, 2603 [1 specimen] ). Selenemetra aranea A. H. Cuark, Crinoids of the Indian Ocean, 1912, p. 139 (synonymy; detailed description; locality), fig. 15, p. 140. Ozxymetra finschii A. H. Cuark, Unstalked crinoids of the Siboga Exped., 1918, p. 99 (in key; range).— Gist£n, Kungl. Fysiogr. Sillsk. Handl., new ser., vol. 45, No. 11, 1934, p. 25. Ozymetra aranea A. H. Cuarx, Unstalked crinoids of the Siboga Exped., 1918, p. 99 (in key; range).— Gist6n, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, p. 25. Oxymetra finshit Gistf{n, Kungl. Fysiogr. Sillsk. Handl., new ser., vol. 45, No. 11, 1934, p. 33. A MONOGRAPH OF THE EXISTING CRINOIDS 403 Diagnostic features.—The longest cirrus segments are but little longer than broad, and the distal cirrus segments are twice as broad as long or even broader; the proximal pinnules are less enlarged than are those of QO. erinacea and, though more or less stiffened basally and moderately elongated, they are flexible distally; there are 40-44 arms 90-140 mm. long; and the cirri have 41-80 segments. Description.—The centrodorsal is hemispherical and is more or less completely covered with cirrus sockets. The cirri are composed of 60-80 segments and may reach a length of 60 mm. In the proximal portion of the cirri the segments, except for a few of the basal ones, are usually slightly longer than broad; the remaining segments become gradually slightly shorter. Ata variable distance from the base of the cirri the distal dorsal edge of the segments begins to project and later develops a sharp spine that retains its distal position on the segment almost to the end of the cirri but here passes back more toward the middle. The radials are more or less completely, and sometimes wholly, visible. Their dorsal surface is approximately parallel with the dorsoventral axis of the animal. The IBr,; are incompletely united laterally; they are as long as the free sides of the rather large and pentagonal IBr, (axillaries). The postradial series as a rule divide three times. Individual IVBr series may be present. All the division series are 2. The IVBr series, if present, occur always on the inner side of the inner IIJBr axillaries. The articulations between the elements of the division series are tubercular. The post-radial series from the IBr, to the IIIBr axillaries often have their outer side more or less strongly produced into a shelflike border with sharp edges and a flattened outer side, but the thickening of the ventrolateral edges of the individual ossicles collectively forming this border may be discontinuous without any flattening. In other cases the production of the outer sides of the postradial series may be lacking altogether, their outer sides being entirely smooth. When their lateral edges are strongly produced the post-radial series are almost in lateral contact, otherwise they are well separated. There are 40, or a few more, arms 140 mm. long. The arms are long, smooth, and rather slender, with the dorsal surface strongly rounded. The lowest brachials may have flattened sides. The first brachials are entirely, or only partially, united interiorly. They are somewhat shorter than the second brachials, which are approxi- mately square. The first syzygial pair (composed of brachials 3++4) is markedly longer than broad. This is succeeded by six to nine more or less discoidal brachials, which bear a small lateral produced point, alternating in position on succeeding brachi- als, on their proximal borders. On their distal borders they sometimes bear a similar small sharp process. These are followed by a series of rather short triangular bra- chials, which before the middle of the arm pass over into bluntly wedge-shaped bra- chials. ‘Toward the end of the arms the brachials become more square. The triangu- lar and bluntly wedge-shaped brachials have on their distal edge a sharp production laterally placed and alternating in position on succeeding brachials. The first syzygy is between brachials 3+-4, the second is from the thirty-ninth to the forty-sixth brachial, and the distal intersyzygial interval is from 4 to 21, often from 5 to 10, muscular articulations. 404 BULLETIN 82, UNITED STATES NATIONAL MUSEUM The lower pinnules are equally long on all the arms, and on individual arms are as long on the inner as on the outer side. They are slender and gradually tapering. The pinnules of the first pair are mostly about as long as those of the second. These last are always somewhat thicker; they measure about 11 mm. in length and are composed of about 20 fairly equal segments, of which most are as a rule somewhat longer than broad. The following three or four pairs decrease in length and in the number of their component segments. The smallest pinnule is about 5 mm. long. The pinnules succeeding slowly increase in size, finally reaching a length of 10 mm. The disk is about 16 mm. in diameter and is strongly incised. The sacculi are closely crowded on the pinnules. The color is uniform gray-brown, sometimes with a touch of green, the cirri lighter, and the disk dark blackish gray. Notes.—The preceding description is adapted from Hartlaub’s original descrip- tion of the species. I have examined both of Hartlaub’s specimens, one of which is now in the United States National Museum. The specimen from Albatross station 5413 was described as a new species under the name of Selenemetra viridis as follows: The cirri are XX XIII, 47-57, from 30 to 35 mm. long, and rather slender. Sharp dorsal spines are developed from the tenth or twelfth segment—a more or less marked transition segment—onward. The calyx and arm bases resemble those of S. gracilipes (=Oxymetra finschii), but the division series and first ten brachials are sharply flattened laterally. The 40 arms are 90 mm. long. P, is slender, 10 mm. long, with 19 segments, which become twice as long as broad distally. P, and P; are similar, 12 mm. long, with 21 segments, slightly stiffer than P,; and proportionately stouter. P, is 9 mm. long, with 18 segments; it is as slender as P, but slightly stiffer. P; is 7 mm. long and resembles Py. Ps is 5.5 mm. long and resembles P; but is less stiffened. P, and the following pinnules are 5 mm. long; their component segments are proportionately slightly shorter than are those of the pinnules preceding. The distal pinnules are very slender and reach 7 mm. in length. The color is greenish yellow. The specimen from Albatross station 5163 was described as a new species under the name of Himerometra gracilipes in the following terms: The centrodorsal is a thick disk with a small flat polar area. The cirri are arranged in two and a partial third crowded, irregular, and more or less alternating rows. The cirri are XIX, 41-52 (usually about 50), 35 to 40 mm. in length. The first segment is about twice as broad as long, and those following gradually increase in length to the fifth or sixth which is about as long as broad. The succeeding segments remain the same, or become slightly longer than broad, to about the twenty-sixth, then gradually become shorter, about half again as broad aslong. After the twentieth to the twenty-second segment dorsal spines gradually begin to develop; these, however, never become very large. The opposing spine is terminal, erect, and about half as long as the width of the penultimate segment. The terminal claw is longer than the penultimate segment and is slender and moderately curved. The radials are very prominent and are about twice as broad as long; their dorsal surface is parallel with the dorsoventral axis of the animal. Each radial bears a low, rounded posterolateral tubercle on each side. The IBr, are trapezoidal, about three times as broad as long proximally and twice as broad as long distally; they ~ A MONOGRAPH OF THE EXISTING CRINOIDS A405 are basally united but diverge very rapidly distally. The IBr, (axillaries) are pen- tagonal, about half again as broad as long. The IIBr, I11Br, and IVBr series are 2 and are widely separated; the IVBr series are developed only on the outer side of the II[Br series. The 44 arms are 90 mm. long. The first ten brachials are oblong, not quite twice as broad as long, those succeeding are wedge-shaped, almost triangular, about twice as broad as long, becoming proportionately longer in the distal part of the arms. Syzygies occur between brachials 3-++4, again from between brachials 44-++45 to between brachials 56457 (usually nearer the latter), and distally at inter- vals of 6 to 11 (usually 7 to 9) muscular articulations. P,; is 10mm. long, slender but somewhat stiffened, tapering evenly from the base to the delicate tip, composed of 22 segments, of which the first is about twice as broad as long, the fourth is about as long as broad, and those following gradually increase in length so that the seventh and following are about half again as long as broad. P, is usually slightly longer than P, and is composed of 17 segments, of which those in the distal part are more elongated than the corresponding segments in P,. P; is 5.5 mm. long, with 12 segments, and is of the same character as the two preceding pinnules. The following pinnules are small, short, and delicate, 4 mm. long, with 12 segments, of which the first three are about as long as broad and the remainder are longer than broad, becoming about twice as long as broad distally. The distal pin- nules are 7 mm. long, with 20 segments of which the first is short, the second is about as long as broad, the third is about half again as long as broad, and the remainder are about twice as long as broad or rather longer. The color in alcohol is brownish ray. ; aa the specimen recorded by Koehler from Billiton the cirri reached a length of 60 mm. The arms were exactly 40 in number, 100 to 120 mm. long; each postradial series divided regularly three times. Professor Koehler said that this specimen conformed in all respects with Hartlaub’s description. The specimen from Investigator station 175 was described as a new species under the name of Dichrometra aranea. This new species was said to be nearest O. finschii, differing from that form in having fewer cirrus segments (53-66) and in having P, nearly or quite as long as P,, while the segments in the distal portion of the proximal pinnules are much elongated instead of subequal, slightly longer than broad, as in O. finschu. The lower pinnules are comparatively slender and are only slightly stiffened. P,is about asstout as P; and 2 mm. longer. Localities.—Albatross station 5413; between Cebu and Bohol, Philippines; Lauis Point light bearing N. 68° W., 10 miles distant (lat. 10°10/35’’ N., long. 124°03/15’’ E.); 77 meters; March 24, 1909 [A. H. Clark, 1911, 1912, 1918] (1, U.S. N. M., 27491). Albatross station 5163; Tawi Tawi group, Sulu (Jolo) archipelago; Observation Island bearing N. 79° W., 6.7 miles distant (lat. 4°59’10’” N., long. 119° 51’00’’ E.); 51 meters; coral sand; February 24, 1908 [A. H. Clark, 1908, 1909, 1911, 1912, 1918] (1, U. S. N. M., 25455). New Britain; Dr. Otto Finsch [Hartlaub, 1890, 1891; A. H. Clark, 1907, 1909, 1911, 1912, 1918; Hartmeyer, 1916] (2, U. S. N. M., 35268; Berl. M., 2602, 2603 [1 specimen)). Biliton, between southeastern Java and Borneo; M. Korotney, 1885 [Koehler, 1895]. 406 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Investigator station 175; off northeastern Ceylon (lat. 8°51’30’’ N., long 8191152” E.); 51 meters; April 20, 1894 [A. H. Clark, 1909, 1912, 1918] (1, U.S. N. M., 35241). Geographical range.—From the Philippines southward to New Britain and west- ward to northeastern Ceylon. Bathymetrical range.—From the shoreline down to 77 meters. History —This species was originally described in 1890 by Dr. Clemens Hartlaub as Antedon finschii from four specimens in the Berlin Museum, which had been collected in New Britain (Neu-Pommern) by Dr. Otto Finsch. It was redescribed in greater detail and figured in the following year. One of the original specimens was subsequently transferred to Géttingen and another to the United States National Museum. Hartlaub said that this fine new species is easily distinguished from all the other members of the Palmata group. In the first place the cirri, which are composed of 60-80 segments and may reach a length of 60 mm., are unusually long for a member of this group, and in the second place the strong flattened productions of the outer sides of the postradial series are highly characteristic. He noted that the development of the shelflike production of the sides of the postradial series is very variable, and on some rays it is wholly lacking. According to Hartlaub this species in these features very decidedly approaches the Spinifera group, but it cannot be included in that group because its pinnules have no ambulacral skeleton. In 1895 Prof. René Koehler recorded and gave notes on a specimen that had been collected by M. Korotnev in 1885 at Biliton. In my first revision of the old genus Antedon published in 1907, Antedon finschii was referred to the new genus Himerometra. In 1908 I described a new species, Himerometra gracilipes, from Albatross station 5163 in the Philippines. In a revision of the family Himerometridae published in 1909, I referred gracil- ipes to the new genus Dichrometra. I said that I had not been able to place Antedon Jinschii satisfactorily because of lack of material for comparison, but that it appears to be most closely related to Pontiometra andersoni, and the diagnosis of the genus Pontiometra may have to be altered for its reception. Later in 1909 I described Dichrometra aranea from a specimen from Investigator station 175. In 1911 I established the new genus Selenemetra with Antedon jfinschii as the genotype, at the same time describing Selenemetra viridis from Albatross station 5413. In another paper published in the same year I compared Selenemetra jinschii with Dichrometra articulata. In a paper on the crinoids of the Hamburg Museum published on November 7, 1912, I said that from the published description and figure of Antedon erinacea I had assumed that this form is allied to the species of Stephanometra with which it was associated by Hartlaub and, as it differed materially from all the forms I had grouped in that genus, I had suggested the generic name Ozymetra for it. Examination of the type specimen, however, showed that it had nothing to do with the species of Stephanometra but instead is closely related to the species I had grouped in the genus Selenemetra, especially to S. tenuicirra. The generic name Oxymetra with Antedon erinacea Hartlaub as the genotype has precedence over the generic name Selenemetra with Antedon finschii Hartlaub as the genotype, so that all the species heretofore A MONOGRAPH OF THE EXISTING CRINOIDS 407 assigned to the genus Selenemetra must be transferred to the genus Oxymetra, and the name Selenemetra must be relegated to the synonymy of Oxymetra. In a paper on the crinoids of the Berlin Museum published on November 20, 1912, I noted that I had examined the two cotypes of Selenemetra finschii in that museum. In my memoir on the crinoids of the Indian Ocean published in 1912, I listed Selenemetra finschii, S. aranea, S. gracilipes, and S. viridis as valid species and rede- scribed S. aranea. In the original description of S. tenuicirra, which was published in 1912, that form was compared with S. finschii. Dr. Robert Hartmeyer in 1916 published a note correcting the catalog numbers of the type specimen of Selenemetra finschii in the Berlin Museum. In my memoir on the unstalked crinoids of the Siboga expedition published in 1918 a key to the species of Oxymetra was given in which Oxymetra finschii and O. aranea were both admitted as valid species; O. finschii was said to be characterized by having the first four or five pairs of pinnules stiffened and elongated, P; about as long as P2, which is markedly longer than P;, and 60-80 cirrus segments of which those in the outer half of the cirri bear small dorsal spines, and 0. aranea was said to have the first six or seven pairs of pinnules stiffened and elongated, P; markedly shorter than P, and P;, which are of the same length and character, and 47-66 cirrus segments of which the outer three-fourths bear prominent dorsal spines. Genus STEPHANOMETRA A. H. Clark Antedon (part) Litken, Mus. Godeffroy Cat., vol. 5, 1874, p. 190 and following authors. Comatula (part) Smrrn, Ann. Mag. Nat. Hist., ser. 4, vol. 17, 1879, p. 406. Himeromeira (part) A. H. Cuarx, Smithsonian Misc. Coll., vol. 50, 1907, p. 356. Dichrometra (part) A. H. Cuark, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 13. Stephanometra A. H. Cuark, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 9 (diagnosis; genotype Antedon monacantha Hartlaub, 1890); Amer. Nat., vol. 48, 1909, p. 254 (represented in Red Sea); Proc. Biol. Soc. Washington, vol. 22, 1909, p. 176 (referred to the Stephanometrinae) ; Vid. Medd. Naturh. Foren. Kgbenhavn, 1909, p. 170 (confirmation of the original generic diagnosis); Proc. U. S. Nat. Mus., vol. 40, 1911, p. 10 (represented in West Indies by Analci- dometra); Notes Leyden Mus., vol. 33, 1911, p. 185 (lateral processes on elements of division series characteristic of this genus absent in Dichrometra flagellata); Mem. Australian Mus., vol. 4, 1911, p. 732 (in key), p. 735 (1 species in Australia), p. 769 (original reference; characters; range; represented in Caribbean Sea by Analcidometra); Rec. Indian Mus., vol. 7, 1912, p. 269 (ventrolateral processes on ossicles of IBr series and early brachials compared with similar structures in Oligometra intermedia); Ann. Mag. Nat. Hist., ser. 8, vol. 10, 1912, p. 35 (lateral processes on division series and lower brachials compared with those of Dichrometra tenuicirra) ; Notes Leyden Mus., vol. 34, 1912, p. 145 (union of pinnule segments compared with the same in Nanometra clymene); Proc. U. S. Nat. Mus., vol. 43, 1912, p. 401 (proximal pinnules compared with those of Tropiometra audouini); Crinoids of the Indian Ocean, 1912, p. 11 (Ceylon the western limit of the large and highly multibrachiate species); p. 12 (represented in southeast African region), p. 13 (corresponds to Analcidometra), p. 57 (in key), p. 132 (original reference; type); Unstalked crinoids of the Siboga-Exped., 1918, p. 93 (key to included species).