.. Db O + 6b B= AA AO OO nk OD Ob ree yd Ued OM AOA HO. ener aoe: peoee dota aeanceate ta eae ee eee Oe anent it 0 ot 0-4-4 9 Ae Stepp pire rte per itaay = atatatate eels stats steerer steteeeat Aa tae 09 ts Doha ah ey eeey Serpe OPTI aa Wh ah hE aE Hie 2 ch ecgct at arbeaewre Ft LAG ISN AIAG O04 D4 dead bearbed ein) Whale 19 cree iebe este OG ae Atha eae oe Otarich opiate, hcp eeaceetogs oro-9 Sade ethene teae 0-0-0. 058205000 910-09 9450 @008 Sy dadadedrery tad top oeyeee ey Waa prdea mete tasaratadeaees ghey oeceoeton SOA 0 90 99 0590-4 2h SPOS 024-00 SY ALIS DY DOOM Mt VOL 0 Ord Oi AO OOO EERE eee ee 0 9 305099 50 091 end hoes pepeareddiedPerchbraree merase ett ean > Didghe pedebee ea derpys weep ee eee bases nape isnsytatasarpaw gitar pei Raiamastistcaweet teats aaa series tot 30.8 wobtetakecen$ So nad tions pohh see apeema rte eeb prea tales nie lglo Mieco dmneoaeee ait erent es 10999109 oe delat ppp erntaote coe Ore teat st tata 29 Pose babrenbies 51 ho9 9 aso -9ebed obs ates, RAqaobroxe-ba-ese v0; SSeS LOS: Shaxeee ieee Sseepaaereenes Pry ert 1 cayaopeeieiy exes peeks oot, saeasestststsecs eonitosessteteestea seat Shani eRHREEEE, iS eye a retats ate etaeatat terete: 1@4dses00e Sky. MMO Er yeweye Wider dy. ferestrorana ps drgrsecscer bance hess cecseah tecate rovrotstsertnicetat stati pepsin Descwrn ir lvtandenteroyepepaeie ne Farsiabstersteaisaeurten at sosteioesatetette ties tat Faopiieeasaeey tien: Seid acetinragetat oh bar peumeb abe ebaees adeees MIRA Sere ody pone a yee wie pomegdeotuntery walstatatie Srecwayooceneprtigisess arene segs roa eecet 0.09 x8 eee sxe 0-0 10-00 S00 0-0-0208 9:4 Sree PD and Ws Pubcon yi We adaoneunaasiekon! titatenatareanler seats enacmssigi peacoat is hase. ueasiae eer een coat a Soest Pe oere 0 Mh ad wets ed by nd nd ww ae dree cope. ah soao4 ee aseaienbae sraichet Sareea Sard secs bcbsy ta sete ns eiraeaeeasteas S060 dt abet eth att S98 198 50 “tase 4z¢tardeacgronneaed freee. oos rhs ot atom ae pry tthe tye 4g yea reo le WSS. PAP RDS ney ey! 03 th ay CUSe eos netey ty 4b ohh ees Cs tone ot od ond ei rie phl eww besnoowe peephwe pipettes waa erty vs pyee ayy new. SS tat ee earners eae ee ee at ars cseneast arte nits SRE OmaEo ry ners Spee aiaatnd oi om tit sso siratecactai te seein erertgueneesepent ora Sop bee yy eee pepe haa tstecpemet aise cash ecaed oa acd caay aasrateee . eee pam Srarycech ebstehete baie sioaadeatsstetererses esthaeeeeel riseresewseeheees eratatansttontceates ely oa OO Se te reds ay oh <0 0-0 S00 =O P1010 0-0 One bereson eens eer a je etek te sans onus ena sganat : Teed sact sd coco eb ee’ ae oa At ag eer pssresegeesne ceases <3 tsttarsecreet yet Soph a han hataent ne aoe his yew yee sere st ey pet iw tas manaads rw vbcasaee tote estas ated tars s Selaceazesa echoes tadcrasecoepacerseca ge roreny Peteas Ne te erat man Toren een Pe a ayaa Brera se ess Syey Scteaterstieted et es pbiniioaese sh-osomibeteies estatenera rs aestese 3 “oe Bianca aiepratatestemmae aunt U aS 5 Sone ed sect oh ee oe Roe ety Stina tesps tenses ae eaartcore sess : at ep enh poets yee ient nests Sagaens aeorwtewespheo rome tes Site te Seat anata Soscaeeee arse tsa e tat eye ey ey ee esta tetpeeties psa -egrera cheneboe arava sent Rpt tehttp yeep ties be enks ape a divans ByPy > Bdge erates peat eee stseagencriset tia coctneatstatnctanats $otstces retake a ser Stes aaa setanatae stat Spasteestsomamemeancastostecarmemstceeteetaie ee 3 wh etiat ato 0247 929°2-4 2934-0 7b2 002006 P-¥ obits sn sagaceewsbeaco srocerepestcaraebee tenon os rae youdnarerth ttn chy cathe 68 were 0-6 v8 oxerend 26; seat corevesoms atihatcees PS Sine Sable cdr oe rivbeaet se se bres ye" meeras seca Berta Bs ne One oan 10 oF COS aes ptr pdr} hab setae emer arenes rk eat stead cea irae L Paar SS whey how pees hanes Nesarecheoee strand gdh st Braet BE Soest tte dary iyeea ye) tae Petbct ahnt nit ot ob pai et Hh DLE DEP ne Reba pbb Ta asats | busritae ; . seed php ornate pane ersense laasisasasion 8 84 tad ge os 198 98 Rrvoss te 2 od oh i808? eA bb ee vege siete cee varetp Bid yess ots w ¥ alto ateedebeicarsraeanseona rasa brosborrovisetatiie Sheawarmeary sacacoaseecncersintareee ye ates Pasi ledides ieereritecaaseses a iacaeanetsea aaioeenre ae fe nbalanta pos} ides) pede 28 oh ba iri ace ep Deneve eh od ds ohret oe ca seiaselestigaeg i tacesh-hetaaehs> seve eaeachebsay acing : ” Veranded trond esearch eee aie wren) Wd Pedant» serine} eo erp ee) Lr STE sie) eatiepeieeoachct es: Te; Orgetebea Srectebed ake wedetrd v6 TTA ve ee nah om pre) ; b09 0.9 hetebubsersetebehe h bras 1a phe debe ee aR Mere N EET Tr hoebd sited sishranegtioetehtregamearhe eaves Bi Sra “apbeisign Wicaresheseseararbepeaeas Viseseep wand va Fiend ovat ehh yb hy oe owha rd bea Bei 33) seo rnGeh: 8 tt wp ted aravasd betsyes v Sane ee 5 Fe aa iy anardctrhed ded apenas ach'h aban bears chtorathy Hy tet bt anh agen Bh sor a eae tte ea Pr Beran) Patt tet roe ele vitae sie-basina-drbapimn Dlocieb assis Peteewrarbenete dt hey, seri asennad taoen rosea cashed vere tteed See hah Te lepers Sy Pd 1 ee erterpans seetedeartoatetetpede he pot 8 od ont Sekeate teebstet ieee sori secdeoubes eboret eae st achemescheastebebednoengh pebes tonbeoeoee thes PUTT ustiersiietides tere ihe mnhet weemnbenewed 1 bot d shed Deg coed ebebyt re-do eh. strorresd tonbai ted “9d erarpelt say one “Gristvactraed ed cacord et th chga TS SES aS ROWS Cea uayerssandta's del Pyy ciuaca peed sich vosbed pies teeast oeoetes teresa dened et 3) vetoes Serr geee etext {¥ ectceseatasy bea saast each Ber eescer pert story statovesb's Sterhavarees ary beasorscaserees 4 od ooo pe ledasebes saan? oes hdelalasedsaeteiad oy wy eye eyes ve Bowe bP bn 4s ha oa we narbed had of 18 bed spe ts a eee Sern oergedebed 2 vorae) oh coe pea eo = mae: Prey ded conavoreed ted tock ca-aeh ac oeo wooded 124eye sede pes beargraravach avo ya ay bed taearttechvorere oe Yevepsr peg the per Rare) ReltenOuiideceise iis iii ri sie sera te bangrpehepences frarsaedarecacas a4 Pebcizh epsorers weoceek toe eared 8 ees oh htt oh 14 op mad arvbeoep rove detbed weds trveeeey seh oy ere nen nes Peaperhes more bogus 48 dno 0x0 28 8 9 beng oo Ua eweees onan eer ae ae Tah ebro ter eases POSSE ee eee eae eee a ediaekaieiad reerunrersh vy woee rere: deu aoe rps? Rr Jesees saeenaee 419-0 ee > a GePoel O24 #o bed sy Oadrde 9 44:40 00 a wd ow seta lal tmitahdendssi eee) ewer ere yeti) Sree ein) : pyre aaeeeds itor Set wards Sore beanies seed regen sates rendyis See ass irene et tee anager? secs" treet er oen eh eee nA noabeowe epppenennate oncedicanbagetey pestatst tne veorevine Deamseeoeses otaestenests soy eryrneee: Sioa 9-4-0 0.9 -b29 99 erence eT oe reeyes br dt oqseepegeere Meeps se Sadao ees Siesta sroceresoras sesestgtatsts0:078 ose contre sesesesere eee sesran rest one sarecbrache O14 sted stah eBags gnbe aed oa SO P53 shee reeataracdacerdroesisverieseand terre eerste eeetat eet perce py $00 caubehenedee debenese atatae Ssdakasy sds i seers toate ante snoeas asset Mnecn ts taeaataa tutti caecnrtmacesiaten ea th est Deeb vteb tea wry oe. gaderrdedelen sped a Aodtttanstetvanbastiaismtypsisid pobthstonsotes Dedemmibel aepeaeh cher Tartseet etet Aetetedarbed peed mwere cise Shion iien erie) dsdebsb feebsprach sorbcoeasdeary ~ toh Soschop vt ovprbrd vi ie Prasbrey et be diysd