—GISLEN, Zool. Bidrag Uppsala, vol. 9, 1924, p. 59, footnote (swimming movements), pp. 64, 66 (absence of muscular fossae on radial articular faces), p. 89 (articulation in IBr series), pp. 99, 100 (pin- nule articulations), p. 235. Lampromeira (part) H. L. Cuarx, Carnegie Inst. Washington Publ. 212, 1915, p. 104. Diagnosis.—A genus of Mariametridae in which one or more of the oral pinnules are enlarged, greatly stiffened, sharp-pointed, and spinelike; the division series are well separated, and the component ossicles bear rounded ventrolateral extensions; 408 BULLETIN 82, UNITED STATES NATIONAL MUSEUM the cirri are short or of moderate length, with less than 40 segments; and the dorsal surface of the division series is smooth, without ornamentation. Geographical range.-—From the Macclesfield Bank and the Philippine Islands to the Pelew, Caroline, and Marshall Islands, Samoa, Fiji, the Tonga Islands, New Caledonia, Torres Strait, and the Lesser Sunda Islands westward to the Red Sea and Madagascar. Bathymetrical range-—From the low-tide mark down to 245 meters. The species of Stephanometra are especially characteristic of the zone from the low-tide mark down to about 50 meters. Of the seven species, all have been taken in shore collecting, five are not known with certainty from a depth greater than 51 meters, one is recorded from 62 (possibly 73) meters, and one is recorded from 245 meters; there is some doubt about the correctness of the last record. Remarks.—The genus Stephanometra includes seven species, which are rather sharply divisible into two groups. In the species belonging to the first group the outer cirrus segments bear long, prominent, and sharp dorsal spines. These two species, S. echinus and S. tenuipinna, are alike in all features except such as might well be attributed to the relative maturity of the individuals assigned to one or the other. Although it is by no means certain that S. tenuipinna is simply the young of S. echinus, this will very likely turn out to be the case. In the five species belonging to the second group the distal cirrus segments are wholly without dorsal spines, being at the most only slightly carinate dorsally. In this group the species are differentiated by the characters presented by the proximal pinnules. In all of them P,is enlarged, much stiffened, and spinelike. In one species, S. spinipinna, P; is stiffened and spinelike, though considerably slenderer and shorter than P,, being intermediate in character between the stout, much stiffened, and spinelike P; found in S. echinus and S. tenuipinna and the slender and flexible P, found in S. oxryacantha, S. spicata, S. protectus, and S. indica. When the characteristic P,; of S. spinipinna is typically developed the species is an easy one to recognize, but in this species P; is rather variable, and there is a suggestion of possible intergradation between S. spinipinna on the one hand and S. spicata and S. oryacantha on the other. The four species in which P, is slender, delicate, and flexible are S. oryacantha, S. spica- ta, S. protectus, and S. indica. In S. oxyacantha the second-fifth or second-fourth pinnules are stiffened and spinelike, in S. spicata the second and third, and in S. protectus and S. indica only the second. But there is really no hard and fast line of division between these forms. Pinnules of an intermediate type more or less enlarged and much stiffened for a greater or lesser distance from their base, but with the distal portion delicate and flexible, sometimes occur. As a result of the occurrence of such pinnules some individuals are intermediate between S. oryacantha and S. spicata, while others are intermediate between S. spicata and S. protectus. Although S. protectus and S. indica when typically developed appear very different, some individuals from Ceylon have P, of a type intermediate between that characteristic of S. protectus and that characteristic of S. indica. The species of the first group are somewhat limited in their distribution, occurring from the Philippines to New Britain, the Paternoster Islands, and Ceylon. The A MONOGRAPH OF THE EXISTING CRINOIDS 409 species of the second group range far to the eastward into Polynesia, southward to northern Australia, and westward to east Africa. The species of Stephanometra are all inhabitants of shallow water and are most numerous in a narrow band just below the low tide mark. All the species are of medium or rather large size. History.—The genus Stephanometra was established by me in 1909, with the genotype Antedon monacantha Hartlaub, 1890. The included forms had previously been placed in the genus Himerometra established in 1907. There has been no change in the status of the genus since its first description. KEY TO THE SPECIES AND SUBSPECIES OF THE GENUS STEPHANOMETRA a!, Outer cirrus segments with long and prominent dorsal spines. b!. 33-40 (usually 30-35) arms 110-170 mm. long; cirri with 25-37 segments (Philippines south- ward to Paternoster Islands and westward to “India” [possibly Ceylon]; 0-38 meters)eoseen ess Loli ak SUE 2 EE ae EE es SER ae ee, echinus (p. 409) b?. 16-24 arms 60-70 mm. long; cirri with 20 segments (Philippines to New Britain, and westward to Geylon;,0=48)} meters): lseu:!o- 222 22h Leereieiee ee [ote et ere tenuipinna (p. 413) a?. Outer cirrus segments without dorsal spines, being merely slightly carinate dorsally. b!, P, stiffened and spinelike, resembling P, but somewhat smaller; 12-32 arms 35-150 mm. long; cirri with 15-23 segments (Borneo southward to the Moluccas and Kei Islands; 0-245 NNIGCCTS) saeco te ee ee ee Se ee eee spinipinna (p. 415) b?, P, flexible, flagellate, delicate, and slender, much smaller and weaker than the enlarged and stiffened P». cl, P, the only pinnule enlarged and stiffened. d', P, much enlarged and stiffened, usually more or less straight, distally ending somewhat abruptly in a strong stout point, composed of 11-15 (usually 12) segments, of which the fourth and fifth are typically abruptly longer than the others (Macclesfield Bank and Philippines to Caroline and Marshall Islands, Samoa, Fiji, Tonga Islands, New Caledonia, and Torres Strait, and westward to Ceylon and Maldive Islands; 0-51 MeLeES) 9 a ek oe ae eee ed indica protectus (p. 443) d?. P; somewhat less enlarged and stiffened, usually more or less strongly recurved, distally becoming very slender and delicate, though not flagellate, composed of 15-20 segments of which the fourth and fifth are not noticeably different from the rest (Madagascar, Seychelles, and Red Sea to Ceylon; Torres Strait; 0-62 [?73] meters)_indica indica (p. 436) c?. P3; and sometimes one or more of following pinnules enlarged and stiffened like P2, though usually shorter. d!, P, and P; the only pinnules enlarged, stiffened, and spinelike, P, being short and flexible like the succeeding pinnules (Macclesfield Bank and Philippines to Pelew and Caro- line Islands, New Guinea, the Admiralty and Solomon Islands, Fiji, New Caledonia, and Torres Strait, and westward to Red Sea; 0-38 [?64] meters) --____- spicata (p. 424) d?, P,, or Py and P;, stiffened and spinelike, resembling P; and P; though smaller (Philippines to Solomon Islands and Moluccas, and westward to Java; 0-16 [?23] meters) oxyacantha (p. 418) STEPHANOMETRA ECHINUS (A. H. Clark) Puate 45, Ficures 205-207; Piate 46, Ficures 210, 211; Prats 47, Ficures 212-216 [See also vol. 1, pt. 2, fig. 125 (arm base), p. 79; fig. 273 (arm and pinnules), p. 207; figs. 467, 468 (pinnule tip), p. 266.] Himerometra echinus A. H. Cuark, Smithsonian Misc. Coll., vol. 52, 1908, p. 218 (description; Albatross station 5147). 208244—40——27 410 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Stephanometra coronata A. H. Criark, Proc. U. 8. Nat. Mus., vol. 36, 1909, p. 639 (description; “India’”’); vol. 39, 1911, p. 541 (Port Busin, Barias Island; “India’”); Smithsonian Misc. Coll., vol. 60, No. 10, 1912, p. 19 (synonym of echinus); Crinoids of the Indian Ocean, 1912, p. 133 (synonymy; detailed description; localities), fig. 13, a, b, p. 1338; Unstalked crinoids of the Siboga Exped., 1918, p. 93, footnote ! (synonym of echinus). Stephanometra echinus A. H. Cxiarx, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 10 (listed) ; Smithsonian Misc. Coll., vol. 60, No. 10, 1912, p. 19 (includes S. coronata; no locality; fragments) ; Crinoids of the Indian Ocean, 1912, p. 182 (synonymy; locality); Unstalked crinoids of the Siboga-Exped., 1918, p. 93 (in key; range), p. 94 (synonymy; notes; stations 40, 99), pp. 271, 272 (listed); Smithsonian Misc. Coll., vol. 72, No. 7, 1921, pl. 15, fig. 52 (arm).—Gisi&n, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, p. 25.—Gistf£n, Kungl. Fysiogr. Sallsk. Lund Forh. vol. 7, No. 1, 1936, p. 4 (French Indochina), p. 5 (range), p. 11 (Pulo Condor; notes), figs. 2, 2a, p. 9. Diagnostic features.—The outer cirrus segments have long, sharp, and prominent dorsal spines; there are 30-40 (usually 30-35) arms 110-170 mm. long; and the cirri have 25-37 segments. Description —The centrodorsal is discoidal, with the moderately large bare polar area slightly concave; the cirri are arranged in two closely crowded more or less alter- nating rows. The cirri are XXIV, 26-30, 25 mm. long. The first four segments are about twice as broad as long, the sixth is about as long as broad, and the seventh-tenth or -eleyenth are about one-third again as long as broad; the segments succeeding are again about as long as broad, and those following gradually become shorter so that the terminal 12 or 14 are broader than long. The tenth and following segments bear large dorsal spines. The opposing spine is terminally situated, erect, and reaches about one-half the width of the penultimate segment in height. The terminal claw is rather longer than the penultimate segment and is slender and moderately curved. The radials project slightly beyond the rim of the centrodorsal. The IBr, are proximally about four and distally about three times as broad as long; they are united in their basal third, but their sides diverge very rapidly from the point of union so that the free lateral borders of two adjacent IBr, form together an even and mod- erately curved line. The IBr2 (axillaries) are broadly pentagonal, nearly twice as broad as long, with large and broad ventrolateral projections. The IIBr, IIIBr, and IVBr series are 2, bearing on the outer side, in common with the first brachials, large and broad ventrolateral processes. The 40 arms are 110 mm. long. One of the internal IIIBr series is missing, but its loss is compensated by the development of an external IVBr series on the same I1Br series. The first 9 or 10 brachials are oblong or very slightly wedge-shaped, about twice as broad as long, those following becoming short-triangular, rather more than twice as broad as long, and short-wedge-shaped in the distal portion of the arms. Syzygies occur between brachials 3+-4, again from between brachials 22+23 to between brachials 32+-33 (most frequently in the vicinity of the twenty-third bra- chial), and distally at intervals of from 9 to 24 (usually 9-13) muscular articulations. P, is large, stiff, and spinelike, resembling P2, with 15 segments and reaching 15 mm. in length. The first two segments are nearly twice as broad as long, the third is about as long as broad, and those succeeding increase in length so that the seventh and following are from one and one-half times to nearly twice as long as broad. P, is 16 mm. long with 12 or 13 segments, of which the distal are rather A MONOGRAPH OF THE EXISTING CRINOIDS 41] longer than the distal segments of P;. P; is 15 mm. long, resembling P,. P, is 12 mm. long, with 11 segments, resembling P;. The pinnules following decrease in length and also slightly in stoutness, P; being 7 mm. long, with 10 segments, then gradually lose their peculiar stiffness, and later gradually increase in length. The distal pinnules are 10 mm. long, with about 20 segments, of which the first two are not so long as broad, the third is about as long as broad, and the remainder are about one-third again as long as broad, becoming half again as long as broad distally. The color in alcohol is dull yellowish, with the division series and arms thickly covered with small red spots and blotches; the ventral perisome is brown. Notes.—The specimen from Barias Island is a fine example of the species closely resembling the type specimen of Stephanometra coronata in the Indian Museum at Calcutta, with which it was directly compared. It has 34 arms about 150 mm. long; four of the postradial series bear seven arms each, and one bears six. As in the type specimen of coronata, the division series omitted are always internal. The cirri are XVIII, 27-29, from 26 to 28 mm. long. The color is dark purple. One of the specimens from Port Galera, Mindoro, has about 30 arms, which are about 170 mm. long; the longest cirri are 30 mm. long and are composed of 34-37 segments. Another has 31 arms 160 mm. long. A third has 33 arms 155 mm. long. A fourth has about 35 arms 150 mm. long. Two others have 38 and 36 arms. The specimen from Siboga station 99 is small with 29 arms; in color it is a uniform dull light yellow. In the specimen from Pulo Condor as described by Gislén the diameter of the centrodorsal is 6.5 mm. and the cirri are XXXIV, 31-32, from 26 to 28 mm. long. The four basal segments are short, the fifth is as long as broad, the eighth is half again as long as broad, and the distal are one-third again as broad as long. From the thirteenth onward there is a triangular dorsal spine, projecting in the distal half of the segment. The opposing spine is considerably larger than the spine on the segment preceding. The terminal claw is curved, and is a little longer than the seg- ment that bears it. There are 6-8 arms on each postradial series, or about 35 in all. They are 115+ mm. long. The arm bases are widely separated but bear ventro- lateral flanges which approach the flanges from the proximal portions of the adjacent postradial series. The synarthrial tubercles are rather inconsiderable, and are not sharply set off. P, is 14 mm. long, with 12 segments. P, is 18 mm. long, with 14 segments. P,; is 13 mm. long, with 12 segments. The pinnules following decrease in length. All the proximal pinnules are similar, with smooth and very long cylindrical segments, of which the longest are between three and four times as long as broad. None of them are specially stiffened, but because of the elongated segments they are rather inflexible. The distal pinnules are 10.5 mm. long, with 28 segments. The color is dark brownish violet. The specimen from Siboga station 40 has 28 arms, which are up to about 85 mm. in length. The cirri are XVI, 22-24, 19 mm. long; the eighth or ninth is a transition segment. P,is 11mm. long, with 12 segments. P, is 12 mm. long, with 11 segments. P; is 9 mm. long, with 9 segments. P,is 7 mm. long, with 9 segments. P; is 5 mm. long, with 9 segments and is slightly stiffened. The first four pinnules are all stiff and 412 BULLETIN 82, UNITED STATES NATIONAL MUSEUM spinelike and are enlarged in proportion to their length. The color is dull whitish with narrow bands of yellow brown on the arms and cirri. The description given for the species is based upon the type specimen from Albatross station 5147. The type specimen of Stephanometra coronata from ‘India’ may be described as follows: The cirri are XXII—XXIIT, 25-30, 22 mm. long, and resemble those of S. tenuipinna; the longest segments are about one third again as long as broad; the ninth, tenth, or eleventh is a well-marked transition segment. The radials project slightly beyond the edge of the centrodorsal. The IBr, are oblong, short, about three and one-half or four times as broad as long, with a rounded ventrolateral process in the proximal half, and are not in lateral apposition. The TBr, (axillaries) are broadly pentagonal, twice as broad as long, with the lateral edges about half as long as those of the [Br, and produced into prominent rounded ventro- lateral processes. The IIBr series, [J Br series, and JVBr series, when present, are 2, resembling the IBr series. All the ossicles of the division series and the first brachials have prominent ventrolateral processes; the synarthrial tubercles are rather prom- inent. The 33-34 arms are 120 mm. long, in general resembling those of S. tenwipinna. P, is long, stout, stiff, and spinelike, composed of 14 segments, of which the first two are somewhat broader than long, the third—-fifth are about as long as broad, and the remainder become gradually elongated and twice as long as broad distally. P, and P, are exactly like P;. P,is 10 mm. long, with 10 segments, resembling the pinnules preceding. P; is 7 mm. long, spinelike as are the preceding but somewhat slenderer, with 8 segments. The following pinnules decrease gradually in length, Ps being 5 mm. long, with 8 segments. The succeeding pinnules remain of similar length but decrease in stiffness and increase in the number of segments, P,; being 5 mm. long, with 12 segments, which become as long as broad on the third and twice as long as broad distally and only stiffened proximally. The distal pinnules are slender, 9 mm. long. Remarks.—It is possible that this form will eventually prove to be simply full sized S. tenuipinna, but until definite evidence of their identity is available it seems best to keep them separate. Localities—Albatross; Port Busin, Barias Island, off the entrance to Ragay Gulf, southeastern Luzon, Philippines; shore; March 8, 1909 [A. H. Clark, 1911, 1912, 1918] (1, U.S.N.M., 35095). Port Galera, Mindoro, Philippines; Lawrence E. Griffin (8, M. C. Z., 629, 684, 685, 686, 687). Siboga station 99; anchorage off North Ubian (lat. 6° 07’ 30’” N., long. 120° 26’ 00”’ E.); 16-23 meters; lithothamnion bottom; June 28-30, 1899 [A. H. Clark, 1918] (1, Amsterdam Mus.). Albatross station 5147; Sulu (Jolo) archipelago, in the vicinity of Siasi; Sulade Island (E.) bearing N. 3° E., 8.4 miles distant (lat. 5° 41’ 40’’ N., long. 120° 47’ 10’ E.); 38 meters; coral sand and shells; February 16, 1908 [A. H. Clark, 1908, 1909, 1912, 1918, 1921} (1, U.S.N.M., 25442). Pulo Condor, Cochin China; littoral; Dr. C. Dawydoff [Gislén, 1936]. Siboga station 40; anchorage off Pulu Kawassang, Paternoster Islands; coral reef; 12 meters; April 2, 1899 [A. H. Clark, 1918] (1, Amsterdam Mus.). A MONOGRAPH OF THE EXISTING CRINOIDS 413 “India.” [A. H. Clark, 1909, 1911, 1912, 1918] (2, U.S.N.M., 35242; I. M.). No locality [A. H. Clark, 1912] (fragments, H. M.). Geographical range—From southern Luzon, Philippines, southward to the Pater- noster Islands (north of Sumbava) and westward to “India” (possibly Ceylon). Bathymetrical range-—From the shoreline down to 38 meters. History.—This species was first described under the name of Himerometra echinus from a specimen dredged by the Albatross at station 5147 in 1908. In the year follow- ing another specimen from the Investigator collection labeled “India” was described under the name of Stephanometra coronata. In 1911 a specimen of Stephanometra coronata was recorded that had been taken by the Albatross at Port Busin on Barias Island, and it was noted that this specimen closely resembled the type (with which it had been directly compared) in the Indian Museum at Calcutta. In a paper on the crinoids of the Hamburg Museum published in 1912, some arm fragments without locality were recorded, and coronata was placed in the synonymy of Stephanometra echinus. In my memoir on the crinoids of the Indian Ocean published in 1912, I rede- scribed the type specimen of Stephanometra coronata in detail and figured it and also listed S. echinus from the Philippines in 21 fathoms. In the report on the crinoids of the Siboga expedition published in 1918, I recorded Stephanometra echinus from stations 40 and 99 and gave notes on the specimens. In this report coronata was placed as a synonym under echinus. In 1936 Dr. Torsten Gislén recorded and gave notes on a specimen from Pulo Condor, Cochinchina. STEPHANOMETRA TENUIPINNA (Hartlaub) PuaTtEe 45, Figures 208, 209 [See also vol. 1, pt. 2, figs. 465, 466 (pinnule tip), p. 266.] Antedon tenuipinna Hartuaus, Nachr. Ges, Gottingen, May 1890, p. 178 (description; Matupi, New Britain); Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 38 (in key), p. 54 (Matupi, New Britain; detailed description and comparisons), p. 58 (in first section of the Palmata group with spiny lower pinnules), p. 113 (in Géttingen Mus.), pl. 3, figs. 28, 30, 34—Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1581 (listed) —A. H. Cuark, Proc. U. S. Nat. Mus., vol. 43, 1912, p. 383 (identity); Crinoids of the Indian Ocean, 1912, p. 37 (identity). Himerometra tenuipinna A. H. Cuarx, Smithsonian Misc. Coll., vol. 50, 1907, p. 356 (listed); vol. 52, 1908, p. 219 (compared with H. echinus). Stephanometra tenuipinna A. H. Cuarx, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 10 (listed); Proc. U. S. Nat. Mus., vol. 36, 1909, p. 399 (Albatross station 5174), p. 639 (comparison with S. coronata); Vid. Medd. Naturh. Foren. Kgbenhavn, 1909, p. 170 (Singapore; description of a specimen), p. 193 (collected at Singapore by Svend Gad); Proc. U. S. Nat. Mus., vol. 48, 1912, p. 383 (identity), p. 397 (New Britain); Crinoids of the Indian Ocean, 1912, p. 37 (identity), p. 135 (synonymy; off Colombo Light; description; summary of previous records).—H. L. Ciarx, Spolia Zeylanica, vol. 10, pt. 37, 1915, p. 93 (occurs at Ceylon).—A. H. Cuarx, Unstalked crinoids of the Siboga-Exped., 1918, p. 93 (in key; range). Stephanometra tenuispina Gisuin, Kungl. Fysiogr. Sillsk. Handl., new ser., vol. 45, No. 11, 1934, p: 20. 414 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Diagnostic features.—The outer cirrus segments have long, sharp, and prominent dorsal spines; there are 16-24 arms 60-70 mm. long; and the cirri have 20 segments. Description.—The centrodorsal is a convex disk. The cirri are about XV, about 20, up to 12 mm. in length, and slender. The fifth- eighth segments are rather long (shown in the figure as about twice as long as broad) and those following gradually become shorter, toward the end of the cirri being about as long as broad. From the ninth onward the segments bear a sharp erect median dorsal spine. In addition, each end of the midline of the dorsal surface is produced into a spinelike process that lies parallel with the dorsal surface; of these horizontal spinelike processes the distal is the larger and overlaps the proximal end of the following segment. The radials are visible, and are partially united laterally. The IBr, are entirely free laterally, and are somewhat shorter than the pentagonal axillaries. Of the five postradial series, two have 2 divisions, one has 3, and two have only 1. The IIBr and IIIBr series are 2. The IIBr, are only partially united with their neighbors. The division series are widely separated laterally. Their lateral edges have no thick- enings but are notched. There are no synarthrial tubercles. The 16 smooth and slender arms are 62.5 mm. long and are composed of rather long brachials. The first brachials are somewhat longer exteriorly than interiorly and are only partially united interiorly. The second brachials have approximately the same form, with a slight posteriorly directed elevation of the middle of the prox- imal border. The first syzygial pair (composed of brachials 3+-4) is slightly longer than broad. The next few brachials are discoidal, and those following are moderately long, short wedge-shaped, distally becoming elongated. The syzygial pairs are long. Syzygies occur between brachials 3+-4 and 16+17 (or in arms arising directly from the [Br axillaries between brachials 9+10), and distally at intervals of 5-8 muscular articulations. In arms arising from a [Br axillary the first two pairs of pinnules (P; and P, and P; and P,) are equally long and stout, 7 mm. in length. They are stiff and spine- like and are composed of about eight segments, which with the exception of the two basal and of the terminal are much longer than broad. The pinnules of the pair following (P; and P,) are of the same character but are only 4 mm. long, with five segments. The pinnules of the next pair (P, and P,) are the shortest, reaching only 3 mm. in length, and are less stiffened. The succeeding pinnules are very slender, almost hairlike, and become about 7 mm. long. The lower pinnules of the other arms differ in that they are a good deal longer on the outer than on the inner side of the arms. The disk is 11 mm. in diameter and is only slightly incised. Sacculi are not numerous but are deeply colored and conspicuous; they occur on the pinnules, on the arms, and on the disk. The color is dorsally light yellowish white; the disk is light brownish gray with a tinge of green. Notes.—The specimen from Singapore has 24 arms 70 mm. long; the five post- radial series bear 8, 2, 2, 7, and 5 arms; in the last one of the arms arises from the IBr axillary, the other branch from the axillary consisting of one IIBr and two IIIBr series, the latter bearing 4 arms. This specimen agrees very well with Hartlaub’s A MONOGRAPH OF THE EXISTING CRINOIDS 415 description, except that P; is nearly as long as P2, and P, is as he describes P;. The division series have slight lateral projections. The specimen from off Colombo Lighthouse has 21 arms 50 mm. long. It agrees well with Hartlaub’s description, and also with the specimen from Singapore just mentioned. P, has 12 segments. Localities.—Albatross station 5174; in the vicinity of Jolo (Sulu), Philippines; Jolo light bearing E., 2.6 miles distant (lat. 6°03/45’’ N., long. 120°57’00’’ E.); 36 meters; coarse sand; March 5, 1908 [A. H. Clark, 1909, 1912, 1918] (1, U.S.N.M., 35256). Matupi, New Britain; Dr. O. Finsch [Hartlaub, 1890, 1891; A. H. Clark, 1907, 1909, 1912, 1918] (1, Berl M., 5374). Singapore; Svend Gad [A. H. Clark, 1909, 1912, 1918] (1, C.M.). Investigator station 152; 11.5 miles S. 83° W. of Colombo Light, Ceylon; 48 meters; sand, shells, and coral; December 12, 1893 [A. H. Clark, 1912; H. L. Clark, 1915] (1, I.M.). Geographical range.—From the Philippine Islands to New Britain and westward to Ceylon. Bathymetrical range.—From the shoreline down to 48 meters. History.—This species was first described by Dr. Clemens Hartlaub in 1890 under the name of Antedon tenuipinna from a specimen in the Berlin Museum collected at Matupi, New Britain, by Dr. Otto Finsch. It was redescribed by him and figured in 1891. In 1907 I placed it in my new genus Himerometra, transferring it in 1909 to the new genus Stephanometra. In 1909 I recorded another specimen from Albatross station 5174 in the Philip- pines, and later in the same year I recorded a third from Singapore collected by Svend Gad, the Danish consul at that port. The fourth specimen, dredged by the Investigator off Colombo, Ceylon, was recorded in 1912. In 1912 I mentioned having examined the type specimen in the Berlin Museum, and in 1915 Dr. Hubert Lyman Clark included the species in his list of the echinoderms of Ceylon, other than holothurians, on the strength of the Investigator specimen from off Colombo. STEPHANOMETRA SPINIPINNA (Hartlaub) Antedon spinipinna Hartuaus, Nachr. Ges. Gottingen, May 1890, p. 179 (description; Amboina); Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 11 (collected by Brock at Amboina), p. 39 (in key), p. 58 (in second section of species of Palmata group with spiny lower pinnules), p. 61 (Amboina; detailed description and comparisons), p. 113 (in Géttingen Mus.), pl. 4, figs. 42, 44.—Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1581 (listed). —A. H. Criarx, Crinoids of the Indian Ocean, 1912, p. 37 (identity). Himerometra spinipinna A. H. Cuark, Smithsonian Mise. Coll., vol. 50, 1907, p. 356 (listed). Stephanometra spinipinna A. H. Cuark, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 10 (listed); Crinoids of the Indian Ocean, 1912, p. 37 (identity), p. 132 (synonymy; Amboina); Unstalked crinoids of the Siboga-Exped., 1918, p. 93 (in key; range), p. 95 (references; notes; stations 81, 89), p. 272 (listed)—Guist£N, Kungl. Fysiogr. Saéllsk. Handl., new ser., vol. 45, No. 11, 1934, p. 20. 416 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Diagnostic features—The outer cirrus segments are wholly without dorsal processes, being merely more or less sharply carinate in the middorsal line; P, is stiffened and spinelike, resembling P; though smaller and slenderer; there are 12-32 arms 35-150 mm. long; and the cirri have 15-23 segments. Description.—The centrodorsal is convex, with the bare dorsal pole small and pitted; the cirri are arranged in two irregular marginal rows. The cirri are XV—XX, about 15, slender and moderately compressed, with smooth segments, of which the fourth—ninth are elongated. The radials are visible. The IBr;, which are as long as the free sides of the axillaries, are not in lateral contact. The IBr, (axillaries) are pentagonal. Four of the postradial series divide only once, but the fifth—which is markedly smaller than the others, perhaps as a result of regeneration—divides twice. The I1Br series are 2. The ventral edges of the division series are smooth, and synarthrial tubercles are not developed. There are 12 smooth arms 35 mm. long. The first brachial is rather short and is only partially united with its fellow. The second brachial is a little longer. The first syzygial pair (composed of brachials 3+-4) is longer than broad. From the seventh onward the brachials are larger and wedge-shaped, finally becoming more oblong. The first syzygy is between brachials 3-++-4, the second is between brachials 94-10, and the distal intersyzygial interval is three or four muscular articulations. P,; on arms arising from a [Br axillary is slender, but rather stiff and styliform, and consists of about 12, or at the most 14, elongated segments. P, is markedly stouter and a good deal longer; it is quite stiff and spinelike, 6 mm. long, and consists of 8-10 very long segments. P; and P, decrease in length and are less stiff. P, is markedly smaller than P,; and is not styliform. P, is about as long as P;. The outer pinnules are moderately long, reaching 7 mm. in length. The disk is 7 mm. in diameter and is not incised. On the pinnules the sacculi are large and closely set. The color in alcohol is light yellowish brown, with dark cross bands at the articu- lations. The disk is gray-brown. Notes.—The preceding description is taken from Hartlaub’s redescription of the type specimen from Amboina. The specimen from Siboga station 81 has 31 arms about 90 mm. long. The cirri are XXIV, 22-23, 20 mm. long. P, is 10 mm. long, with 18 segments. P, is 12 mm. long, with 12 segments. P,; is 10 mm. long, with 11 segments. P,, which is of the same character as the pinnules preceding, is 8 mm. long, with 10 segments. The example from Siboga station 89 has 25 arms about 80 mm. long. The cirri are XXIII, 22-23, from 15 to 20 mm. long. P, is 11 mm. long, with 20 segments, of which the first two are about as long as broad, the third is longer than broad, and the remainder are much elongated; the pinnule is slender but very stiff. P, is 11 to 12 mm. long, stouter than P,, with 13 segments. P; is 9 mm. long, with 10 segments. P,, which resembles the preceding pinnules, is 5 mm. long, with 9 segments. One of the specimens collected by the Danish Expedition to the Kei Islands at Vatek van Toeal has the cirri XXI, 23, from 20 to 23 mm. long. The first segment A MONOGRAPH OF THE EXISTING CRINOIDS 417 is very short, the second is about twice as broad as long, the third is slightly longer, the fourth is nearly or quite as long as broad, and those succeeding are longer than broad, the longest (eighth-twelfth) about half again as long as broad, with the terminal only slightly longer than broad. The last 10 or 11 segments have a narrow and very low median dorsal carination. There are 32 arms 140 mm. long. When present IIIBr series are developed on the outer side of the postradial series, occasionally also on the inner side of the IIBr series. P, is 11 mm. long with 21 segments, tapering evenly from the base to the tip, somewhat stiffened but not spinelike. The first two segments are broader than long, the third is slightly broader than long, the fourth is longer than broad, the sixth is about twice as long as broad, and the distal are about three times as long as broad. P, is 15 mm. long, much stouter than P,, stiff and spinelike, with 14 segments. The first two segments are nearly half again as broad as long, the third is about one-third again as long as broad, and the remainder are mostly about three times as long as broad, except for the terminal, which is minute and conical. P; is 12.5 mm. long, with 12 or 13 segments, similar to P,; but proportionately smaller. P, is 7.5 mm. long, spinelike, but smaller at the base and tapering more rapidly than those preceding, and more slender distally; it is composed of 12 segments of which the outer are very greatly elongated. P,; is 6 mm. long, with 12 segments, more slender than P, and less stiffened. The next two pinnules resemble Ps, and those following are more slender and more flexible. The distal pinnules are 9.5 mm. long, with 20 segments, very slender, with the third and following segments between three and four times as long as broad. The second specimen from Vatel van Toeal has 30 arms 150 mm. long; each postradial series bears an external IIIBr series on each side. The cirri are XXII, 21-23, from 20 to 25mm. long. The color of both specimens is deep purple, more or less marked with lighter or darker. In the larger specimen from the Danish Expedition to the Kei Islands station 3 the centrodorsal is discoidal, thin, with the flat dorsal pole 2 mm. in diameter. The cirrus sockets are arranged in two closely crowded and irregular marginal rows. The cirri are X XIX, 15-17, from 14 to 17 mm. in length. The first segment is three times as broad as long, the second is twice as broad as long, the third is from one-third to one-half again as long as broad, the fourth is nearly or quite twice as long as the median width, and those following slowly decrease in length so that the outermost are about as long as broad, very slightly longer than broad, or, more rarely, slightly broader than long. The segments up to and including the proximal two-thirds of the sixth have a dull surface, the distal third of the sixth segment and the segments fol- lowing having a highly polished surface. The dorsal profile of the segments following the sixth is moderately convex, the maximum height of the convexity being near the distal end, and their dorsal surface is sharply carinate in the median line. The opposing spine is prominent, conical, arising from the entire dorsal surface of the penultimate segment, with the apex terminal or subterminal; its height is equal to about half the width of the penultimate segment. The terminal claw is slightly longer than the penultimate segment and is moderately and regularly curved. The radials are visible as a narrow line above the rim of the centrodorsal which is wider inter- radially than elsewhere and is usually interrupted in the midradial line. The IBr, are extremely short and bandlike, very slightly longer in the midradial line than 418 BULLETIN 82, UNITED STATES NATIONAL MUSEUM laterally, 10 to 12 times as broad as the median length. The very short lateral borders are strongly convergent, making approximately a right angle with each other. The IBr, (axillaries) are triangular, nearly or quite twice as broad as long. The very short lateral sides converge strongly proximally to the anterolateral angles of the IBr,. The anterior edges are rather strongly concave. The proximal portion rises to a prominent though broadly rounded synarthrial tubercle. When the animal is viewed in a plane at right angles to the dorsoventral axis the radial portions of the centrodorsal are seen to have deep V-shaped incisions which make an angle somewhat greater than a right angle accomodating the IBr,. The IIBr series are 2. The division series and arm bases are rather widely separated laterally. The 14 arms are about 70 mm. long. The other specimen from station 3 is an exceedingly small 10-armed individual. Remarks.