ed tryed ete oh Stet heted ba. poy ey yhreteys haa thot ha unitate) borace visdvace-bebsped perpas arene Wat geeebearard abwd-eer ce 9109 19 0998 D8 web ta minister owt cteortan sonia teweoecet: Saas sPrt PSO a-G-0-9) SOS hers alatatateitatetat iets Ne Tey. yyy ae ay roy strene ee Rp yy ty yew rte prety dnd eadgrnnmngnnmn pabaheone OSS poten Ann apne 0 5029-2 020-09 31 SHAS POSS. Wit Ab Sebi narens inter neiee ene poncnnos ae Eee eae : eroe hts ctasea tree srbnataghracncgehae oat Srectb0 O89 F-Ace-Srasted so sdrarpensdebehterecoceeesocachocbes Pep eee nee ss posaeoutebvospeacaranecbeccampravecsc geben sstoescae stamrey estan aree rarer tone! peeeesg ants rats statans tates Scoetnacoeretesuece totes eer eS Pept d ame pere Seerererees 5 erst tae seee Ose-tn 00209 sabes pe ot we sy tovth sot wate vias teer be vow wis peoratge twat yy Wepntis vite bh erwedysgarauregergstorsanqapogiey apd Sowlutaowarveanpisssetioct py tere pon otehaoh AER yr eete waspearecavnrectras rast secticsnenssreesyseacaregsencebsrid arhctee anh wecgeecaesosteen serve .: labstebiotedinat oe SE net Settee eran woes sneer 8 radune Sark wince Sm beet indao dirt scdet 19 Se non biet bee hae te} seeariea denote listers isE eee peas: eset -0 oes nerbuest oh baet baraten, sionsedetet besapbanaouheers ter Mtoe pW vbbhnnie \babeuseessmupbaaye wyesseon cesses brovaregsesataanensessanpedeoroetertslomtetient etn pier deney ort poesestgla coe geconbades ed bubvmnastirntcinetestooee porte y sorts tuners peste torienes ae seceitecsnanatatatresépbeecegseigtcac tpuoant ort iaeneee enpeamenes 3 sists bytes Scent eee tes-presteowastes Rides) PPA bray vpadis tebe rameter prerercreds, serereieees SShreee so rons 19-8 heh 0 <0 9408 1b 9 1b Mob vpede ene Pe 46 2B ot bu enh roenersia sesed 04d 9 8 de have Bu adbs H8} beds avnbebalbb-aneencnehiicag Ue aeh ence ae sehen aoe tatcdeoed whet Vy midband phe se bded “i vh-vtv vs stebeavtoa-oeseh p-achetghachetireehek ack ot maid vaea aia. G chet, deetrentiesisepessrrinsstsbae ests cdciosbctvers tesarorereeoracacieore pean otis eoebebears rhea vpehenats Sareea pear leas ycnesden esate eet trrdesy Dal ps yood strbsoemthesen sents le'ptavascespenearany “ if Teac aea sere beth 4b Speen oe porte peseeeeeneny se] Rithaslshesani nies egeane spon ; H:e-Rnneaeg upriaa) p1Re Pode Peoe ge a-hobet ote be te beat ty "etguebea bs Patgaete Sp terittsecnttet tienes optpeat ites e ecper See sce + ut ndtroteie inns eens Fo4 etrdiad oa eta bed-o-4 6 OnPuh chmibon boetate! oi tebchee—evoaben eeeereaeey Weishastrieh eben hed eomegeh eoebsttbetas o nearperteerten sb icoshnsame teed daa theteretiocerice sted eebeaeree es oh tyneee heb uped th che ben pheg reebed ef een sepst Sistas heaese bees Neieshtaceatertios ret) tase O58 9-4 4040408 0804 hed ee AaAeds Loiepwitussdiessabeie: Loe tebe tcbetereaene wares Myipigh tatedeasd beds jubedclagege short ct-iebeds esd atarbebrgvecbepehe emedeae } 0-t br) bedet petkae eeepc aeets S. “b:Pehbeopteneee Mars iewriscrgsrcnstecnt Riptehtama teres eens sere ealesgeteri er cece a8 6) bah eds beb-Debepchepey oS orepelothesemeaiese sinters RtseROgT Neca seats trons 24 VO Pahot ots anh Ohi) (7 po scents o “t+ aponebegeg. Sosa evsey: oot aetna plese ita sedels bongs oe oe pobre arrest evens pay ae eet -tooy Yori taens ted boputasebe aurea Depsbeacgnd coupes ta ber Creatrewrreitser renee <2 ferseafeenties sprite shouts nopisbosebeorseanosbate ebeoete isa sh-eenpeb-adrenpeitragpebehes tooteavoy RA Setoeer ss asesteat Satis eet eters sar p-brdseebenctet ci teenebe peeaeaeee Pes seevementeorseteds yh sa Teseaed ges tease taetose renters site reasartay ue places sae ie ae Searsigssiatebeksetitrenrtecet oe Rattamteltecrse ert in! ats Citatisiets b aeare sven ried oeber tant. oie sdivted od Foe bea eo n Ritvemee staves |otoeg mieate at Prado warily rotate sty aster erepaeetien teen tare siatispibeisanardnoysrocganriowlstos aatierotretieirteetes Paraatercaresieeltieey telee anton aaateree etretiereres basen dSatsdetenageice-ararted teorpeen Se leottehamscieri tiers ane a Se ie deamee seve we ge joe pebeveded Oos8s trang “tien eoe me ae bee oY al Ae 4ctrhegendna-a Dhndnaen w Dofeacae} send Brorboeeiey yeh sepa See BeR eeoe Bf “etree sear oo sae budbiensuelaneluiestaensese te iiascend Soxex x3 ey fees ee eset ger a orvered vane, eae ptr yishtteerws Spey y he) gases no-9 0-4-8 br bode Oot YO Pr bvinang ara ad 14 sd ewehupupaaen—ieinerd ded betenlerdeaan owebeateras aa Roster ort) “4 Aetwtes eee atbaattbase rebate eke eOeeekee Hoenn 078 habe oh 4-0-4 od wtndstvh 4-0 v=t Geer-Or- hd ro yf pemeetrepepenee ec enon heir rreny . shihove-e-opeacecerasesd Sriterraroeetresaresh spear nt pobew wh Seabee ieksebhonrasenl poe dose deebaonel terete ns pettae tata. ¢ 0-96 bed wing 1G) OP Hetng wn One te ene p-tnen ge om an gna eee fe Ohm ones, Tissritie tena: terete yeas tari enigieememootann conan estecreemsess 404 hangs ad wonpeaeene™e arereseteabtnserte went atten at Tete A) Perey ongnd wrereaney, Lee nnn SS perSceromenenens Rend teee gs teerdeesbeecbnw ct eeemycgrartees MepMeebnauet ooteseoesebeesthtest a sent eee eed Saegebee wees ma neee) Ring-o beg bre-o-subemracarere-eroeg tah 14-4 np me Ang NG Oohaemhnd br epepigeans aa TOe Swe ory bm pere erence ary ssraecdsss ed eonacanton estos Spe a3h ; : : mais, Casbeicheratgetioestosaasoes So ; : eae EEE x Bie ore see eran rat b-4-4 none he. eeinerneee" esetens giytetvosenenoreractss s7 cane ony i tas en ene ane aedre se “pms: ee a eee we wee SELES tee a leat pt otmane ewetentnte@ bo a-eee paren et yh i er) Fen Getabed 2 > Petter bo ercitcorsser) pooearesy aor rents apareperrite eee Dantas ret a ree ek yo aind-ecere dyastearb b-0canbOcbtend taba te-a-teacdsashsbeargrecdsoamareeh ne tebeenne’ pptheerreresnrefarrnoearertroreenssd] Cites sewer st eer wee nd Br pb Scents sen +09: 48.0-0: 0-0-8r@ trteaabe renin best be beba rore-aes. oc peteb bares Bat SoCs serpreves F ar mabwet ae ears rerares aeteekon vo sorashwses penonbet, vbr oeipeeastepeees SS peheerdabe hea are bekee : - oepestorerearenes beisitsescorihioe en nreter nae aww © EET eta sore goca Slupessreedebrcbetrae cnchinek ences age rbearebrerhsvrbete bee CUS CU ie Crhcirerreesceestvenrsrartoes peer oee- eo. aware eats A eevt ny. STRUCTURE AND DEVELOPMENT OF THE SKULL IN THE BATRACHIA. PART III. BY WILLIAM KITCHEN PARKER, F.R.S. @ From the PHILOSOPHICAL TRANSACTIONS OF THE ROYAL SOCIETY.—Parr IJ, 1881. Aceanion iy tins oS Ca “tN a) GILL 4) COLLECT! Voip. 30,0 AMS Hiei Aly LONDON: 5s 3 2 = a 5 a t g Z ae PHILOSOPHICAL TRANSACTIONS. I. On the Structure and Development of the Skull in the Batrachia.