—In describing Antedon spinipinna Hartlaub said that the specimen upon which the description was based is not sexually mature, but nevertheless he felt justified in considering it as a new species because it is so markedly different from the related forms that it is difficult to interpret as a young stage of any one of them. He said that spinipinna is a small and very slenderly built form. P, is inter- mediate in character between P; in S. tenuipinna on the one hand and S. spicata and S. oxyacantha on the other. It is rather stiff and styliform and is composed of about 12 elongated segments, but as in oryacantha and spicata it is markedly shorter than P,, which has only 8-10 segments. Localities —Siboga station 81; Pulu Sebangkatan, Borneo Bank; 34 meters; coral bottom and lithothamnion; June 14, 1899 [A. H. Clark, 1918] (1, Amsterdam Mus.). Siboga station 89; Pulu Kaniungan ketjil, between northwestern Celebes and Borneo; 11 meters; coral; June 21, 1899 [A. H. Clark, 1918] (1, Amsterdam Mus.). Amboina [Hartlaub, 1890, 1891; A. H. Clark, 1907, 1909, 1912, 1918]. Danish Expedition to the Kei Islands; Dr. Th. Mortensen; Vatek van Toeal; about 2 meters; March 23, 1922 (2). Danish Expedition to the Kei Islands; Dr. Th. Mortensen; station 3; 245 meters; March 31, 1922 (2). Geographical range-—From Borneo southward to the Moluccas and the Kei Islands. Bathymetrical range-—From the shoreline down to 245 meters. History.—This species was originally described as Antedon spinipinna by Dr. Clemens Hartlaub in 1890 from a single specimen that had been collected by Dr. J. Brock at Amboina. It was redescribed and figured by him in 1891. I placed it in the new genus Himerometra in 1907 and removed it to Stephanometra in 1909. In 1918 I recorded and gave notes upon two specimens that had been dredged by the Siboga at stations 81 and 89, STEPHANOMETRA OXYACANTHA (Hartlaub) Puate 47, Figure 217; Puate 48, Figures 218-221 Antedon ozyacantha Hartiaus, Nachr. Ges. Gédttingen, May 1890, p. 178 (description; Amboina); Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 6 (with A. ludovici the most usual species A MONOGRAPH OF THE EXISTING CRINOIDS 419 of Antedon at Amboina), p. 11 (collected by Brock at Amboina), p. 15 (at Amboina next com- monest to A. ludovici), p. 39 (in key), p. 55 (Amboina; detailed description and comparisons), p. 58 (in second section of the Palmata group with spiny lower pinnules), p. 113 (in Géttingen Mus.), pl. 3, figs. 35, 37.—Prerrer, Abh. Senck. naturf. Ges., vol. 25, 1900, p. 85 (Ternate).— Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1581 (listed).— A. H. Cuarx, Smithsonian Misc. Coll., vol. 60, No. 10, 1912, p. 2 (identity); Crinoids of the Indian Ocean, 1912, p. 37 (identity). Himerometra oxyacantha A. H. Cuarx, Smithsonian Misc. Coll., vol. 50, 1907, p. 356 (listed). Stephanometra oxyacantha A. H. Cuark, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 10 (listed); Notes Leyden Mus., vol. 33, 1911, p. 183 (Amboina); Rec. Australian Mus., vol. 9, No. 1, 1912, p. 82 (detailed description of a specimen from Ugi, Solomon Islands); Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p. 2 (identity), p. 19 (Amboina; description of one of the specimens recorded by Hartlaub) ; Crinoids of the Indian Ocean, 1912, p. 37 (identity), p. 132 (synonymy; Amboina; Solomon Islands).—ReEIcHENSPERGER, Abh. Senck. naturf. Ges., vol. 35, Heft 1, 1913, p. 83 (Am- boina), p. 102 (characters of the specimens).—A. H. Cuarx, Unstalked crinoids of the Siboga- Exped., 1918, p. 94 (in key; range; references; notes; station 99; Enkhuizen Island), pp. 272, 276 (listed).—GisLin, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, p. 25.— A. H. Cuarx, Temminckia, vol. 1, 1936, p. 302 (near Bongao; near Koepang; Obi latoe; Am- boina; Lembeh Strait, Celebes; notes). Diagnostic features—The outer cirrus segments are wholly without dorsal processes, being merely more or less sharply carinate in the middorsal line; P; is flexible, flagellate, delicate, and slender, much smaller and weaker than P2; P2, Ps, and P,, and sometimes also P;, are much enlarged, stiffened, and spinelike, pro- gressively decreasing in length; there are 24-32 (usually 30) arms 100-150 mm. long; and the cirri have 23-29 (usually about 25) segments. Description.—The centrodorsal is moderate in size, discoidal, with the bare polar area slightly concave, 2.5 mm. in diameter. The cirri are XXXII, 23-24, from 20 to 25mm. long. The first segment is short, the next two are about twice as broad as long, and those following gradually increase in length, becoming about as long as broad on the fifth or sixth. The next three or four segments are slightly longer than broad, and those succeeding very gradually decrease in length, becoming about half again as broad as long distally. The segments in the proximal half of the cirri are slightly constricted centrally and have somewhat prominent ends, and those in the distal half are slightly compressed laterally and may be bluntly carinate. The opposing spine is median to terminal, blunt, and triangular, equal in height to one-quarter the width of the penultimate segment. The terminal claw is somewhat longer than the penultimate segment and is moderately slender, evenly tapering, and moderately curved. The radials are visible only in the interradial angles. The IBr; are very short, in contact basally, about four times as broad as long. The IBr, (axillaries) are short and broad, almost triangular, twice as broad as long, with the lateral edges swollen and produced into rounded lateral processes. The IIBr series are 2; the IIBr, are interiorly united for about three-quarters of their length. The elements of the divi- sion series and the first brachials have rounded lateral processes. In the specimen described the IIIBr series are present exteriorly on three postradial series. The 24 arms are about 120 mm. long. The first two brachials are short, wedge- shaped, twice as broad as long exteriorly, the first interiorly united for most of their length. The first syzygial pair (composed of brachials 3+ 4) is half again as broad 420 BULLETIN 82, UNITED STATES NATIONAL MUSEUM as long. The next three brachials are oblong, slightly more than twice as broad as long, those succeeding are wedge-shaped, soon becoming almost triangular, twice as broad as long, distally gradually becoming less and less obliquely wedge-shaped, and in the terminal portion of the arm longer, about as long as broad. Syzygies occur between brachials 3+-4, again from between brachials 10+ 11 to between brachials 15-+16 (usually between brachials 144-15 or 15+-16), and distally at intervals of from 7 to 10 (usually 8) oblique muscular articulations. P,; is 11.5 mm. long, slender, evenly tapering and becoming very slender distally, with 23 segments of which the first is about half again as broad as long and those fol- lowing gradually increase in length, becoming about as long as broad on the third and twice as long as broad, or somewhat longer, distally. P, is much stouter than P,, stiff and spinelike, 15 mm. long, with 12 segments of which the first two are subequal, about half again as broad as long, the third is half again as long as broad, and the remainder are about two and one-half times as long as broad. P; is 12 mm. long, resembling P2, and is composed of about 10 segments, of which the distal are much longer than those of the preceding pinnule. P, is 7 mm. long, resembling P; but not quite so stout, and is composed of 9 segments. P; is smaller than P, basally and becomes very slender distally; it is 6 mm. long and is composed of 13 segments. The following pinnules resemble P;, but rapidly become less stiffened basally. The distal pinnules are 9 mm. long. Notes.—The preceding description was drawn up from the specimen from Ugi, Solomon Islands, in the collection of the Australian Museum. In one of the specimens from Port Galera, Mindoro, with 30 arms P, is 17 mm. long, with 28 segments, and is very slender and flagellate. P, is 23 mm. long, with 18 segments. P;is 15 mm. long, with 13 segments. P, is 8 mm. long, with 11 segments, resembling the immediately preceding pinnules but smaller. On another arm P; is 20 mm. long, with 16 segments. P; is 16 mm. long, with 14 segments. P, is 9 mm. long, with 11 segments. On a third arm P, is 22 mm. long, with 18 segments. P; is 16 mm. long, with 14 segments. P, is 9 mm. long, with 12 segments. Another specimen has 30 arms 135 mm. long, and the longest cirri 26 mm. long with 24 segments. Two others have 30 arms 150 mm. long, and 32 arms 140 mm. long. Of the remaining specimens one has 32 arms and five have 30 arms. The specimen from Siboga station 99 has 31 arms about 110 mm. long. The cirri are XXIV, 24-29 (usually 24-26), 25 to 30 mm. in length. P; is 12 mm. long, very delicate, with 20 segments, of which the outer are three times as long as broad. P; is 15 mm. long, stiff and spinelike, with 13 segments. P3 is 14 mm. long, with 12 seg- ments, and resembles P,. P, is 9 mm. long, with 10 segments, and resembles P, and P;, but is proportionately smaller. Ps is 6 mm. long, with 9 segments, and is of the same character as the pinnules preceding. P, is weak and delicate, 4.7 mm. long, with 9 segments. The immediately following pinnules are similar to P. One of the two specimens from near Bongao has 19 arms 140 mm. long. P, is 18 mm. long, with 15-17 segments. P; is 10-12 mm. long, with 11-13 segments. P, resembles P;, but is only slightly longer than the pinnules succeeding, 7 mm. long, with 13 segments. All but one of the IIBr series are present. The other specimen is sim- ilar. It has 20 arms, of which 8 are in process of regeneration. All the I1Br series are present. A MONOGRAPH OF THE EXISTING CRINOIDS 421 In the specimen from near Koepang P2, P3, and P, are stiffened and enlarged. The curi have 19-21 segments. The specimen from Obi latoe has 27 arms 90 mm. long. P, is 13 mm. long, with 11 segments. P;is 10 mm. long, with 9 segments. P,is similar to P;,7 mm. long, with 8 segments. The example from Lembeh Strait, Celebes, is small and is in process of adolescent autotomy; it has 20 arms. The specimen collected by the Willebrord Snellius at Amboina has 30 arms 120 mm.long. The six arms on each postradial series are arranged in 2, 1, 1,2 order. The cirri are 30-33 mm. long, with 25 or 26 segments. P, is 23 mm. long, with 39 segments, very flexible, tapering more rapidly in the proximal fourth than beyond. P, is 30 mm. long, with 23 segments. P; is 22 mm. long, with 18 segments. P, is 18 mm. long, with 19 segments. P; resembles P, and is 12 mm. long, with 15 segments. The 10 specimens collected by Dr. J. Brock at Amboina were thus described by Hartlaub: The centrodorsal is thick, with convex sides and sometimes with a markedly pentagonal lower border. The free dorsal pole is usually rather small and is marked with small pits. The cirri are arranged in two and a partial third irregular marginal rows. The cirri are XXX~XXXV, about 25, from 18 to 28 mm. long, rather slender and laterally compressed. The fifth-eleventh segments are elongated, and often haye thickened ends. In the proximal half of the cirri the ventral side of the distal ends of the segments tends to overlap the proximal end of the segments following. There are no dorsal tubercles or spines with the exception of the opposing spine. The radials are entirely concealed. The IBr, are as long as the free sides of the shortly pentagonal IBr, (axillaries) and are entirely free laterally. The IJBr and IIIBr series are 2. The IIIBr series occur only on the outer sides of the postradial series so that as a rule each of the latter bears 6 arms. Slight synarthrial tubercles are present. The elements of the division series have slight thickenings on their lateral edges. The two ossicles immediately following each axillary are partially united interiorly. The arms are usually 30 in number and are 100 to 140 mm. long. The brachials are short. The first seven or eight are thick discoidal, the 30 or so following are triangular with slightly overlapping distal ends, and those succeeding are short wedge-shaped, later becoming short discoidal, toward the ends of the arms about as long as broad, and finally some- what elongated. Syzygies occur between brachials 3-++-4, again usually from between brachials 13-14 to between brachials 26-++27 (often at about the twenty-fourth bra- chial), and distally at intervals of usually 8-11 muscular articulations. On arms arising directly from a [Br axillary the second syzygy is between brachials 9+10 and the distal intersyzygial interval is four or five muscular articulations. The pinnules of the first pair are slender and flagellate, but those of the three or four following pairs are much thicker, stiff, straight, more or less sharp at the tip, and spinelike. P, is usually longer on the outer arms of each postradial series than on the inner. It measures up to 15 mm. in length and is composed of 20-27 cylindrical seg- ments which, with the exception of the basal, are longer than broad. P, is a good deal shorter than P;. P, and P;, the two following stiff and spiniform pinnules on the outer side of the arms, are usually markedly longer, and may reach 20 mm. in length. Some- times they are both of the same size, but their relative length varies, and P,; may be either longer or shorter than P;. Both pinnules are composed of 9-15 (usually about 422 BULLETIN 82, UNITED STATES NATIONAL MUSEUM 12) segments, of which the middle are twice, and the distal, with the exception of the two last, are three times as long as broad. The two following pinnules decrease markedly in length and in the number of their component segments. At least the first of these possesses the stiff and spinelike character of those preceding. The max- imum length of the distal pinnules is 10 mm. The length and form of the proximal spiniform pinnules are somewhat subject to individual variation. They are always longer on the outer side of the arms than they are on the inner. The disk is 15 mm. in diameter and is incised. Sacculi are numerous on the disk, arms, and pinnules. The dorsal skeleton is chocolate-brown, with the central region sometimes somewhat lighter; or the arms are grayish black with broad bands of light brown and the central region also light brown; or the color is light brown with very dark cross stripes at the brachial articulations. The disk is dark brown. I have examined two of the specimens that served as the basis for Hartlaub’s description, one at Leyden and one at Hamburg. The specimen at Hamburg has 30 arms 140 mm. long. The centrodorsal is low hemispherical, 6 mm. in basal diameter, with the dorsal pole small and flat, 2.6mm. in diameter; the cirrus sockets are arranged in three closely crowded irregular rows. The cirri resemble those of S. monacantha or S. tuberculata; the longest proximal segments are twice as long as broad or slightly longer, and the segments in the outer third or half of the cirri are about as long as broad. This is a more robust form than S. tuberculata or S. monacantha, and P2, Ps, and P, are similar. In the specimen at Hamburg P, is 12 mm. long, slender, with 21 seg- ments. P, is 13-16 mm. long, enlarged, stiffened, and spinelike, with 13 or 14 seg- ments. P; is 13-16 mm. long, with 12 or 13 segments, and is similar in every way to P,. P, is 9 mm. long, similar to the preceding pinnules, but proportionately smaller, with 10 segments. P,; is 6 mm. long, similar to P, but proportionately smaller, with 10 segments. The following pinnules are small and weak, 4.5 or 5 mm. in length, gradually increasing in size outwardly. The distal pinnules are 9 mm. long, with 20 segments, and very slender. The specimen collected by Professor Strubell at Amboina, according to Dr. August Reichensperger, has the cirri XXXII, 22-26, from 22 to 30 mm. long. With the exception of the first and the last the segments are longer than broad, and there are no dorsal spines. There are 29 arms 130 mm. long. P, is slender, somewhat stiffened though filiform, and is composed of 22-25 segments. P, to P; are stiff, styliform, and sharp-pointed. P, consists of 15 segments and reaches 21 mm. in length. The following pinnules measure 16, 11, and 7 mm. in length and are composed of 12, 10, and 9 segments. The structure of the arms, etc., agrees with that given in Hartlaub’s description, but the lateral thickenings of the ossicles from the 1Br axillary to the last axillary are very pronounced, and they differ from the adjacent parts in their lighter color. The color is chocolate-brown, with the distal portions of the arms abruptly much lighter. One of the specimens from the Danish Expedition to the Kei Islands taken at Toeal has 30 arms about 130 mm. long; the cirri have 23-25 segments. The other specimen has 29 arms about 110 mm. long; there are nine IIIBr series, all of them on the outer sides of the postradial series so that the arms are developed in 2,1,1,2 order. A MONOGRAPH OF THE EXISTING CRINOIDS 423 Of the four specimens collected by the Siboga at Enkhuizen Island, one has 30 arms 135 mm. long and the cirri XXXV, 24-25, 30 mm. long. P, is 11 mm. long, with 20 segments, very slender and delicate, with the outer segments three times as long as broad. P, is 15 mm. long, stiff and spinelike, with 12 segments. P, is 14 mm. long, with 11 segments, and resembles P2, but is slightly more slender. P, is 10 mm, long, with 11 segments, and resembles P;. Ps is 7 mm. long with 9 segments and is of the same characte~ as the immediately preceding pinnules. The following pinnules are short. The other specimens are similar. Localities.—Port Galera, Mindoro, Philippines; Lawrence E. Griffin (23, M.C.zZ.., 615, 630, 690-696). Siboga station 99; anchorage off North Ubian, west of Jolo (Sulu), Philippines (lat. 6°07’30’” N., long. 120°26’00’” E.); 16-23 meters; lithothamnion bottom; June 28-30, 1899 [A. H. Clark, 1918] (1, Amsterdam Mus.). Ugi, Solomon Islands [A. H. Clark, 1912, 1918] (1, Austr. M.). Ternate, Moluccas |Pfeffer, 1900; A. H. Clark, 1918]. Anchorage near Bongao, Tawi-tawi Islands; 27 meters; Willebrord Snellius, September 9, 1929 (2, L.M.). Near Koepang; 6-15 meters; Willebrord Snellius, December 4, 1929 (1, L.M.). Obi latoe; shore and reef; Willebrord Snellius, April 23-27, 1930 (1, L.M.). Lembeh Strait, Celebes; Willebrord Snellius, September 25, 1930 (1, L.M.). Amboina; Willebrord Snellius, September 17, 1930 (1, L.M.). Amboina; Dr. J. Brock [Hartlaub, 1890, 1891; A. H. Clark, 1907, 1909, 1911, 1912, 1918] (2, H.M., L.M.). Amboina; Professor Strubell [Reichensperger, 1913; A. H. Clark, 1918]. Danish Expedition to the Kei Islands; Dr. Th. Mortensen; Vatek van Toeal; about 1-2 meters; rocky coast; March 27, 1922 (2). Siboga, Enkhuizen Island, near Batavia, Java [A. H. Clark, 1918] (4, U.S.N.M., E. 469; Amsterdam Mus.). Geographical range——From the Philippines to the Solomon Islands and the Moluccas, and westward to Java. Bathymetrical range.—From the shoreline down to 16 (?23) meters. Occurrence.—Hartlaub said that the commonest comatulids at Amboina seem to be this species and Heterometra amboinae (‘‘Antedon ludovici’’). History—This species was first described as Antedon oxyacantha by Dr. Clemens Hartlaub in 1890. In 1891 he redescribed and figured it, and compared it in detail with related species. It was recorded by Prof. Georg Pfeffer from Ternate in 1900. On my establishment of the new genus Himerometra in 1907, oryacantha was included in it, but it was removed to the new genus Stephanometra in 1909. In 1911 I published notes on one of Hartlaub’s specimens from Amboina, which I had seen in the Leyden Museum, and in 1912 I published notes on another that I had examined at the Hamburg Museum. In 1912 also I described in detail a specimen in the Australian Museum from Ugi in the Solomon Islands. Dr. August Reichensperger in 1913 described a specimen collected at Amboina by Professor Strubell, and in 1918 I recorded and gave notes upon five specimens collected by the Siboga at station 99 and at Enkhuizen Island. 424 BULLETIN 82, UNITED STATES NATIONAL MUSEUM STEPHANOMETRA SPICATA (P. H. Carpenter) Prats 49, Fiaures 223, 224; Puare 91, Figure 447 Antedon spicata P. H. Carpenter, Notes Leyden Mus., vol. 3, 1881, p. 190 (description; Banda Sea).—Be.1, Proc. Zool. Soc. London, 1882, p. 533 (listed) —P. H. Carpenter, Proc. Zool. Soc. London, 1882 (1883), p. 746 (specific formula).—Be 1, Proc. Linn. Soc. New South Wales, vol. 9, 1884, 1885, p. 497 (Ugi); Proc. Zool. Soc. London, 1888, p. 389, footnote 5 (collected by Dr. Anderson at Mergui).—P. H. Carpenter, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 225; Journ. Linn. Soe. (Zool.), vol. 21, 1889, p. 305 (specific formula), p. 310 (King Island, sublittoral; characters of the specimen; comparison with indica and tuberculata), pl. 27, figs. 3-5—Harriaus, Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 38 (in key), p. 58 (in second section of species of the Palmata group with spiny lower pinnules; characters).— HAMANN, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1581 (listed.)— A. H. Crark, Rec. Australian Mus., vol. 9, No. 1, 1912, p. 81 (known from Ugi); Crinoids of the Indian Ocean, 1912, pp. 35, 36 (of P. H. Carpenter 1888 and 1889=Stephanometra spicata), p. 41 (of Bell, 1909=Cenometra emendatriz), p. 136 (B.M., MS.=Stephanometra monacantha). Antedon tuberculata yon Grarr, Challenger Reports, Zoology, vol. 20, pt. 61, 1887, pp. 4, 6 (station 174; myzostomes).—P. H. CarrentER, Challenger Reports, Zoology, vol. 26, pt. 60, 1888, p. 232 (description; station 174, B, C, or D), pl. 45, figs. 2, 3; Journ. Linn. Soe. (Zool.), vol. 21, 1889, pp. 310, 311 (comparison with A. spicata).—Harruaus, Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 38 (in key), p. 58 (in second section of species of Palmata group with spiny lower pinnules), p. 59 (characters).—Burtt, Willey’s Zool. Results, pt. 2, 1899, p. 133 (Lifu)—Hamann, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1581 (listed).—A. H. Criark, Proce. U. S. Nat. Mus., vol. 43, 1912, p. 385 (of Hartlaub, 1891 = Stephanometra tuber- culata); Crinoids of the Indian Ocean, 1912, pp. 35, 39 (of P. H. Carpenter, 1888, and Bell, 1899 = Stephanometra tuberculata). Antedon ?spicata Bru, Proc. Zool. Soc. London, 1894, p. 396 (Macclesfield Bank, 20-35 fathoms) .— A. H. Cuark, Crinoids of the Indian Ocean, 1912, p. 38 (identity). Himerometra spicata A. H. CLarx, Smithsonian Mise. Coll., vol. 50, 1907, p. 356 (listed). Himerometra tuberculata A. H. CtArx, Smithsonian Mise. Coll., vol. 50, 1907, p. 356 (listed). Stephanometra spicata A. H. Cuark, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 10 (listed); Notes Leyden Mus., vol. 33, 1911, p. 176 (= Antedon spicata P. H. Carpenter), p. 183 (Banda Sea; redescription of the type); Rec. Australian Mus., vol. 9, No. 1, 1912, p. 84 (Bell’s record from Ugi); Crinoids of the Indian Ocean, 1912, pp. 35, 36 (identity of previous records), p. 132 (synonymy; Banda Sea; Ugi; detailed description of the type; comparison with S. tuberculata) ; Unstalked Crinoids of the Siboga-Exped., 1918, p. 94 (in key; range), p. 95 (references; notes; station 99), p. 272 (listed), pl. 7 (colored figure).—GusLin, Zool. Bidrag Uppsala, vol. 9, 1924, pp. 41, 44 (reversion), p. 78 (syzygies), p. 88 (articulations), p. 283, figs. 53, 54, p. 67, fig. 70, p. 75, fig. 99, p. 87, figs. 154-157, p. 98 —A. H. Cuark, Proc. Biol. Soc. Washington, vol. 47, 1934, p. 11 (Blakang Mati; notes).—GuisL6n, Kungl. Fysiogr. Sillsk. Handl., new ser., vol. 45, No. 11, 1934, p. 25; Kungl. Fysiogr. Saillsk. Lund Férh., vol. 7, No. 1, 1936, p. 3 (Macclesfield Bank), p. 4 (French Indochina), p. 5 (range), p. 11 (locality; notes)—A. H. Criarx, John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1936, p. 87 (listed), p. 88 (Mabahiss station 10; notes; range), p. 100 (range), p. 104. Stephanometra tuberculata A. H. CuarK, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 10 (listed); Smithsonian Misc. Coll., vol. 60, No. 10, 1912, p. 19 (compared with S. oryacantha), p. 20 (Ruk; Pitilu; Pelew Islands; descriptions of specimens); Proc. U. 8S. Nat. Mus., vol. 43, 1912, p. 385 (Antedon tuberculata of Hartlaub, 1891, is this species), p. 396 (New Guinea; detailed descrip- tion; no locality; descriptions of 2 specimens) ; Crinoids of the Indian Ocean, 1912, pp. 35, 38, 39 (identification of previous records), p. 133 (synonymy; localities; description of the type speci- men); Smithsonian Mise. Coll., vol. 61, No. 15, 1913, p. 28 (published references to specimens in the B. M.; localities represented by the specimens).—HARTMEYER, Mitt. zool. Mus. Berlin, vol. 8, Heft 2, 1916, p. 235 (New Guinea, No. 2971). Stephanometra marginata A, H. Cuarx, Vid. Medd. Naturh. Foren. Kgbenhayn, 1909, p. 169 (Singa- pore; description of a specimen), p. 170 (a typical Stephanometra); Crinoids of the Indian Ocean, A MONOGRAPH OF THE EXISTING CRINOIDS 425 1912, pp. 34, 40, 41 (identity of previous records), p. 185 (synonymy; localities; notes on the type specimen).—H. L. Crarx, Spolia Zeylanica, vol. 10, pt. 37, 1915, p. 93 (occurs at Ceylon). Stephanometra sp. A. H. Cuark, Bull. Mus. Hist. Nat., Paris, 1911, p. 252 (New Caledonia; notes). Stephanometra stypacantha H. L. Crarx, Carnegie Inst. Washington Publ. 212, 1915, p. 103 (descrip- tion; Mer).—A. H. Cuarx, Unstalked erinoids of the Siboga-Exped., 1918, p. 94, footnote 2 (synonym of S. spicata)—H. L. Crarx, The echinoderm fauna of Torres Strait, 1921, p. 8 (collected by the Carnegie Exped., 1913), p. 22 (Mer; notes; color in life), pp. 192 ff. (range in Australia), pl. 22, fig. 4, pl. 36, figs. 3, a-e. Diagnostic features.—The outer cirrus segments are wholly without dorsal proc - esses, being merely more or less sharply carinate in the middorsal line; P, is flexible, flagellate, delicate, and slender, much smaller and weaker than P,; P, and P,; are much enlarged, stiffened, and spinelike, P; being smaller than P; though otherwise similar to it; P, is small and weak like the pinnules following; there are 14—33 (usually about 30) arms 70-120 (most commonly about 100) mm. long; and the cirri have 18-28 (usually 20-25) segments. Description.—The centrodorsal is discoidal, usually rather thin though sometimes thick, with strongly sloping sides and a small concave dorsal pole, more rarely low hemispherical, with an approximately flat, small dorsal pole. The cirrus sockets are arranged in from one and a partial second to three or more, but usually in two, irregu- larly alternating marginal rows. The cirri are XXV-LII (usually XXV-XXXV), 18-28 (usually 20-25), 15 to 30 mm. (usually 20-25 mm.) long. The basal segments are short, and those following gradually increase in length to the fourth or fifth, which with a few of those succeeding are usually from one-third to one-half again as long as broad but vary from slightly broader than long to twice as long as broad. The following segments gradually decrease in length so that the distal are usually somewhat broader than long, though they may be as long as broad, or even slightly longer than broad. The short outer segments are rather strongly compressed laterally and bear a more or less well developed keel in the middorsal line. The opposing spine is prominent, subterminal, and di- rected obliquely forward. The terminal claw is slightly longer than the penultimate segment and is stout and rather strongly curved. The radials are wholly concealed, or their distal ends are just visible beyond the rim of the centrodorsal. The anterolateral angles of adjacent radials are separated by a notch. The IBr, are short, oblong, rather strongly convex dorsally, and quite free laterally. The IBr, (axillaries) are broadly pentagonal and are from about half again to more than twice as long as the I[Br;._ The IIBr, IIIBr, and IVBr series are 2. The ITIBr series are usually developed only on the outer sides of the postradial series so that typically each postradial series bears six arms arranged in 2, 1, 1, 2 order. The ossicles immediately following each axillary are almost completely united interi- orly. The division series are well separated, rather narrow, and rather strongly convex. More or less well developed synarthrial tubercles are present. The ventro- lateral borders of all the ossicles from the IBr, to the second brachials inclusive bear usually small irregular projections toward the ventral side; but they may be produced into prominent, though narrow, flangelike extensions. The arms are 14~33 (usually about 30) in number and are 70 to 120 mm. (most commonly about 100 mm.) in length. They consist of about 120-150 brachials. The first two brachials are wedge-shaped, the second having more oblique ends than the 208244—41——_28 426 BULLETIN 82, UNITED STATES NATIONAL MUSEUM first. The first syzygial pair (composed of brachials 3-4-4) is about as long as broad, or slightly broader than long. The next four or five brachials are oblong, and those succeeding are wedge-shaped or even triangular, broader than long, with very slightly produced and overlapping distal ends, gradually becoming quadrate and more dis- coidal again in the middle of the arm. Syzygies occur between brachials 3+4, again from between brachials 8+9 to between brachials 22-+-23 (usually from between brachials 13-+14 to between brachials 16-++17, and on arms arising from a [Br axillary with an extra one between brachials 8+9 or 9+10), and distally at intervals of 4-11 (usually 6-8) muscular articulations. P, is weak, slender, and flagellate. It varies in length from half as long as P2 to quite as long as the latter but is usually two-thirds to three-fourths the length of P». It measures 4 to 14 mm. in length but is usually between 9 and 11 mm. long. It is composed of 14-27 (usually 17-22) segments, of which the first is twice as broad as long, the third is about as long as broad, and those following gradually increase in length so that the distal are three times as long as broad or even longer. P, is stouter, usually much stouter, than P,, usually markedly longer and never shorter, composed of fewer segments, very stiff and styliform with a sharp spinelike tip, straight and erect, recurved at the base, or recurved for its entire length. It varies from 8 to 17 mm. in length and is usually between 10 and 15 mm. long. It is composed of 9-18 (usually 10-15) segments of which the first is twice as broad as long, the third is about as long as broad or somewhat longer than broad, and most of the remainder, except the terminal, are three times as long as broad or even longer. The terminal 4-6 segments are sometimes abruptly less in diameter than those preceding, so that the pinnule appears to have a regenerated tip. Although three instances of this have been reported from widely separated regions (the Philippines, Singapore, and Ceylon), it may be the result of injury or of a diseased condition. P; resembles P, in character, but it is always shorter. It is commonly about two-thirds as long as P:, but may be longer and even very nearly as long as the latter. Its length varies from 5 to 14 mm., being usually between 8 and 12mm. _ Its stoutness in relation to P, is proportionate to the length of the two pinnules. Pg ; usually is composed of one to four fewer segments than P., but occasionally it has the same num- ber of segments, and rarely it has from one to three more. The segments number from 8 to 14 and are most commonly 10-12. They are of the same proportions as the segments of P). P, is variable. It is typically and usually small, slender, and flexible like the following pinnules and thus very different from the enlarged and styliform P, and P3. But it may be slightly longer than the succeeding pinnules and somewhat stiffened, and in extreme cases it may, though much smaller than P;, be nearly three-fourths of its length and more or less like it distally, differing more from the slender, weak, and flexible P; than it does from P;. It varies from 3 to 9 mm. in length and is composed of 8-12 (usually 9-11) segments. Notes.