—Part UL. By Wru1AM Kircuen Parker, F228. Received April 29,—Read May 27, 1880. [Puates 1-44. | INTRODUCTION. My first attempt at working out the morphology of the Batrachian skull (Phil. Trans., 1871), instead of satisfying my mind, only served to increase tenfold the desire to know the meaning of the mysterious changes undergone by that part—the main part —of the organisation of the Frog. Since then no opportunity has been lost of laying up in store fresh and fresh materials for further work in this field of research. Moreover, an additional strip, of ground has since then been cleared and cultivated (Phil. Trans., 1876); in that second essay I was greatly helped by Professor HuxLEy, who showed me what was wrong in the first attempt, and also cut through some of the thickest and thorniest parts of this tangled subject. I am also indebted to him for materidls, and also to Professor A. AGAsstz, and Mr. Garman (of Harvard University, U.S.) ; also to Professor Rurperr Jones, Dr. Gtntuer, Professor W. H. Ftowrer, Dr. Murtn, Dr. Doxsson, Mr. T. J. Moore (of Liverpool), W. FeErcuson, Esq. (Ceylon), James Woop-Mason, Esq. (Calcutta), Arrep C. Happon, Esq., and George Dives, Esq. Primarily, the aim of this extended research into the meaning of the skull in one “Order,” or main group, is to get light upon the great cranial problem. A second use will be to rectify the classification of the group itself. To make a systematic classification of these metamorphous animals upon such MDCCCLXXXI. B 2 MR. W. K. PARKER ON THE STRUCTURE AND characters as are easily seen by the Zoologist is to build upon a sandy foundation ; none will be more ready to acknowledge this than those who are most familiar with the group. Even the cranial characters, which lie deeper down, are extremely variable, so that in formulating them for any division or sub-division of the Order it is always necessary to give some qualification of the scheme, and to say that “as a rule” such and such modifications of structure exist in the group under notice. Our “genera” suffer from this weakness; and in some cases, notably in the great genus Rana, there are morphological modifications and variations such as are not to be seen in whole groups of Families in the Osseous Fishes. On the whole, my own views correspond very accurately with those of my friends Messrs. GUNTHER, Mrvart, and WALLACE (whose works are referred to in the biblio- graphical list); they, I am satisfied, will be struck with the evidence here shown of the common origin of groups of the Batrachia that now are very widely dispersed, and marked by every variety of external character. For there can be no doubt that these curious fishy air-breathers are, as Mr. WALLACE has suggested to me, a very ancient kind of Vertebrates. They have not struggled for life through one but through many epochs; they have been put to every kind of shift to live, and with infinite readiness and adaptability they have become all things to all conditions. Here, undoubtedly, we get light upon the mystery of the great perfection of the various organs seen in the members of so lowly a group ; for they are mere anamuniotics at the best, and their upspring has been from some of the lowest of the Vertebrate stocks, In every kind of facility for motion, in organs of sense wonderfully perfect, in power of speech and of song, and in instincts and habits innumerable, the Frogs and Toads teach the order, and anticipate the life of the peopled kingdoms of the nobler tribes that have risen above them in the scale. The Salamandrian tribes (‘ Urodela”) have branched up and beyond the “ Dipnoi” (Lepidosiren, Ceratodus), and more or less, as a rule, lose their gills after a time, acquiring, in each type, a fenestral passage and a stapedial plug to their ear-capsule. They also, in harmony with their more and more terrestrial habits, acquire a rudimentary larynx, so that the beginnings of the better kinds of organs of hearing and of voice are found in them. , But the Batrachians, springing from another part of the “stock,” and indeed from a far lower “ node,” rise high above the Salamandrians in the metamorphosis of their organs, especially those of voice and hearing; their general intelligence and their gymnastic powers are also of a much higher kind. Supposing the Urodeles to have arisen from some archaic forms of the ‘ Dipnoi,” the height in the scale of such double-breathing Fishes must have corresponded very closely with the Crossopterygii, and these again are manifestly a mere subdivision of he great ‘ Ganoid ” order. the great “G d” ord DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 3 But the ideal Protobatrachians came up by the way of the Sucking Fishes (“ Marsipo- branchii”), and grazing the edge of the Chimeeroids and ordinary Selachians, developed into the larval form (or Tadpole) first. Iam under the impression that in many cases the tailed gill-bearing condition has only lately been completely departed from, and that the anwrous form or stage of such a type as Pseudis has only become universal in the newer geological times. Whilst these tailed forms—hypothetically the primary condition of the Batrachia —have been yielding to terrestrial influences, and undergoing more and more cur- tailment and general metamorphosis, they have also, from time to time, dropped out of their organisation much of their old bony armour; and that which has been retained has become less and less superficial (or “ dermosteal ”). They have to a remarkable degree lost their mandibular, and in many cases their maxillary teeth also; this has been evidently a correlate of the development over the ventral end of the hyoid arch of a peculiar cushioned fold of the floor of the mouth—the tongue. As this fold has become a more and more effective prehensile organ, the teeth have become less and less useful to the creature, whose succulent “articulate ” or “molluscous ” food is caught suddenly, and swallowed whole. The size of these types has evidently undergone a steady secular diminution ; this, and in many cases an exquisite specialisation of the fingers and toes, has all been in their favour ; up high in the trees of the forest, and variously painted to resemble their surroundings of bark, leaves, and flowers, these marvels are wrought in them by the “ Archchemic Sun ;”—their small size, I say, their curiously mimicking coat, and their high nestling, give them a chance of life and of life’s enjoyments equal to that given to any tribe of animals whatever. As in the common living forms now, the dilemma for the larval Batrachian is to “¢transform or perish” in short annual periods; so, I opine, in long-past secular periods, again and again, those tribes—the forefathers of our existing kinds—havye been put to the same extremity of shifting for their lives. As the earth was made to be inhabited, and as the evolution of its tribes takes place through the harmonious inter-action of the forces within the organisms and the influences surrounding them, it has come to pass again and again that the extremity of some archaic form has been Nature’s opportunity; its threatened extinction has been the occasion of its transformation into a higher kind of being, and the drying up of the old waters has been followed by the peopling of the new land with fresh and fresh forms, enjoying in many ways “ newness of life.” There are two main groups of dwarf forms, namely, the highly developed and typical Tree-frogs, and the arrested, low kinds, such as the “ Engystomide” and “Phryniscide ;” these latter appear to be waifs and strays from old and extensive tribes that have gone down in the world and are becoming gradually extinct; the “ Hylide ” (like the smallest “Carinate” birds), being Frogs of high degree, are in no such danger of extinction. B 2 4 MR. W. K. PARKER ON THE STRUCTURE AND The range of size is greatest in the genus Rana, the highest or main typical form ; here we see that a Frog, as a Frog, should not be too large, nor too small; the largest have in them, as the American Bull-frog, much of an old generalised nature ; whilst the smallest, as 2. cyanophlictis and R. pygmea of the Oriental region, and some very small kinds of true Frogs in both the Neotropical and Ethiopian regions, are mani- festly kinds that have been arrested at a stage similar to that seen in similar-sized young of the typical or medium species. BIBLIOGRAPHY. I here append a list of papers and works that have been of service to me in the present work ; there is here no pretence of completeness, and several of the works treat of other types of Ichthyopsida :— Acassiz, Professor A., and Garman, 8. W., Esq. ‘On Fishes and Reptiles of the Lake Titacaca.’ Cambridge, Mass., U.S.; Nov. 26, 1875. Anpverson, Joun, M.D., F.L.S., F.Z.8. “On some Indian Reptiles.” Proc. Zool. Soc., Feb. 21, 1871, pp. 149-211. Barrour, F. M., M.A., F.R.S. ‘On the Development of Elasmobranch Fishes.’ London: Macmrnnan and Co. 1878. Bett, THomas, F.R.S. ‘A History of British Reptiles.’ London: JoHn Van Voorst. 1839. Born, Dr. Gustar. ‘Ueber die Nasenholen und der Thranen Nasengang der Amphibien.’ Leipzig : 1877. Brince, Professor T. W., B.A. 1. “On the Cranial Osteology of Amia calva.” Jour. of Anat. and Phys., vol. xi., pp. 605-622; pl. 23. 2. “On the Osteology of Polyodon folium.” Phil. Trans., Part II., 1878, pp. 683-733 ; Plates 55-57. Corr, Professor E. 1. “On the Classification of the Anura.” Jour. Nat. Se. Phil., U.S., vol. vi. 2. “Geology of Montana.” ‘American Naturalist,’ July, 1879, pp. 432-441. 3. ‘ Erpetology of Tropical America.” Proc. Amer. Phil. Soc., 1879, pp. 261-277. Cuvier. ‘Ossemens Fossiles.’ Atlas, plate 252. Paris: 1836. Ducks. ‘Recherches sur lOstéologie et la Myologie des Batraciens & leurs différens Ages.’ Paris: 1834. GarMAN, 8S. W., Esq. (Cambridge, Mass., U.S.). 1. “ Peruvian Fishes and Reptiles.” Bulletin of the Museum, 1875. 2. “Fishes and Reptiles of the West Coast of South America.” Proc. Bost. Soe. Nat. Hist., vol. xviil., pp. 202-205 ; Dec., 1875. 3. “Reptiles and Batrachians.” Ibid., vol. xviii., pp. 402-413 ; June 21, 1876. 4, “On Pseudis.” A note on the so-called Batrachichthys, described in the Ann, Mus. Nae. Rio de Janerio, vol. i., plate 6, by Dr. Pizarro, DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 5) GecEenBAur, Dr. Cart. ‘Untersuchungen zur vergleichenden der Wirbelthiere’ (Selachians). Leipzig: 1872. Gray, Dr. Joun Epwarp, F.R.S. “On Dactylethra.” Proc. Zool. Soc., 1864; p. 458. Gtuntu_er, Dr. A., F.R.S. 1. “ Batrachia Salientia.” Brit. Mus. Cat., 1858. 2. “ Reptiles of British India.” Ray Soc., 1864. 3. A series of papers descriptive of the Batrachia in the Proc. Zool. Soc., as follows :—1862, p. 190; 1863, p. 249 (and Ann. and Mag. of N. H., 1863, p. 26); 1864, pp. 46 and 303; 1868, p. 478; 1869, p. 500; 1870, p. 401; S72, p.5865 1873, p. 165; 1875,'p. 567. 4, “Description of Ceratodus.” Phil. Trans., Part Il., 1871, pp. 511-571; Plates 30-42. Huprecut, Dr. A. A. W. ‘Kopfskelet der Holocephalen’ (Callorhynchus and Chimera). Leyden and Leipzig: 1877. Hiuamnny. 0; B., GED: ERS. . “The Croonian Lecture.” Proc. Roy. Soc., Nov. 18, 1858. . ‘Elements of Comparative Anatomy.’ London: 1864. . ‘Anatomy of the Vertebrated Animals.’ London: 1871. . “On Menobranchus lateralis.” Proc. Zool. Soc., March 17, 1874, pp. 186-204 ; plates 29-32. 5. Article “ Amphibia.” Encycl. Brit., 9th edit., pp. 750-771. 6. “On Petromyzon.” Jour.of Anat. and Phys., vol. x., pp. 412-419 ; plates 17, 18. 7. “On Ceratodus Forster.” Proc. Zool. Soc., Jan. 14, 1876, pp. 23-59. Mra, Professor L. C. 1. “On the Labyrinthodonts of the Coal Measures.” Rep. Brit. Assoc., 1873, pp. 225-249 ; plates 1-3. 2. Ibid., 1874, pp. 149-192 ; plates 4—7. Mivart, Professor St. GEorGs, F.R.S. 1. “On the Classification of the Anourous Batrachians.” Proc. Zool. Soc., 1869, pp. 280-295. 2. “On the Common Frog.” ‘Nature’ Series. MacmILian and Co. 1874. 3. “On Pachybatrachus robustus.” Proc. Zool. Soc., 1868, p. 557. Mittter, Jon. “Vergleichende Anatomie der Myxinoiden, der Cyclostomen, mit durchbohrtem Gaumen.” Abh. Berl. Akad., 1835; Wiegm. Arch., 1836, i1., p. 245. Parker, W. K., F.R.S. 1. “On the Structure and Development of the Skull of the Common Frog.” Phil. Trans., 1871, pp. 187-211 ; Plates 3-10. 2. “On the Skull of the Eel.” ‘ Nature,’ June 22, 1871, pp. 146-148. 3. “On the Skull of the Salmon.” Phil, Trans., 1873, pp. 95-145 ; Plates 1-8, _ Seok m CO bo MR. W. K. PARKER ON THE STRUCTURE AND 4, “On the Sturgeon’s Skull.” M. Micr. Jour., Juhe 1, 1873, pp. 254-257 ; plate 20. 5, “On the Skull of the Batrachia.” Phil. Trans., 1876, Part II., pp. 602-669 ; Plates 54-62. 6. “On the Skull of the Urodeles.” Phil. Trans., 1877, Part I., pp. 529-597 ; Plates 21-29. 7. “On the Skull of Sharks and Skates.” Trans. Zool. Soc., vol. x., part iv., 1878, pp. 189-234; plates 34-42. 8. “On the Skull of the Urodeles.” Trans. Linn. Soc., ser. 2, Zool., vol. ii, 1879, pp. 165-212 ; plates 14-21. Swinnor, Rosert, H.M. Consul, Formosa. “On Reptiles and Batrachians from China.” Proc. Zool. Soc., 1870, pp. 409-412. Traquair, Ramsay, M.D. 1. “On the Osteology of Polypterus.” Jour. of Anat. and Phys., vol. v., pp. 166-182 ; plate 6. 2. ‘On Ganoid Fishes’ (Palzeoniscidze). Palzeont. Soc., 1877, pp. 1-60 ; plates 1-7. WALLACE, ALFRED Russet, Esq., F.L.8S. ‘The Geographical Distribution of Animals,’ London: MacmiLuan and Co., 1876. Pp. 414-423, WIEDERSHEIM, Dr. R. 1. ‘On Salamandrina perspicillata and Geotriton fuscus.’ Genoa: 1875. 2. ‘Das Kopfskelet der Urodelen.’ Leipzig : 1877. 3. “On Labyrinthodon Riitimeyert.” Abh. der Schwerz Palzont. Gesell., Zurich, 1878. 4. “Zur Anatomie des Amblystoma Weismann. Zeitschr. f. Wiss. Zool., bd. xxxii., pp. 216-236 ; plates 11, 12. 5. ‘Die Anatomie der Gymnophionen.’ Jena: 1879. Wiper, Professor Burt. G. ‘The Branchize of the Embryo Pipa.” ‘The American Naturalist,’ July (or August), 1877. Zoological list of the species whose skulls are described or referred to* in the present paper. In drawing up this list I shall make free use of the labours of my friends, Dr. A. Ginruer, F.R.S., and Professor St. GEorcE Mivart, F.R.S. The “ Batrachia * The species, the descriptions and illustrations of whose skulls are referred to here, and only partially described, are :-— Ist, Rana temporaria (see Phil. Trans., 1871, pp. 137-211, Plates 3-10; and Phil. Trans., 1876, pp. 603-605, Plate 54, figs. 1, 2). 2nd, Rana esculenta (see Huxtnry, article “‘ Amphibia,” Encyc. Brit., 9th Edit., vol. iii., pp. 750-771). 3rd, Bufo vulgaris (see Phil. Trans., 1876, pp. 605-625, Plates 54, 55). ; 4th, Dactylethra capensis vel levis (see Phil. Trans., 1864, pp. 625-648, Plates 56-59). 5th, Pipa Americana vel monstrosa (see Phil. Trans., 1864, pp. 648-665, Plates 60-62). DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 7 Salientia ” and various papers in the ‘ Zoological Proceedings’ give me the views of the former, and the article on the “ Classification of the Batrachia” (Proc. Zool. Soc., 1869, pp. 280-295) those of the latter, author. I shall use as few of the zoological characters as possible, and refer my reader to the works just mentioned for further information. IL.—ANURA PHANEROGLOSSA.* (Inctupine “ OpistHoGLossa”’ anp ‘ PROTEROGLOSSA.”) Il.—ANURA AGLOSSA. I.—PHANEROGLOSSA. A. With teeth in the maxillaries, premaxillaries, and generally in the vomers. a. With sharp fingers and toes. b. With dilated digital disks. B. With no teeth in either upper jaws or vomers. a. With sharp fingers and toes. b. With dilated digital disks. I. A. a. 1.—Frogs with sharp toes, teeth i jaws and vomers, toes more or less webbed, a bony shaft to manubrium (“omosternum”); cylindrical processes to sacral vertebra, and without parotoid glands. First Family. “ Ranip2.” Genera—Rana, Tomopterna, Pyxicephalus. First genus. Rana. . &. clamata, Daup.—Larve, three stages (A, B,C). Cambridge, Mass., U.S. . RR. pipiens, HAru.—Three stages of larvee (A, B, C) and Adult. N. America. R, ? sp.—-Larva. India. . R. palustris, LEcontE.—Newly metamorphosed young. Cambridge, Mass., U.S. . R. halecina, KAtmM.—Immature male. N. America. . R. temporaria, Linn.—Great Britain. Various stages described (see Phil. Trans., 1871, Plates 3-10, pp. 137-211.) . R. esculenta, Liyn.—Adult. Europe (see Huxuiey, Art. ‘ Amphibia,” Encye. Brit., 9th edit., vol. ii., pp. 750-771.) aur whe ~I * A few types enjoy some freedom of the tongue in front, namely, Rhinophrynus and Xenorhina; I have also found this character developed, in some degree, in Rhinoderma Darwinii, but Dr. Ginruer tells me that he does not think it a modification of any great importance ; yet the rule is for the tongue to be only free behind. Another character, namely, the possession of mandibular teeth, as in Hemiphractus and Grypiscus (Mivart, pp. 294, 295) is more important, but unfortunately I have not at present been able to procure either of these kinds. 8 MR. W. K. PARKER ON THE STRUCTURE AND 8. &. gracilis, Wrez.—Adult male. Ceylon. 9. R. cyanophlyctis, ScuNEtD. (see GUNTHER, Rept. of Brit. Ind., p. 406).— Ceylon. 10. Rana pygmea, Grur. (see P. Z. S., 1875, pp. 567, 568).—Adult male, ? inch long. Anamallays Mountains, Malabar, 8.W. India. 11. R. tigrina, Daup.—Adult female. Ceylon. 12. R. hexadactyla, Luss.—Adult female. Ceylon. 13. R. Kuhli, Scutec.—Male, two-thirds grown. Ceylon. 14. R. 4 sp—Adult female. A small species, 1 inch long. Lagos, W. Africa. Second genus. Tomopterna. 15. Tomopterna breviceps, SCHNEID., sp. (Pyaxicephalus breviceps in GUNTHER’s Rept. of Brit. Ind., p. 411).—(A) Half grown female. §. India. (B) Adult female. Ceylon. Third genus. Pyxicephalus. 16. Pyaxicephalus rufescens, JERDON.—Adult male. India. I. A. a. 2.—Like the last in most things, but have their toes free, or but little webbed, and the “ omosternum” not ossified, or even absent. Second Family. ‘“CysticNaTHIDé.” First genus. Pseudis. 17. Pseudis paradoxa, LiNN., sp.—Larvee in four stages, and adult. §. America. Second genus. Gomphobates. 18. Gomphobates (Leiuperus) % sp. (see P. Z. S., 1868, p. 478, same as Leiuperus, “ Batr. Sal.,” pp. 22 and 135).—Adult (?), 10 lines long. River Plate. Third genus. Cystignathus. 19. Cystignathus ocellatus, var. LINN., sp.—Adult male, 55 inches long, Dominica; with larve 2$ and 34 inches long. Brazils. PAC Oe ? sp.—Tadpole, 1 inch long ; tail, } inch; hind legs, 1 line. Lake Jannarg, Manaoo, Brazils. 21. C. typhonius, Daup., sp. (C. typhonius, Dum. and Bre.).—-Adult female, 14 inch long. Porto Rico. Fourth genus. Pleurodema. 22. Pleurodema Bibronii, Tscuupi.—Adult female, 14 inch long. Chili. Fifth genus. Lymnodynastes. 23. Lymnodynastes tasmaniensis, Grur.—Adult female, 13 inch long. Tas- mania. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 9 Sixth genus. Camariolius. 24. Camariolius tasmaniensis (?) (see GUNTHER on “Batrachians from Australia,” P. Z. §., 1864, pp. 46-49, plate 7, fig. 3; there called Pterophrynus tasmamensis. Dr. GUNTHER is nearly satisfied that my specimens belong to that species)—(A) Adult female, ? inch long; and (B) Larva, 2 inch long ; legs, yg inch. Australia. Seventh genus. Cyclorhamphus. 25. Cyclorhamphus marmoratus, Dum. and Brs.—Adult female, 12 inch long. Vinco Caya, Peruvian Andes. Height 16,000 feet. 26. Cyclorhamphus culeus, GARMAN.—Larva, 31 inches long; tail, 2 inches; hind legs, 7 lines. Puno, Lake Titacaca, Peru. The next group differs from the Ranidz in having the sacral apophyses dilated. I. A. a. 3.—Third Family. “ Drscoctossip &.” First genus. Discoglossus. 27. Discoglossus pictus, OrrH.—Adult male, 2§ inches long. §. Europe. Second genus. Pelodytes. 28. Pelodytes punctatus, Dauv.,sp., Frrz., gn.—Adult male, 14 inch long. Europe. Third genus. Xenophrys. 29. Xenophrys monticola, Grur. (see Rept. Brit. Ind., plate 26, figs. H, H’, p- 414).—Adult male, 3 inches long. Darjeeling. Fourth genus. Calyptocephalus. 30. Calyptocephalus Gay, Dum. and Brs.—(A) Adult female, 54 inches long, and (B) Tadpole, 4% inches long ; tail, 2? inches; hind legs, } inch. Chili. I. A. a. 4.—Fourth Family. “ Atyrmp2.” In this family the Ranian type is modified by some new characters. The sacral “e ? apophyses are dilated, and there are “parotoids” as in the Toad; ribs present, and vertebrae opisthocalian. Genus Alytes. 31. Alytes obstetricans, LAuR., sp.—Adult female, 1 inch 10 lines long. Europe. I. A. a. 5.—Fifth Family. “ Hyperotnp.” This type differs from the last in many things, notably in possessing, besides the neck-glands and wide sacrum, which is “not much extended” (GUNTHER, “ Batr. Sal.,” p- 39), proccelian vertebrae, and no ribs (Mivart, p. 291). Genus Hyperolius (formerly Uperoleia; the species of so-called Hyperolius described MDCCCLXXXI. © 10 MR. W. K. PARKER ON THE STRUCTURE AND in the “ Batrachia Salientia” are now put into the genus Rappia) ; they are narrow-backed Tree-frogs. 32. Hyperolius marmoratus, GRAY, sp.— Adult female, 1f inch long. Paramatta, Australia. I. A. a. 6.—Sixth Family. ‘‘ BomBrnaTorIp&.” This type has dilated sacral apophyses but no parotoids, it has ribs, and abnormal (opisthocelian) vertebree, but no columella, or tympanic cavity. Genus Bombinator. 33. Bombinator igneus, ROsEL, sp—Adult female, 13 inch long. Europe. I. A. a. 7.—Seventh Family. “ PELopatip&.” The last has one of the softest, this has one of the hardest, skulls. There is no cavum tympani, but there is a small bony columella. There are no ribs, and the vertebree are proccelian ; it has a broad sacrum, but no parotoids. Genus Pelobates. 34. Pelobatus fuscus, LAuR., sp-—Adult male, 2 inches 5 lines long. Europe. I. A. b.— With dilated digital disks. First sub-division of Tree-frogs. First Family. “ PoLypEpDATID&.” Ear perfect. Sacral apophyses cylindrical. ‘Omosternum” with a bony shaft (Mivart, p. 292). “No parieto-frontal fontanelle” (except in such small types as the Australian Rappia bicolor). No neck-glands. First genus. Polypedates. 35. Polypedates chloronotus, Grur. (see P. Z. S., pp. 567-576, plate 65, fig. A). Adult male, 2 inches long (GiinrHER gives the length of the body in the male as 51 millims., and of the female 93 millims; p. 570). India. 36. Polypedates maculatus, GRAY, sp.—Adult male, 2 inches 1 line long. India. Second genus. Lhacophorus. 37. Rhacophorus maximus, Grur.—Adult male, 3§ inches long. N. India. Third genus. xalus. 38. Ivalus variabilis, Grar.—Adult female, 1 inch 1 line long. Ceylon. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 11 Fourth genus. Hylarana. 