—The specimen from the Macclesfield Bank is smaller than the type of Carpenter’s Antedon tuberculata from Challenger station 174 but is otherwise exactly like it. The specimen from Annam, according to Gislén, has the cirri 11-14 mm. long, with 19-21 segments. There are no dorsal spines, but an opposing spine is present. A MONOGRAPH OF THE EXISTING CRINOIDS 427 I1IBr series are developed on the outer sides of the postradial series. The division series are constricted ventrolaterally at the articulations, the lateral margins thereby acquiring a knobby appearance. There are 28 arms. P, and P, are stiffened. P, is the longest pinnule and is composed of 13 segments. P; has 10 segments. The disk is much incised, without calcareous granules, white spotted with brown. The color otherwise is white, banded with brown, especially at the articulations. Of the three specimens from Bantayan Reef, Cebu, one has 30 arms, one has about 30 arms, and the third is fragmentary. In the specimen with 30 arms P, is 11 mm. long with 20 segments, P, is 17 mm. long with 15 segments, and P, is 11 mm. long with 14 segments resembling P, but proportionately smaller. The specimen from Siboga station 99 has 20 arms 85mm.long. The cirri are XXV, 22-23, 20 mm. long. P, is comparatively short, 11 mm. long, with 15 segments. P; is 8 mm. long, with 13 or 14 segments. The specimen from the Pelew Islands has 30 arms 115 mm. long. The cirri are 25 to 30 mm. long and consist of 20-23 segments, of which the longer proximal are twice as long as broad and the shorter distal are slightly longer than broad. The cirri are moderately slender and very long as in the specimen from Pitilu (see below), though they are not quite so slender as in that individual. P, is 13 mm. long, slender and flagellate, with 26 segments. P, is 15 mm. long, very stiff and spinelike, but slightly more slender than usual, with 13 segments. P; is 12 mm. long, with 11 seg- ments, and resembles P;. P, is 9 mm. long, with 11 segments, resembling P; but much more slender. The example from Ruk has 31 arms 85 mm. long. The cirri are about 23 mm. long and consist of 26-27 segments, of which the longest are usually not quite so long as broad, rarely as long as broad, and the outermost 9 to 12 are broader than long. The enlarged lower pinnules have 10 or 11 segments. P, is more like the following pinnules than it is in the specimen from Pitilu (see below), and the shorter and somewhat stouter cirri with shorter segments give this individual a somewhat different appearance. The specimen from New Guinea has 15 arms about 60 mm. long. One IIIBr series is developed externally. P, is 9 mm. long, with 10 greatly elongated segments. P; is similar to P, but smaller, 6 mm. long, with 8 segments. P, has 8 segments, of which the outer resemble those of the preceding pinnules; it is 4 mm. long and is slightly stiffened. P, and P; are comparatively slender, possibly because of the small size of the animal. The specimen from Pitilu, Admiralty Islands, has 27 arms 95 mm. long. The cirri are XXX, 20-23, from 25 to 27 mm. long. The longest segments of the longer cirri are twice as long as the median width, or somewhat longer. P, is 10 or 11 mm. long, weak, and slender, with 17 or 18 segments. Pz, is stout and spinelike, 10 mm. long, with 10 or 11 segments. PP, is similar to P, but slightly smaller, 9 to 10 mm. long, with 9 or 10 segments. P,; is 7 mm. long, smaller than the preceding pinnules, but stiffened. P; is 5 mm. long, with 10 segments. The following pinnules are similar but slightly smaller and weaker. The distal pinnules are very slender, 9 mm. long, with 18 or 19 segments. 428 BULLETIN 82, UNITED STATES NATIONAL MUSEUM The specimen from Challenger station 174 was described by Carpenter as repre- senting a new species, which he called Antedon tuberculata. According to Carpenter the centrodorsal is saucer-shaped. The cirri are about XL, 20-25, rather stout with tolerably uniform segments, few of which are longer than broad; there is a small opposing spine. The radials are concealed by the centrodorsal. The IBr, are short and rather convex dorsally and are not united laterally. The IBr, (axillaries) are more than twice as long as the IBr; and are broadly pentagonal. The articulation between the IBr,; and IBr. is rather tubercular. The division series are well sepa- rated, and the postradial series may divide three times. The I1Br and IIIBr series are 2 and bear synarthrial tubercles. III[Br series are usually developed only on the outer sides of the postradial series, so that each postradial series bears six arms arranged in 2,1, 1,2 order. The ventrolateral borders of all the ossicles of the division series bear small tubercles, which project somewhat toward the ventral side. The 31 arms are about 90 mm. long and consist of 120 or more brachials, the first 8 or 10 of which are thick disks, those following being shortly triangular and later becoming more discoidal again. The first syzygy is between brachials 3-++4, the second from between brachials 8-+-9 to between brachials 22+ 23, generally about brachials 13-+14 or 15+16, and the distal intersyzygial interval is from 6 to 11, usually 8, muscular articulations. The pinnules of the first pair (P; and P,) are about 9 mm. long and consist of some 25 longish segments. The pinnules of the second pair (P2 and P,) are much stouter and very stiff and tapering, reaching 15 mm. in length; they are composed of about 12 segments all of which, except for those at the two ends, are much longer than broad. The pinnules of the third pair (P; and P,) are of the same character but are rather shorter than the pinnules of the second pair. The pinnules of the fourth pair (Py and Py) are much smaller and less stiff. The disk is lost, but was probably about 12 mm. in diameter. Sacculi are abundant along the ambulacral grooves of both arms and pinnules. The color in alcohol is purplish white with occasional dark purple bands, the ventral perisome greenish gray. I examined this specimen at the British Museum in 1910. There are 30 arms. P, is slender and flexible, becoming very slender and flagellate distally. P: is much enlarged, stiff and spinelike, nearly half again as long as P3, with 12 segments. P3; resembles P, but is shorter and slightly less stout. Py, is very short and small but is more or less stiffened. P; resembles P,. Carpenter said that twberculata, as well as spicata (from the Banda Sea; see beyond) and indica (p. 439), is distinguished by the characters of the pinnules of the second pair which are considerably longer than the pinnules of the first pair, though com- posed of a smaller number of segments, which, however, are of very large size, some of them reaching 1.5 mm. The segments decrease gradually in diameter from the base to the tip of the pinnule so as to give the latter a remarkably stiff and tapering appearance. He said that there is some indication of this in marginata (p. 447), but its large pinnules are less stiff with relatively shorter segments, which are more uni- form in diameter so that the pinnules lack the tapering and styliform appearance which is so marked in spicata and tuberculata. In marginata also the cirri are both smaller and have fewer segments than those of tuberculata, while the IBr, and IBr, (axillaries) are more nearly equal in length, and portions of the radials are visible. A MONOGRAPH OF THE EXISTING CRINOIDS 429 Carpenter noted that twberculata has many points of resemblance with spicata from the Banda Sea (see beyond), and it may be that a larger knowledge of both types will eventually lead to their union He said that in tuberculata the cirri are both considerably more numerous than in spicata and reach a larger size, though the actual number of segments composing them is the same in both forms. The IBr, of tuber- culata are short as compared with the axillaries, not reaching half their length, while in spicata the axillaries are short as compared with the IBr;. In spicata the arms are also longer than in tuberculata, and the muscle plates are more prominent at the sides of the ambulacra. Of the two specimens from Lifu one bas 30 arms 120 mm. long, and the cirri LIL, 25-28 (usually 25), 20 to 30 mm. long. P, has 15 segments, and P; has 11 or 12 segments. The other is similar. In the specimen from New Caledonia the cirri are XVI, 18-19. There are 24 arms. P, is much enlarged and very stiff and spinelike. P, resembles P; but is slightly smaller. P, and the pinnules following are short but stiffened. The specimen from Mer was described by Dr. Hubert Lyman Clark as a new species under the name of Stephanometra stypacantha. He said that the centrodorsal is large, thick, slightly convex, and almost completely covered by the cirri. Its diameter is 3.5 mm., but the bare dorsal pole is less than 1 mm. across. The cirri are XX XIII, 18, about 12 to 13 mm. long, and are distally very distinctly compressed. The basal three or four and the terminal half dozen segments are more or less shortened, and the others are longer than broad. The terminal six or seven segments may show a longitudinal keel dorsally, but this is never marked. The opposing spine is distinct but small. The division series are all 2. The axillaries are pentagonal, nearly as long as broad. The division series have no lateral processes, and no synarthrial tubercles. The 16 arms are about 70 mm. long. At the base of the arm and distally the brachials are quadrate, but the seventh-thirtieth (or thereabouts) are more or less markedly wedge-shaped. Syzygies occur between brachials 3+4 and 16+17 and then at intervals of seven to nine muscular articulations. Sometimes they occur between brachials 9-10 and 14++-15 (or 15+16) and then at intervals of six or seven muscular articulations. P, is about 4 mm. long, flagellate, with 14 segments, of which the basel are stout, but the distal are long end very slender. P, is similar but much smaller. P, is very rigid, sharp, and spinelike, 8 mm. long, with 9 segments, of which 3 to 6 are greatly elongated, three times or more as long as broad, and the ninth is minute. _ P, is similar but is evidently smaller in every way. P, is similar to P, but is only 5 or 6 mm. long. P, is similar but much smaller. P, is more flagellate, less than 3 mm. long, and with only 8 segments. P, is similar, and about equal, to Py. The succeeding pinnules gradually become longer but do not exceed 6 or 7 mm., with 14 or 15 segments. All the pinnules are more or less cylindrical and are composed of smooth segments. Dr. Clark said that this species seems to be very well characterized by the numer- ous, few-jointed, nearly smooth cirri, the absence of ventrolateral processes on the arms, and the small number of segments in P,; the form and proportions of the lower pinnules also seem to be characteristic. According to Carpenter the type specimen from the Banda Sea has the centro- dorsal thick and discoidal with a small slightly hollowed dorsal pole and very sloping 430 BULLETIN 82, UNITED STATES NATIONAL MUSEUM sides, which bear about 25 cirri in an irregular double row. The cirri are XXV, 25. The fourth segment is longer than broad, the next three are the longest, and those following diminish gradually in length. The penultimate segment bears a tolerably strong opposing spine. The radials are barely visible. The IBr, are oblong and are not united laterally. The IBr, (axillaries) are pentagonal, relatively short, with a wide distal angle; they are only half again 9s long as tbe IBr,. The ventrolateral borders of the IBr axillaries and of the three or four following ossicles bear small irregular tubercles. The IIBr series, IIIBr series, and IVBr series (when present) are 2. The ossicles immediately following each axillary are almost completely united interiorly. The 28 arms are 100 mm. long and consist of nearly 200 brachials. The first brachials are widely rhomboidal. The second brachials are of about the same length, but are more wedge-shaped. The first syzygial pair (composed of brachials 34-4) is nearly square. The next four brachials are oblong, and those following are wedge-shaped and of medium length, with very slightly overlapping distal ends. The later brachials are blunter and more oblong, becoming squarer and slightly elon- gated at the arm ends. The muscle plates of successive brachials stand up rather prominently, alternately on either side of the ambulacral groove. Syzygies occur between brachials 3+-4, again from between brachials 16+17 to between brachials 19+20, and distally at intervals of 4-11 (usually 6-8) muscular articulations. P, is moderately long, but slender, and P, is shorter than P;. The pinnules of the next pair (P, and P,) are longer than P,, and are stiff, tapering, and styliform, consisting of about 15 elongated segments. P, is longer than P,, reaching 15 mm. in length. PP, is stiff but shorter again, and the following pinnules decrease until about the tenth brachial, after which the size increases slowly. Toward the arm ends the pinnules become slender and filiform, but they never reach the length of the pinnules of the second pair. The disk is 17 mm. in diameter, naked and much incised. Sacculi are closely set along the pinnule ambulacra. The color in alcohol is dorsally light purplish red, with darker bands at the articulations; the perisome is very much darker, almost black. Hartlaub reexamined the type specimen of spicata at Leyden and compared it in detail with his new species Antedon (=Stephanometra) oryacantha. He said that judging from the type specimen, spicata is a much more slenderly built species though otherwise in its general habitus very similar to oryacantha. The postradial series, how- ever, are not widely separated laterally as is the rule in oryacantha but are more nearly in lateral contact. The pinnules of the first pair (P; and P,) are slender and flagellate and reach, as given in Carpenter’s latest statement (concerning the specimen from the Mergui Archipelago), almost the length of the pinnules of the following pair (P, and P,), while in oryacantha they are usually markedly shorter than the pinnules of the second pair. P, is composed of 16 and more segments and is less markedly spiny than P, in oryacantha, which is composed of 12, or at the most 15, segments. P,; is markedly shorter than P, and, as is shown in Carpenter’s figure, is less procumbent and stiff, while in oxyacantha occasionally it is longer than P, and is quite as markedly spini- form. In spicata the outer pinnules are slender and filiform, a feature that Hartlaub considered as very characteristic of that form. I examined this specimen at Leyden in 1910. The cirri are XXIII, 22-25, rather slender, resembling those of such species as S. protectus. The longest cirrus A MONOGRAPH OF THE EXISTING CRINOIDS 431 segment is about twice as long as the median width, and the longer proximal segments are somewhat constricted centrally, or ‘dice-box shaped.” The IIIBr series are externally developed. The lateral tubercles on the rays are well developed and thick. P, is the largest pinnule and is composed of 16 or 17 segments. P is similar but is not quite so long. P,is much shorter than P;, but stiff, with 11 or 12 segments. P, is slightly shorter than P, and resembles the succeeding pinnules instead of being stiff like those preceding, though it may be a trifle stiffer than its successors. The specimen from Cocos Island has 30 arms. A magnificent specimen from Singapore has exactly 20 arms 95 mm. long and the cirri XXV, 24, 20 mm. long. P, is 11 mm. long with 18 segments, stiff and spinelike ; the terminal 4-6 segments are abruptly less in diameter than those immediately preceding, giving the pinnule tip the appearance of having been broken off and subse- quently repaired as described by Carpenter in Antedon marginata, and also noticed by Chadwick in a specimen from Ceylon. P; resembles P:, but is only 8 mm. long and is proportionately less stout. The following pinnules are small and weak. A very fine specimen from Singapore has 31 arms about 150 mm. long. The cirri are XXVII, 26,28 mm.long. The dorsal pole of the centrodorsal is 3.5 mm. in diam- eter and is slightly concave. P, is 12 mm. long and is composed of 11 segments. A third specimen from Singapore has the centrodorsal small and discoidal with the dorsal pole slightly concave, 2.5 mm. in diameter. The cirri are XXVIII, 20-23, from 15 to 20 mm. long, and slender. The first segment is short, and those following gradually increase in length to the sixth, which, with the three or four following, is twice as long as broad. The succeeding segments gradually decrease in length so that those in the outer third of the cirri are about as long as broad. The shorter distal segments are faintly carinate dorsally. The opposing spine is prominent, and is subterminal in position. P, is 11 to 14 mm. long, with 22-27 segments, equal in length to P, and stiff, but much more slender. P, has 12-13 segments, of which the distal are about three times as long as broad. P; is stiff like P., but smaller, 6 to 8 mm. long, with 12 segments. P, and the following pinnules are small and weak. Another specimen with 31 arms resembles the preceding. P, is 14 mm. long and is composed of 14 segments. A fifth specimen from Singapore has 30 arms 100 mm. long. The cirri are XXX, 21-23, from 17 to 20 mm. long. The dorsal pole of the centrodorsal is slightly con- cave, 2.5 mm. in diameter. A smaller specimen from Singapore has 14 arms 85 mm. long. On one of the arms arising from a [Br axillary P; is exactly like P2, though usually it is more nearly like P,. In arms arising from ITBr axillaries the relationships of the lower pinnules are the same as in fully grown individuals. The specimen from Pulau Obin, Singapore, has 30 arms. The specimen from Blakang Mati is a fine example of the species. It has 28 arms 125mm. long. P,is 19 mm. long, with 17 segments, and P; is 14 mm. long, with 15 segments. Carpenter said that a single individual that he considered as representing spicata was taken at King Island in the Mergui Archipelago. But it has rather more curri than the type specimen at Leyden and in this respect approaches the allied species indica (= prote:tus) and tuberculata, which resemble it in having long and stiff pinnules 432 BULLETIN 82, UNITED STATES NATIONAL MUSEUM on the fifth and sixth brachials. In indica, however, P; is much smaller than P2, as is well shown in Smith’s figure, while in spicata it may nearly equal its successor in