39. Hylarana malabarica, Dum. and Bis. sp.—Young, § inch long. India. 40. Hylarana temporalis, Grur.—Adult male, 25 inches long. Ceylon. Fifth genus. Rappia. 41. Rappia ? sp.—Adult female, § inch long. Lagos. (Probably 2. lago- ensis, GTHR., see P. Z. S., 1868, pp. 478-490, plate 40, fig. 2.) 42. Rappia (Hyperolius) bicolor, GRAY, sp.—Adult female, ? inch long. Dog-trap Road, Paramatta, Australia. (See ‘ Batr. Sal.,” p. 89; Dr. GUnrHeER there remarks that, “this species is very probably the type of a separate genus, but the condition of this single specimen does not enable me to give the characters with certainty.” The fact of the case is, that the skull of this specimen is extremely membranous, as much so as in that of Acris Pickeringii (see Plate 19, figs. 6-10, and Plate 30, figs. 1-5). The Rappia from Lagos had, on the contrary, a highly ossified skull, with a well-covered roof (see Plate 28, figs. 6-10). Second Family. “ HyLoprp&.” (I make up this family of Professor Mivart’s sub-families AcriprInA and HyLop1Na.) No bone in the manubrium (“omosternum”). Vertebrze proccelian ; fontanelle very open (Acris), or covered (Lylodes). First genus. Hylodes, 43. Hylodes martinicensis, Brs., sp.—Adult female, 14 inch Jong. Martinique. Second genus. Acris. 44, Acris Pickering, Hors.—(A) Adult female, 10 lines long (% inch); and (B) larva, 1 inch 2 lines long ; tail, # inch ; hind legs, } inch. Cambridge, Mass., U.S. Second sub-division of Tree-frogs. Family “ Hyp.” These have the sacral apophyses dilated; and have no parotoids. The ear is perfect, and, as a rule, the fontanelle is open; I find it covered in Hylu rubra, and almost covered (by sculptured solid bones) in Nototrema. First genus. Hyla. 45. Hyla Ewingii. Dum. and Brs.—Adult female, 14 inch long. Van Diemen’s Land. 46. H. phyllochroa, Grup. (see P. Z. S., 1868, p. 481).—Adult female, 1 inch 5 lines long. Cape York, Australia. 12 MR. W. K. PARKER ON THE STRUCTURE AND 47. Hyla arborea, LInN., sp.—Adult male, 13 inch long. 8. Europe. 48. H. albomarginata, Sprx.—Adult female, 23 inches long. Brazils. 49. H. rubra, Daup.—Adult male, 1 inch 11 lines long. §. America. 50. H. ? sp.—Tadpole, 1 inch long; hind legs, 5 lines. Rio Janeiro. Second genus. Litoria. 51. Lntoria marmorata, Dumérit.— Adult male, 14 inch long. Australia. In this type the digital disks are very small and the head is very large in proportion to the body. I suspect that it comes very close to an “ Oxydac- tyle” Frog from the same region, viz.: Myxophyes fasciolatus, Grur. (see P. Z.S., 1864, p. 46, plate 7, fig. 1). Third genus. Nototrema. 52. Nototrema marsupiatum, Dum. and Brs., sp.—(A) Adult male, 13 inch long. S. America; and (B) Tadpole, 23 inches long; body, 1 inch; tail, 14 inch ; hind legs 13 lines; right fore leg free, } inch long; left fore leg still under the operculum. Third sub-division of Tree-frogs. Family ‘‘ PeLopryapip™.” Parotoids present ; other characters Hyline. First genus. Pelodryas. 53. Pelodryas ceruleus,* Wuuire, sp. (see “ Batr. Sal.,” p. 199, and plate 9, fig. B.)—Adult male, 3 inches long. N.S. Wales. Second genus. Phyllomedusa. 54. Phyllomedusa bicolor, Bopparrt.—Adult female, 3} inches long. Santarem, River Amazon, lat. 2° 20’ S., 8. America. I. B. a.—Toothless “ Anura.” Without digital disks. Oxydactyle Toads. First Family. ‘“‘ Buronip&.” _ Typical Toads, with parotoids, and processes of sacral vertebree dilated. Ears perfect. Toes webbed. Skull generally strongly roofed; an open fontanelle in Bufo calamita. First genus. Bufo. 55. Bufo pantherinus, Bors.—Adult female, 44 inches long. Africa. 56. B. melanostictus, ScHNEID.—Half-grown male, 24 inches long. India. 57. B. agua, Larr.—Old female, 64 inches long; and younger female, 5 inches long. §S. America. * Evidently named from a bleached spirit specimen ; the living Frog is bright green. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 13 58. B. chilensis, Tscuup1.—(A) Adult male, 3 inches long; and (B) Tadpole @ inch long; tail, 3 inch; hind legs, 35th of an inch. Arequipa, Peru. 59. Bufo lentiginosus, SHsw.—(A) Tadpole, § inch long; hind legs, 45 inch long; and (B) newly metamorphosed young male, 3 inch long. Penekese Island Mass., U.S. 60. B. vulgaris, LauR.—Adult and Tadpoles many stages. England. 61. B. calamita, Laur.-—Adult female, 2} inches long. England. 62. B. ornatus, Spix.—Adult female, 2$ inches long. §S. America. Second genus. Otilophus. 63. Otilophus margaritifer, Larr—(A) Half-grown female, 14 inch long; Vene- zuela; and (B) adult female, 23 inches long (Hyrri’s prepn. Mus. Coll. Surg., Eng.), Brazils, Second Family. “ RatmvopERMATID.” This group agrees with the “ Bufonidze” in many things, but there are no parotoids, and the ear is less perfectly developed. First genus. Rhinoderma. 64. Rhinoderma Darwin, Dum. and Brs.—Adult male, 1 inch long. Chili. Second genus. Diplopelma. 65. Diplopelma ornatum vel rubrum, Dum. and Bis., sp.—Adult male, 11 lines long. India. 66. Diplopelma Berdmorei (?), Buyra (GUNTHER, Rep. of Brit. Ind., p. 417).— Adult female, 1 inch 1 line long. Moulmin, Tenasserim. Third Family. “ BracHycEPHALID&.” These Toads have “no tympanum nor cavum tympani” ;—‘‘ free toes, with the pro- cesses of sacral vertebrze dilated, and without parotoids.” (“ Batr. Sal.,” p. 45.) Genus Pseudophryne. 67. Pseudophryne Bibronu, Grur.—Adult female, 1 inch long; adult male, 3 inch long. N. 8. Wales. Fourth Family. “ Paryniscip®.” Ear imperfect (sometimes with a collumella, as in Phryniscus cruciger); webbed toes; dilated sacral apophyses ; and no parotoids. Genus Phryniscus. 68. Phryniscus cruciger, Martixs and Ginruer.—Adult male, Lj inch long, Interior of Brazils. 69. P. varius, STAN., sp.—Adult female, 1} inch long. Costa Rica, 70. P. levis, Grar.—Adult female, 13 inch long. Ecuador. 14 MR. W. K. PARKER ON THE STRUCTURE AND Fifth Family. ‘“ ENcystomip&.” Ear rather imperfect ; dilated sacral apophyses; without parotoids ; and with free toes. Genus Hngystoma. 71. Engystoma carolinense, HoLpr.—Adult male, 11 lines long. Florida. I. B. b. Loads with digital disks. First Family. “ HyLapactyiip&.” Ear perfect ; webbed toes ; dilated sacral apophyses ; and no parotoids. Genus Callula. 72. Callula pulchra, Gray.—Adult female, 2§ inches long. Pegu. Second Family. “ HyLapLesip&.” Ear perfect; toes free; processes of sacral vertebree cylindrical; and without parotoids. Genus Hylaplesia. 73. Hylaplesia tinctoria, ScHNEID., sp.—Adult female, 14 inch long. South America. Il._—ANURA AGLOSSA. Without tongue; ear perfectly developed; Eustachian tubes united ; cavum tym- pani entirely bony. a. With maxillary teeth, Family “ DAactyLeTHRip&.” With webbed toes ; dilated processes of sacral vertebrve ; and without parotoids. Genus Dactylethra, 74. Dactylethra levis, Daup., sp. (D. capensis, Cuv., see Phil. Trans., 1876, pp- 625-648, Plates 56-59.)—Adult and four larval stages. 8S. and W. Africa. b. Without maaillary teeth. Family “ Pirrp&.” With webbed toes; and with the processes of the sacral vertebrae dilated; no parotoids. Genus Pipa. 75. Pipa Americana, Sepa, (see Phil. Trans., 1867, pp. 648-665, Plates 60-62.) —Adult, ripe young, and larvie 9 lines long, DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 15 ON THE TYPICAL BATRACHIAN SKULL. In the following descriptions I shall take the skull of Rana temporaria as the “norma” or pattern form; it is the best on the whole ; one or two exceptional cha- racters exist in it, viz.: the mark of the originally separate metapterygoid (a rare character) ; the annulus tympanus is not perfect until old age; and the stylo-hyal is a long while fusing with the floor of the skull and ear-capsule, if indeed it ever becomes fused ; but, with these exceptions, this may be taken as the highest kind of Batrachian skull, and the best rule to measure the others by. ON THE EARLY STAGES OF THE BATRACHIAN SKULL. No known kind of Vertebrate shows so many and such instructive stages in its development as the Batrachian. The skull of the newly-hatched embryo of the common Frog or Toad is strictly comparable to that of a larval Lamprey (Ammocetes), whilst that of the well-grown Tadpole comes very close to the skull of the adult Petromyzon. But whilst the ordinary Batrachia have a very Petromyzine or suctorial larva, the “‘ Aglossa” in their early stages have a cranium and face very similar to what is seen in the “Siluroid” Teleostean Fishes, and their skull suggests to the observer the most probable form of endocranium likely to have existed in such Ganoids as Plerichthys and Coccosteus. Both these kinds of larval endocranium (chondrocranium) are figured in my “ Batrachian Skull,” Part II. (Phil. Trans., 1876, Plates 54 and 55, Bufo vulgaris; Plates 56-58, Dactylethra; and Plate 60, fig. 3, Pipa). My present business is with the suctorial type of larval Batrachian skull; for full details and figures of the wide-mouthed Siluriform type of skull I must refer the reader to the paper just referred to. My most successful dissection of the earliest cartilaginous skull in these types was that of Bufo vulgaris (op. cit., Plate 55, figs. 1, 2); the embryo was 4 lines (}rd of an inch) in total length; the next to this (Plate 55, fig. 3) was of a Tadpole of the same species, 5 ines long. I shall make the first of these my First Stage, referring the reader to the plates in the published paper. Of the skull in more advanced stages the present work will give many instances and illustrations ; after describing the simple foundation, as seen in Bufo, I shall describe a series of stages in various species of the genus Rana. The larval skulls of other types will be described in their proper order, with the adult condition of the skull in the same and other species. 16 MR. W. K. PARKER ON THE STRUCTURE AND I must here remark that the early cranium of the Aglossal types, especially that of Dactylethra, is in some respects more Petromyzine than that of an ordinary Tadpole. In some things the chondrocranium of the Tadpole of Bufo vulgaris comes nearest to that of the Lamprey, as in the fusion together of the various plates of cartilage ; but the skull of the larval Ductylethra comes very near to that of the adult Petromyzon in several respects. ; This is seen in the complete fusion of the elements of the branchial skeleton, and in the histological condition of those bars and pouches: these are composed, as in the extra-branchial basket-work of the Lamprey, of a very light kind of cartilage, with large cells and scarcely any intercellular substance: the large amount of superjicial cartilage shows also the same relationship. The Common Toad differs from most of the Opisthoglossa in retaining the primordial “pedicle” to the pier of the mandible, which is, from the first, continuous with the trabecular bar: this is a remarkable survival of a suctorial character—even after metamorphosis. SKULL OF LARVAL BATRACHIANS. First Stage.—Craniofacial cartilages of embryo of Bufo vulgaris soon after hatching ; 3rd of an inch in total length. Nearly all the truly cartilaginous part of this cranium (“Skull of Batrachia,” Plate 55, figs. 1, 2) lies at the base of the fore half of the head (t7.); two globes of soft cartilage unfinished in their upper third, are to be seen right and left in the hind half: these are the very distinct auditory capsules (av.). j At present the whole body of the embryo exists as a sort of tail-like appendage to the huge and precocious head: the development of cartilage appears, therefore, almost entirely in front of the axis of the organism (notochord), which stops at the post- pituitary space, before the “ pituitary body” is formed. Thus the cranium, as yet, is nearly all pro-chordal; and the para-chordal part, like the side-walls and roof of the cranium, is still membranous. Yet these prochordal tracts, or ‘“trabecule cranii” are manifestly true paraxial elements or parts; they are homologous with the paired cartilages that appear in any region along the sides of the sheath of the notochord. All growths above these (dorsad) are of the nature of neural arches; all growths below these (ventrad) are ‘“ visceral” or “ pleural” arches. But these rods are continuous, and they are not parallel the one with the other ; how is this to be explained on the theory of their paraxial nature ? A consideration of the development of this, or of any, Vertebrate embryo, will help us to understand the meaning of these first foundations of the cranium. The neural axis rapidly enlarges at its cephalic end, forms three vesicles there, the hind, mid, and fore brain, and this beaded structure is bent suddenly upon itself so DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 17 that the fore brain is turned downwards and somewhat backwards, and the mid brain is tilted upwards and forwards, and forms the actual end of the embryo, lying against what is called the “frontal wall.”* The fore and hind vesicles are thus brought into contiguity, and the organic apex, or real fore end of the embryo, is a little in front of the hind vesicle ; it looks down- wards and a little backwards. At that point a remarkable process takes place ; the notochord turns downwards, taking, as the skeletal axis, the same direction as the neural axis : the fore brain after a time sends a budding process owt, and the oral mucous membrane sends a budding process in, and the two, gradually becoming mutually engrafted, the one upon the other, open freely into each other, and then the lower process becoming closed below, we have the pituitary body formed. In this quasi-archaic condition of the head and its parts, the body is a mere appendage, solid hyaline cartilage forms first where most pressure is, and the paraxial bands grow first of all from near the apex of the notochord to near the frontal wall. But the fore brain swells and hangs down; the skeletal parts respond to this condition and wind their way round its base, embracing its sides, and then meet, or nearly so, in front, to diverge again in the nasal region of the face. After a few days the paired bands have developed backwards under the hind brain (ibid., figs. 3 and 4), and behind the head they appear in patches that alternate with the somatomes (muscle-segments, &c.) along the spinal region. Then, in Tadpoles jfive-twelfths of an inch long (op. cit., Plate 55, fig. 3), the chondrocranium is simply a pair of planks on which the brain lies; it has become much straighter, but the mid brain lies high, still. Strength, however, has been gained by fusion at two points (in the antorbital and postorbital