length, though it is far more slender. In tuberculata it is also short, though stiffer than in the other two forms. Carpenter said that tuberculata approaches spicata in the character of P3, which is composed of elongated segments like those of P, and not of numerous smaller seg- ments as in indica. In some arms of spicata P; is smaller on the outer than on the inner side of the arm—that is, P; is smaller than P,. But the component segments of these large and stiff pinnules have not the extreme length that they reach in tuberculata. He remarked that the characters of this Mergui example of spicata confirm his separation of this species from tuberculata, for the latter has but 12 seg- ments in a P, that reaches 15 mm. in length, while in spicata a pinnule of the same length consists of 16 segments or more. In both alike, however, the division series have marginal projections that seem to be absent in indica. But the cirri of tuber- culata are stouter though with a somewhat smaller number of segments than occur in spicata. The type specimen of spicata has no IVBr axillaries, but one occurs in the Mergui specimen. The specimen recorded by Chadwick from Ceylon at station XX XIX was con- siderably mutilated. It had 11 arms. In most cases P, agreed well with P, in the type specimen of marginata, seeming to have been broken off by some accident and incompletely repaired. Chadwick said that a small specimen from Welligam Bay (station XXXIV) may possibly belong to this species (marginata). The number of arms can not be deter- mined, but in one of the postradial series the outer face of the I/Br axillary bears two IIIBr series. The specimen from Mabahiss station 10 has 20 arms 105 mm. long. The two cirri remaining have 18 and 20 segments; the longest is 16 mm.long. P3; may resemble P, and be of the same size, though usually it is smaller, or it may be more or less reduced and weak, resembling, or approximating in appearance, P,. P» is composed of 17 segments and is slender and delicate distally instead of stout and spinelike as usual. This specimen is more or less intermediate between spicata and indica, though more like the former. One of the specimens without locality (Berl. M., 5287) is large with 33 arms 100 mm. long. The cirri are XXXV, 18-25 (usually the latter), 15 to 25 mm. long, the peripheral cirri being the longest. The first two segments are very short, and those following gradually increase in length so that the fourth or fifth is about as long as broad. Those succeeding are about as long as broad or slightly longer than broad, after the middle of the cirrus gradually becoming shorter so that the segments in the outer third are slightly broader than long. The short distal segments are strongly compressed laterally and bear a faint median dorsal carination. There are no dorsal spines. ‘The opposing spine is prominent, subterminal, and directed obliquely forward. The terminal claw is slightly longer than the penultimate segment and is stout and rather strongly curved. The sides of the ossicles of the division series are ventro- laterally produced into prominent flangelike borders, which, however, are com- paratively narrow. P,is 10 mm. long, slender, and flagellate, composed of 20 segments, of which the first is twice as broad as long, the third is about as long as broad, and A MONOGRAPH OF THE EXISTING CRINOIDS 433 those following gradually increase in length so that the distal are three times as long as broad. Py is stout and very stiff, 13.5 mm. long, with 14 segments, of which the first is twice as broad as long, the third is somewhat longer than broad, the fourth is half again as long as broad, the fifth is twice as long as broad, and the remainder are nearly three times as long as broad. P; is 9 mm. long with 10 segments and is similar to P:, but the segments are relatively more elongated. P, is small, slender, and weak, 4.5 mm. long, with 11 segments. The other specimen without locality (Berl. M., 5288) has 28 arms about 100 mm. long. The centrodorsal is low hemispherical, about 5 mm. in diameter at the base, with the approximately flat dorsal pole 1.5 mm. in diameter and studded with small deep pits representing obsolete cirrus sockets. The cirri are XXVII, 21-22, from 20 to 25 mm. long, and are arranged on the centrodorsal in two and a partial third closely crowded alternating marginal rows. The longer proximal cirrus segments are half again as long as their proximal width and have a slightly concave ventral profile so that their ends are prominent. The outermost 10 segments are about as long as broad, and the distal, especially the last three or four, are rather strongly carinate dorsally. The opposing spine is prominent and subterminal. The terminal claw is stout, rather short, and strongly curved. All but one of the postradial series bear external IIIBr series. The exception is a postradial series with the IIBr and IIIBr series missing on one side. P, is slender and flagellate, 7 mm. long, with about 17 segments. P, is stiff and spinelike, 12 mm. long, with 11 segments. P3 is 10.5 to 11 mm. long, with 10 segments, very slightly slenderer than P,. P,is7.5 to 8 mm. long, with 9 or 10 segments, much smaller than P; though more or less like it distally. P,; is 4.5 to 5 mm. long, with 10 segments, and is slender and weak. The following pinnules are short and soft with short segments. Localities —Macclesfield Bank; 36-64 meters [Bell, 1894; A. H, Clark, 1912, 1913] (1, B.M.). Nha’trang, Annam; littoral; Dr. C. Dawydoff [Gislén, 1936]. Bantayan reef, Cebu, Philippines; Lawrence E. Griffin (2+, M.C.Z., 387, 396 [original Nos. 68, 99]). Siboga station 99; anchorage off North Ubian (lat. 6°07’30’’ N., long. 120°26/00/ E.); 16-23 meters; lithothamnion bottom; June 28-30, 1899 [A. H. Clark, 1918] (1, Amsterdam Mus.). Pelew Islands [A. H. Clark, 1912] (1, H.M.). Ruk, Caroline Islands [A. H. Clark, 1912] (1, H.M.). New Guinea [A. H. Clark, 1912; Hartmeyer, 1916] (1, Berl. M., 2971 [5347]). Pitilu, Admiralty Islands; Dr. G. Duncker [A. H. Clark, 1912] (1, H. M.). Ugi, Solomon Islands [Bell, 1885; A. H. Clark, 1912]. Challenger station 174 (B, C, or D); near Kandavu, Fiji (lat. about 19°06’ S., long. about 178°18’ E.); 411, 1,115, or 383 meters; coral mud; bottom temperature, at 1,115 meters, 39° F.; August 3, 1874. The specimen could not have come from any of the depths given, but was probably taken in shore collecting. [von Graff, 1887; P. H. Carpenter, 1888, 1889; Hartlaub, 1891; A. H. Clark, 1907, 1909, 1912, 1913] (1, B.M.). Lifu, Loyalty Islands [Bell, 1899; A. H. Clark, 1912, 1913] (2, B. M.). New Caledonia; M. Francois, 1894 [A. H. Clark, 1911] (1, P. M.). 434 BULLETIN 82, UNITED STATES NATIONAL MUSEUM Mer, Murray Islands, Torres Strait; among corals on the southwestern reef; H. L. Clark, October 14, 1913 [H. L. Clark, 1915, 1921; A. H. Clark, 1918] (1, M. C. Z., 600). Banda Sea; Semmelink, 1881 [P. H. Carpenter, 1881, 1883, 1888; Bell, 1882; Hartlaub, 1891; A. H. Clark, 1907, 1909, 1911, 1912, 1918] (1, L. M.). Cocos Island (in the Indian Ocean south of Sumatra); Doctor Wood-Jones [A. H. Clark, 1912, 1913] (1, B. M.). Singapore; Svend Gad [A. H. Clark, 1909, 1912] (1, C. M.). Same (4, C. M.). Pulau Ubin, Singapore [A. H. Clark, 1912, 1913] (1, B. M.). Blakang Mati, near Singapore; 1899 [A. H. Clark, 1934] (1, Raffles Mus.). King Island, Mergui archipelago; sublittoral [Bell, 1888; P. H. Carpenter, 1889]. Ceylon Pearl Oyster Fisheries station XX XIX; southern coast of Ceylon from 2 miles south of Point de Galle westward to Gallehogalle Bank; 29-55 meters; bottom, fine sand; stones and nullipores on the bank [Chadwick, 1904; A. H. Clark, 1912; H. L. Clark, 1915]. 2Ceylon Pearl Oyster Fisheries station XXXIV; southern coast of Ceylon; various parts of Welligam Bay; 4-13 meters; bottom, shell sand and a little mud; temperature at 7 a. m., 77.8° F.; specific gravity 1.0225 (Chadwick, 1904; A. H. Clark, 1912; H. L. Clark, 1915]. Mabahiss station 10; Red Sea (lat. 13°31’00’’ N., long. 42°31’00”’ E.); 55 meters; September 17, 1933 [A. H. Clark, 1936] (1, B. M.). No locality [A. H. Clark, 1912] (2, U.S.N.M., 35237 [from Berl. M.]; Berl. M., 5287, 5288). Geographical range.—From the Macclesfield Bank and the Philippine Islands to the Pelew and Caroline Islands, New Guinea, the Admiralty and Solomon Islands, Fiji, the Loyalty Islands, and Torres Strait, westward to the Red Sea. Bathymetrical range-—From the shore line down to 55 (?64) meters. Most of the specimens have been taken in shore collecting. History.—This species was first described as Antedon spicata by Dr. P. H. Carpenter in 1881 from a specimen in the Leyden Museum that had been collected by Semmelink in the Banda Sea. In October 1882, Prof. Jeffrey Bell published a specific formula for this form, which was emended by Carpenter in April, 1883. In 1885 Professor Bell recorded it from Ugi (or Ughi) in the Solomon Islands, and in 1888 he said that it had been collected by Dr. John Anderson in the Mergui Archipelago. In the Challenger report on the myzostomes published in 1887 Prof. Ludwig von Graff mentioned Antedon tuberculata from station 174 as a host for myzostomes. The name, supplied by Carpenter, appears as a nomen nudum. In the Challenger report on the comatulids published in 1888 Carpenter described Antedon tuberculata as a new species. He said it has many points of resemblance with spicata, and he suspected that it might prove to be a synonym of the latter. Carpenter did not discuss spicata in detail in the Challenger report, as it was not taken by the expedition, but he inserted it in the key to the species of the Palmata group, including it with tuberculata and indica in a special section set apart by the possession of IIBr and IIIBr series, and sometimes additional division series, by having the division series free laterally, and by having P» stiff and styliform, with 12-18 A MONOGRAPH OF THE EXISTING CRINOIDS 435 much elongated segments. He included tuberculata and spicata in a group character- ized by postradial series with marginal projections and by having P; not greatly shorter than P2; indica, with which tuberculata and spicata were contrasted, was said to have the margins of the postradial series smooth and P,; considerably shorter than P,. Later in the text he said that indica differs from spicata and tuberculata in the slighter development of marginal processes at the bases of the rays. The difference between tuberculata and spicata was said to be that tuberculata has XL cirri and the the IBr, (axillaries) more than twice as long as the IBr,, while in spicata there are XXV cirri and the [Bry (axillaries) are less than half again as long as the IBr;. Car- penter included marginata with clemens (=Heterometra quinduplicava) in a special section of the Palmata group characterized by the possession of IIBr series, but no further division, and therefore in the key did not compare it directly with spicata or tuberculata. In 1889 Carpenter recorded and gave notes upon a single specimen from King Island in the Mergui Archipelago and referred it to spicata. He again compared spicata with tuberculata and indica and said that in the latter marginal projections seem to be absent from the sides of the division series. In this case he seems to have interpreted indica wholly on the basis of Smith’s figure, as he did when preparing the key to the species of the Palmata group in the Challenger report; but in other cases his references to indica indicate that he considered as belonging to that species the form herein described as Stephanometra protectus (see p. 443). In his key to the species of the Palmata group published in 1891 Dr. Clemens Hartlaub placed spicata, tuberculata, and indica among the species characterized by the possession of IIIBr and sometimes additional division series. The two first are included in a section characterized by the possession of marginal processes on the elements of the division series, while indica was said to have no marginal processes on the division series, 30 arms, and P, with more than 12 segments. He grouped spicata and tuberculata in a division characterized by haying P; of the same character as P, and only slightly or not at all smaller, contrasting them with monacantha (=protectus) in which P; is much shorter than P, and is not stiff and styliform; spicata he said has P, with 16-20 segments and the outer pinnules long and filiform, while in tuberculata P, has less than 16—usually about 12—very long segments. Hartlaub had examined the type specimen of spicata at Leyden, and also a specimen that he identified as tuberculata—evidently the one from Ruk, as that is the only one of the three in the Hamburg Museum in which P, resembles the succeeding pinnules more than it does P,—and compared the forms in considerable detail. In my first revision of the old genus Antedon published in 1907 spicata and tuberculata were referred to the genus Himerometra, from which they were removed to the new genus Stephanometra in 1909. Later in 1909 I recorded Stephanometra marginata from Singapore, my records and notes being based upon a misidentified specimen of this species. In 1911 I published a redescription of the type specimen of Antedon spicata, which I had examined in the Leyden Museum the previous year, and in another paper I recorded, as Stephanometra sp., a specimen from New Caledonia. 436 BULLETIN 82, UNITED STATES NATIONAL MUSEUM In a paper on the crinoids of the Hamburg Museum published in 1912 I recorded and gave notes on specimens of Stephanometra tuberculata from Ruk, Pitilu, and the Pelew Islands. In a paper on the crinoids of the Berlin Museum published in 1912 I recorded a specimen of Stephanometra tuberculata from New Guinea and two without locality, giving notes on all three. In my monograph on the crinoids of the Indian Ocean published in 1912 I recog- nized Stephanometra spicata and S. tuberculata. The type specimen of spicata was redescribed. In a paper on the crinoids in the British Museum published in 1913 notes on the type specimen of tuberculata were given, and also on the specimen of Antedon ?spicata from the Macclesfield Bank and of Antedon tuberculata from Lifu recorded by Bell. Other specimens were recorded from Pulau Obin, Singapore, and from Cocos Island. In 1915 Dr. H. L. Clark described Stephanometra stypacantha from Mer Island, Torres Strait. In the report on the unstalked erinoids collected by the Siboga published in 1918 I placed tuberculata, stypacantha, and (erroneously) marginata as synonyms under spicata. I recorded and gave notes on a specimen from station 99. In 1924 Dr. Torsten Gislén gave a detailed account of the structure of Stephano- metra spicata, supplementing this in 1934, and in the latter year the author recorded a specimen from Blakang Mati, near Singapore. In 1936 Dr. Gislén recorded and gave notes upon a specimen from Annam. STEPHANOMETRA INDICA INDICA (Smith) Puate 51, Ficures 233, 234; Prars 91, Fiaurms 449, 450 Comatula indica Smirn, Ann. Mag. Nat. Hist., ser. 4, vol. 17, 1876, p. 406 (description; Rodriguez). — Smrru, Phil. Trans. Roy. Soc., vol. 168, 1879, p. 564 (description; Rodriguez), pl. 51, figs. 3, 3b (but not fig. 3a] —A. H. Cuark, Proce. U.S. Nat. Mus., vol. 40, 1911, p. 3 (history); Crinoids of the Indian Ocean, 1912, p. 3 (same). Antedon indica P. H. Carrentsr, Proc. Zool. Soc. London, 1882 (1883), p. 746 (specifie formula) ; Challenger Reports, Zoology, vol. 26, pt. 60, 1888, pp. 35, 54, 210, 225, 232, 233, 366, 379; Journ. Linn. Soe. (Zool.), vol. 21, 1889, pp. 310, 311 (comparison with A. spicata) —HarTLavps, Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 39 (in key), p. 58 (in second section of species of the Palmata group with spiny lower pinnules).—CxHapwick, in Herdman, Report Ceylon Pearl Oyster Fisheries, pt. 2, Suppl. Rep. 11, 1904, p. 153 (occurs at Ceylon), p. 154 (Sta. LIV).— HAMANN, Bronns Klassen und Ordnungen des Tier-Reichs, vol. 2, Abt. 3, 1907, p. 1581 (listed). — A. H. Cuark, Crinoids of the Indian Ocean, 1912, pp. 34, 39, 40 (identity of previous records) ; Smithsonian Mise. Coll., vol. 61, No. 15, 1913, p. 83 (of Bell, 1899=Lamprometra protectus {i. e., palmata]; of Bell, 1902=Comaster gracilis). Antedon marginata Cuapwick, Journ. Linn. Soe. (Zool.), vol. 31, 1908, p. 45 (Suez Bay, 10 fathoms; characters of the specimens).—A. H. Cuarxk, Proc. U. 8. Nat. Mus., vol. 40, 1911, p. 5 (identity of Chadwick’s Red Sea record); John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1936, 1937, p. 89 (identity). Antedon palmata Bex, Trans. Linn. Soc. (Zool.), ser. 2, vol. 13, pt. 1, 1909, p. 20 (Farquhar Atoll, North reef) —A. H. Crarx, John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1936, 1937, p. 89 (identity). Antedon spicata Brevi, Trans. Linn. Soc. (Zool.), ser. 2, vol. 13, pt. 1, 1909, p. 20 (Seychelles, 39 fathoms). A MONOGRAPH OF THE EXISTING CRINOIDS 437 Antedon monacantha Harruaus, Nachr. Ges. Gottingen, May 1890, p. 179 (in part; specimen from Torres Strait); Nova Acta Acad. German., vol. 58, No. 1, 1891, p. 61 (specimen from Torres Straic; comparisons), pl. 3, fig. 33; Mem. Mus. Comp. Zool., vol. 27, 1912, p. 412 (specimen from Torres Strait). Himerometra indica A. H. Cuarx, Smithsonian Mise. Coll., vol. 50, 1907, p. 356 (listed). Himerometra monocantha A. H. Cuark, Smithsonian Mise. Coll., vol. 50, 1907, p. 356 (in part; listed). Stephanometra indica A. H. Cuarx, Proce. Biol. Soc. Washington, vol. 22, 1909, p. 10 (listed); Proce. U.S. Nat. Mus., vol. 40, 1911, p. 6 (Antedon palmata recorded by Bell from southeast Africa is this species), p. 8 (occurs on the southeast coast of Africa), p- 13 (common to southeast Africa and Ceylon), p. 26 (synonymy; localities; description of a specimen; comparisons); Bull. Mus. Hist. Nat., Paris, 1911, p. 252 (Madagascar; description); Smithsonian Mise. Coll., vol. 60, No. 10, 1912, p. 21 (discussion of Hartlaub’s specimen from Torres Strait) ; Crinoids of the Indian Ocean, 1912, p. 3 (history), p. 11 (represented in the Ceylon region), pp. 34, 40, 41 (identity of previous records), p. 135 (synonymy; summary of records); Smithsonian Mise. Coll., vol. 61, No. 15, 1913, p. 29 (published references to specimens in the British Museum; Rodriguez; Seychelles [specimens from Tuticorin and the Maldives are protectus]); Unstalked crinoids of the Siboga-Exped., 1918, p. IX (range), p. 94 (in key; range), p. 97 (synonymy; notes; Torres Strait and Madagascar [specimens from station 301, Ceylon, and Mer are protectus}).—H. L. Cuark, Spolia Zeylanica, vol. 10, part 37, 1915, p. 93 (occurs at Ceylon).—Guisiin, Kungl. Fysiogr. Sallsk. Handl., new ser., vol. 45, No. 11, 1934, p. 20——A. H. Cuarx, John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1936, p. 87 (listed), p. 88 (Mabahiss stations 10 M. B. I. d.; notes; range; comparison with protectus), p. 104. Stephanometra monacantha A. H. Cuarx, Proc. Biol. Soc. Washington, vol. 22, 1909, p. 10 (in part; listed); Vid. Medd. Naturh. Foren. Kgbenhavn, 1909, p. 169 (specimen from Torres Strait) ; Mem. Australian Mus., vol. 4, pt. 15, 1911, pp. 718, 769 (record from Torres Strait); Die Fauna Siidwest-Australiens, vol. 3, Lief. 18, 1911, pp. 440, 443 (record from Torres Strait); Internat. Revue gesamt. Hydrobiol. und Hydrogr., 1915, pp. 228 ff. (record from Torres Strait); Smith- sonian Mise. Coll., vol. 60, No. 10, 1912, p. 21 (Hartlaub’s specimen from Torres Strait appears to be nearer indica than monacantha [= protectus)]). Lamprometra callipecha H. L. Cuarx, Carnegie Inst. Washington Publ. 212, 1915, p. 104 (description; Mer).—A. H. Cuarx, Unstalked crinoids of the Siboga Exped., 1918, p. 100, footnote 1 (in synonymy of Lamprometra protectus [=palmata}). Stephanometra callipecha H. L. Cuarx, The echinoderm fauna of Torres Strait, 1921, p. 8 (collected by the Carnegie Exped., 1913), p. 21 (Mer; possibly is S. indica; notes), pp. 192 ff. (range), pl. 1, fig. 9 (colored), pl. 36, figs. 2, a-e. [Stephanometra] callipech H. L. Cuarx, The echinoderm fauna of Torres Strait, 1921, p. 22, footnote 1 (typographical error). Stephanometra indica indica A. H. Cuark, John Murray Exped. 1933-34, Sci. Reports, vol. 4, No. 4, 1936, p. 88 (comparison with S. 7. protectus), p. 100 (range). Diagnostic features—The outer cirrus segments are wholly without dorsal processes, being merely more or less sharply carinate in the middorsal line; P, is flex- ible, flagellate, delicate, and slender, much smaller and weaker than P,; P, is very long, much longer than is S. protectus, enlarged, though not excessively so, stiffened, distally gradually tapering to a fine stiff point, and recurved, and is composed of 16-20 (usually 17-20) segments, of which the fourth and fifth are not abruptly longer than the others; P; is small, flexible, and weak, like the pinnules following; there are 18-30 arms 65-153 mm. long; and the cirri have 22-25 segments. Description.—The centrodorsal is small, flat or convex, more or less subpentagonal, less than 5 mm. in diameter, with the bare dorsal pole about 3 mm. across. The cirrus sockets are arranged in from two to three crowded alternating marginal rows. The cirri are XXI-XLV, 22-25, about 12 mm. long. The first three segments are about as long as broad or slightly broader than long, the fourth-eighth are longer 438 BULLETIN 82, UNITED STATES NATIONAL MUSEUM than broad, the longest, the fifth or sixth, being about half again as long as broad, and those following are broader than long. The longer proximal cirrus segments are slightly constricted centrally. The distal portion of the cirri is laterally compressed, and the distal segments bear a low dorsal median keel. There are no dorsal processes. The opposing spine is well developed. The radials and division series resemble those of S. 7. protectus (page 443); the I11Br series are normally externally developed, but in one specimen examined they are internal. The ossicles of the division series and first two brachials have more or less developed ventrolateral processes. The 18-30 arms are 65-153 mm. long. The smooth and short brachials are at first practically oblong, then becoming obliquely wedge-shaped, almost triangular, and after about the thirtieth oblong again, and terminally longer than broad. Syzygies occur between brachials 34-4, from between brachials 14+-15 to between brachials 18+-19, and distally at intervals of 6-11 muscular articulations. Sometimes there is an additional syzygy between brachials 9+-10. P, is about 6 mm. long, with about 20 segments, and resembles P, in S. 7. protectus. P, is much smaller than P,, with about 11 segments. P» is long, stiff, and spinelike, becoming slender distally, 7-17 mm. long, with 16-20 segments. The first two seg- ments are short and those following are elongate, those in the middle of the pinnule being the longest. Ps is small and weak, the smallest pinnule, with about 10 segments. The pinnules following gradually increase in length, those in the middle of the arm being 6-7 mm. long, with about 20 segments. Notes.—The material from Mabahiss station 10 consists of arm fragments and a detached disk representing two individuals. The specimen from motorboat station Id has 20 arms; the cirri have 17 or 18 segments; P, is 7 mm. long with 14-16 segments and becomes slender and delicate terminally; on one arm P; resembles Po. Two small and imperfect specimens from Suez Bay were referred to marginata by Chadwick with some little doubt. He said that as in the type specimen the centro- dorsal is a saucer-shaped disk marked with cirrus sockets almost to the center of the dorsal surface and bears the cirri in an irregular row around its margin. The cirri are XVII, 15-17 or even 19, the segments gradually increasing in length from the first to the seventh, beyond which they are scarcely longer than broad and become increasingly compressed, and are slightly carinate dorsally or even bear a minute dorsal spine. The radials are distinctly visible and are not in contact except at their proximal ends. In the type of marginata they are only just visible, but this point of difference, Chadwick said, is possibly due to the immature condition of the Suez specimens. The IBr series agree with those of the type. In one of the specimens only IBr series are present. In the other there are two IIBr ossicles on one branch of each postradial series, and in the case of one postradial series two IIIBr ossicles. Syzygies occur between brachials 3+4, 9+10, and 16+17, and distally at intervals of five or six muscular articulations. The proximal pinnules have fewer segments than those of the type of marginata. P, has 14 segments, and P, has 14 0r 15. Py is little larger than P,, but has 15 or 16 segments, while P, has 11 or 12 segments and is much smaller than P,. The specimen from the Seychelles is small. A MONOGRAPH OF THE EXISTING CRINOIDS 439 In the specimen from Madagascar the cirri are XXX, about 22, and smooth like those of S. 7. protectus. The sixth segment is the longest, half again as long as broad. The longest cirrus segments are slightly constricted centrally. There are about 25 arms. The IIIBr series are in all cases developed internally instead of externally as is usual in the species of Stephanometra. Pz is long, stiff, and spinelike, with about 16 segments. P3 is small and feeble. Dr. E. A. Smith described Antedon indica from Rodriguez in 1879 as follows: Rays 30; dorsal disk, small, convex, subpentagonal, sculptured with little contiguous and well- defined pits which are the sockets of the cirri; these are about 45 in number; radial] joints two, the axillary without a syzygium; between this and the next bifurcation are two articulations; after two more joints the two outermost arms bifurcate; none of the brachial axillaries with a syzygium; each third segment above an axillary has a syzygium, and on the arms the next joints with syzygies are at very unequal distances, the most proximate being separated by as many as 20 joints, but usually by about 14; then nearer the extremity of the rays, they are rather closer together, the interlying joints varying in number from 6 to 10. The second pinnulae are very long and composed of much elongated joints. The colour of the specimen is pale purplish brown, with the sutures of the ray-articulations blackish brown. Diam- eter of disk 6 mill., length of rays about 6 inches (153 millim.) * * * The remarkably elongate second pinnulae (17 mill. in length) are situated on the first and second segments above those joints which always have a syzygium, and which are the third joints above the last brachial axillaries, or in other words, they rest upon the fourth and fifth joints above these axillaries. They consist of about 17 joints, which are elongate, cylindrical, those at the middle being the longest, and the two or three basal ones the shortest. Dr. Smith called attention to the fact that in the preliminary description pub- lished in 1876 he had, by an oversight, stated that the two innermost arms following the first axillaries bifurcate; it should have read outermost or exterior—that is, the IIIBr series are developed externally, not internally. He figured a cirrus that was found among debris at the bottom of the bottle containing the specimen, but he said it is not absolutely certain that it belongs to the species described, although the evidence points to that conclusion. The cirrus he figured belongs not to S. indica but to some species of the family Comasteridae. I examined this specimen at the British Museum in 1910. It is best described as resembling a specimen of S. protectus, but with P, longer and composed of more numerous segments; though slender distally, P, is not flagellate. The specimen recorded by Chadwick from Ceylon had 29 arms. The original visceral mass was almost completely displaced but remained in organic continuity with a new one in an early stage of formation. Dr. Hubert Lyman Clark described Lamprometra callipecha from Mer Island, Torres Strait, in the following terms: The centrodorsal is rather small, flat, less than 5 mm. in diameter, with the bare dorsal pole 3 mm. across. The cirri are XXXI, 23-25, about 12 mm. long. The first-third segments are short and thick, the fourth- eighth are longer than broad, and those following are broader than long. The cirri are laterally compressed distally, and the distal segments have low and minute longi- tudinal dorsal ridges. The opposing spine is well marked. The 18 arms are about 65 mm. long by 10(?) mm. broad. There are four I] Br and four IIIBr series present, all 2. The IIBr series are not appressed; the ossicles of the division series have slight lateral expansions. The brachials are smooth, not overlapping, wedge-shaped, becoming 440 BULLETIN 82, UNITED STATES NATIONAL MUSEUM quadrate after the thirtieth, and at the tip of the arm longer than broad. Syzygies occur between brachials 3+-4, 9-+10, 14+15, 18+19, and 23+-24, and then at intervals of 8 to 11 muscular articulations. All the pinnules are cylindrical, and the lower ones are stiff and little flagellate. P, is about 6 mm. long and consists of 19 segments. P, is much smaller, little longer than Ps, and consists of 11 segments. P, is the largest pinnule, 7 or 8 mm. long, and consists of 18 segments. P, is longer than P,, and con- sists of 11 segments. P is the smallest pinnule and consists of 10 segments. P, is very similar to P;. The succeeding pinnules gradually increase in length until at the middle of the arm they are 6 or 7 mm. long, with about 20 segments. The color is rich purple, with the base of the arms and a broad distal band white, the pinnules often with yellowish or rusty tips. The cirri are cream color dorsally and deep brown on the ventral side. Dr. Clark said that the colors were almost perfectly preserved in alcohol. The specimen from Torres Strait included in the material from which Hartlaub described Antedon monacantha has about 20 arms. The second syzygy is between brachials 16-+-17 or 17-+18, and the distal intersyzygial interval is six or seven muscular articulations. While in the specimens from Mortlock Island (=protectus) P2 is very stout and strong and is composed of only 12 segments, in the present specimen from Torres Strait it is much more slender, with about 20 segments, of which the fourth and fifth are not set off from those succeeding by their greater length. Habits —Dr. Hubert Lyman Clark said that the type specimen of Stephanometra callipecha proved to be an active individual and was a very graceful swimmer. Localities —Mabahiss station 10; Red Sea (lat. 13°31’00’’ N., long. 42°31’00/’ E.); 55 meters; September 17, 1933 [A. H. Clark, 1936] (fragments, B. M.). John Murray Expedition to the Indian Ocean, 1933-1934, motor boat station Id; Red Sea; bay between Great Hanish and Suyul Hanish Islands (lat. 13°39/30’ N., long. 42°43/00’’ E.); 26 meters; sand, shells, and coral; September 17, 1933 (1, B. M.). Suez Bay; 18 meters; mud [Chadwick, 1908; A. H. Clark, 1909, 1911, 1912). Seychelles; 62 meters [Bell, 1909; A. H. Clark, 1911, 1912, 1913, 1918] (1, B. M.). Madagascar; M. Grandidier, 1905 [A. H. Clark, 1911, 1912, 1918] (1, P. M.). Rodriguez; Henry H. Slater, Transit of Venus Expedition, 1874 [Smith, 1876, 1879; P. H. Carpenter, 1883, 1888, 1889; Hartlaub, 1891; A. H. Clark, 1907, 1909, 1911, 1912, 1913, 1918] (1, B. M.). Farquhar Atoll, north reef [Bell, 1909; A. H. Clark, 1911]. Ceylon Pearl Oyster Fisheries station LIV; in the northern part of the Gulf of Manaar, south of Adam’s Bridge ; 7-73 meters; bottom varied, from sand to living coral (Chadwick, 1904; A. H. Clark, 1911, 1912, 1918; H. L. Clark, 1915). Mer Island, Murray Islands, Torres Strait; southwestern reef; H. L. Clark, October 27, 1913 [H. L. Clark, 1915, 1921; A. H. Clark, 1918] (1, M. C. Z., 592). Torres Strait [Hartlaub, 1890, 1891; A. H. Clark, 1909, 1911, 1912, 1918}. Geographical range-——From Madagascar, the Seychelles, and the Red Sea to Ceylon; Torres Strait. Bathymetrical range.—Littoral and down to 62 (?73) meters; most of the records are from the shore line. A MONOGRAPH OF THE EXISTING CRINOIDS 441 History.—This species was originally described in 1876 under the name of Coma- tula indica from a single specimen collected at Rodriguez by Henry H. Slater while on the Transit of Venus expedition in 1874. It was redescribed in practically the same terms and figured by Dr. Smith in 1879. Dr. P. H. Carpenter published a specific formula for Antedon indica in 1883, and in the Challenger report on the comatulids published in 1888 he mentioned it several times, though he added no new information in regard to it. In the key to the species of the Palmata group he gave as the chief distinguishing characters: “Margins of rays smooth; third pinnule considerably shorter than the second,” contrasting it with tuberculata and spicata in which the “rays have marginal projections; third pinnule not greatly shorter than the second.’’ In 1889 Carpenter compared indica in some detail with spicata. In 1890 Dr. Clemens Hartlaub described Antedon monacantha (=protectus), which was based on two specimens from Mortlock Island and one from Torres Strait. In 1891 Hartlaub redescribed and figured this new species, his figure of the proximal pinnules being drawn from one of the two specimens from Mortlock Island and his figure of the centrodorsal and arm bases being drawn from the specimen from Torres Strait. This last specimen Hartlaub had received for examination from Dr. P. H. Carpenter, but he said nothing further regarding its origin. In 1904 Herbert Clifton Chadwick recorded a specimen of Antedon indica from Ceylon. There is a possibility that this specimen in reality represents S. protectus. In my first revision of the old genus Antedon published in 1907 indica was assigned to the new genus Himerometra, and in my revision of the family Himerometridae pub- lished in 1909 it was transferred to the new genus Stephanometra. In 1908 Chadwick recorded and gave notes on two specimens from Suez Bay, which he recorded under the name of Antedon marginata. In 1909 Prof. F. Jeffrey Bell recorded Antedon palmata from Farquhar Atoll, and A. spicata from the Seychelles in 39 fathoms. There is a specimen of S. indica from the Seychelles in 34 fathoms in the British Museum on which apparently the latter record was based, at least in part. I have a note that Bell’s A. palmata is the same species, but no record of having seen the specimen or specimens. Regarding A. spicata Professor Bell said that he was in some doubt as to the correctness of the identification. He noted that P. H. Carpenter described this species in 1881 from, it would seem, a single specimen, and that no subsequent writer that he knew of had ever mentioned it. In a discussion of Stephanometra monacantha (=protectus) published in 1909 in a paper on the crinoids of the Copenhagen Museum, I said that the specimen therein described from the Nicobar Islands (see page 452) agrees more nearly with Hartlaub’s specimen from the Caroline Islands (=protectus) than with the one from Torres Strait (=