regions), and between these two points the bars are becoming crested ; these ridges are the beginning of the ethmoidal and sphenoidal side walls. Returning to the earliest stage we find that there are, indeed, two pairs of cartilage on each side nearly equal, with a third pair of much shorter bars. In these temporary sucking fishes the chief parts of the organism are all crowded into the head and throat; all else is their paddling fin, and the creature is called a bu/l- head, on account of this cephalic preponderance. The second, more external, cartilaginous band (pd., g.), carrying at its end a short inturned segment (Mk.), is the “pier” of the mandible, with its free swinging joint, the rudiment of the lower jaw. Under that pier, one-fourth of the way from its distal end, the third cartilage is seen (c.hy.); this is the free joint of the hyoid (lingual) arch whose pier or epi-hyal element does not appear for three or four months to come, and when developed is devoted to new purposes, does not become the practical “suspensorium ” of the lingual arch at all, but forms part of the outworks of the auditory labyrinth, * See Batrour’s ‘ Elasmobranchs,’ plate 7. MDCCCLXXXI, D 18 MR. W. K. PARKER ON THE STRUCTURE AND Behind these parts, in relation to the hinder visceral clefts, four other arches are beginning to form. These, the branchials, are still more feebly developed, never acquire an upper element (“ epi-branchial”), but in their arrest and feebleness are supplemented by four pairs of extra-branchial cartilages or pouches ; these I shall describe in the developed embryos of the larger species of the genus Rana. Until then, also, I shall not describe the other superficial cartilages that are found in the Tadpole’s head, viz: that over the “ 1st cleft,’ and those so largely forming the suctorial mouth of the Tadpole. In this early stage the cranial nerves pass over the trabeculz through the membrano- cranium. Here, if anywhere, we ought to find rudiments of pre-oral arches, if there be any ; there are certain things in other kinds of embryos, and even adult forms, that suggest such a possibility, but none here.* At present the mandibular pier has two points of fusion with the trabecula ; the hinder of them is the pedicle (pd.), and the front one the palato-pterygoid bar (p.pg.) ; the pedicle will be largely absorbed, and the palato-pterygoid enormously developed during the metamorphosis of the Tadpole. Before passing to the description of the well-developed Tadpole, I must refer again to the intermediate stages (ibid., Plate 55, figs. 3, 4, 5). On its way to the coronoid region of the mandible, the temporal muscle passes over the palato-pterygoid bar, and under a leafy growth of the “suspensorium”—the “ orbitar process” (or.p.). In Bufo vulgaris the apex of this leaf coalesces with the ethmoidal cartilage, a rare character. ‘This fact, and its nonconformity with what is seen in the Tadpole of the Frog, were pointed out to me by Professor HuXueEy five or six years ago. In other species (even of Bufo) I do not find this very peculiar condition of things, but the process itself is well developed, even in the wide-mouthed larvee of the ‘“ Aglossa” (ibid., Plates 56-61). In the half-grown larvee figured in my paper (ibid., Plate 55, figs. 4, 5, 5a), we see the formation of the “fenestra ovalis” as an oblique lateral cleft in the auditory capsule ; and of the “stapes” as the solidification of the soft tissue left in this space. In larvee three-fourths grown ossification begins in the tissue over and under the chondrocranium. 7 * Professor MitNes MarsHALL and the writer, from a consideration of the ‘‘ segmental nerves” of the head, from the development of the mouth itself and of the lacrymal and nasal passages, and from what is seen in various cartilages around and in front of the mouth, hold to the opinion that there are rudiments of at least two, perchance three, pre-oral visceral arches. But Professor Huxtey and Mr. F. M. Batrour will not see with our eyes, and are not in the least satisfied with the evidence which seems so conclusive to us. The Lamprey and the Tadpole are indeed great stumbling-blocks; if they represent primordial or archaic forms, and if what we see in other kinds bear to be interpreted as specialisations having relation to modification of the oral aperture and its framework, the onus probandi will still rest with my talented young friend MarsHaLi and me. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 19 SKULLS OF LARVAL ‘‘RANID&®.” Chondrocranium of perfect larva of Bull Frogs. First Stage.-—1. Tadpole of Rana clamata (A), whose total length was 34 inches; of the tail, 2 inches ; and of the hind legs, 5 lines (;°y winch). The skull (Plate 2, figs. 5, 6) has now attained its full cartilaginous (Petromyzine) condition, and has gone beyond it; there is a pair of cartilage-bones, and there are three membrane bones. I shall deal with these signs of metamorphic progress after describing the chondrocranium. The cranium, proper, is now a small flat-bottomed boat, with a semi-cartilaginous deck; this has been constructed upon the original frame in the following manner. The large open space in the orbital region has acquired a thin floor of cartilage, and this median (“intertrabecular”) tract has grown forwards to the region of the inner nostrils (.n.), where the trabeculze have grown together, and are diverging again as free, flat, broad-ended “ cornua ” (c.tr.). Behind, the occipital arch is as perfect as may be, for the auditory capsules (vb.) fit into large fenestre in its sides; its top (s.0.) is flat, and extends to the fore- most third of the capsules ; its floor is flat, and contains the diminishing notochord (nc.) in its middle: in this kind this rod scarcely acquires any true cartilage in its outer sheath. In the large orbital region the side walls are cartilaginous, except where the optic nerves (II.) pass out of their “fenestrae ”—spaces three or four times as large as the actual foramina. As in the occipital region, so also in this, there is a cartilaginous “tegmen ;” its halves have not united in the middle, and a bilobate “ fontanelle ” (fo’.) exists between these growths and the superoccipital cartilage. Then there is a large oval fontanelle ( fo.) or membranous space up to a short distance of the end of the cranial cavity, where a tract of cartilaginous roof margins this egg-shaped space. Here there is a perfect cartilaginous cincture, which becomes in the adult the en ceinture ;” its front end is extended as a cartilaginous beak formed from below upwards as a crested tract of the “intertrabecula ;” this is the mesethmoid (p.e.) in rudiment ; the lateral regions are the right and left “ ecto-ethmoids” (al.e.). “e Os These lateral parts grow outward into wings whose scooped front margins form a back wall to the nasal sacs ; the olfactory nerves pass to these sacs through “ fenestre that lie between the middle and outer parts of the cincture. The nasal roofs are still soft, and in the dissection the olfactory sacs were removed to display the cartilages. But the internal nostrils (i.n.) are shown in these figures; they are large and neatly circular ; much of their margin is seen to be formed by the narrow, clinging part of the trabecule, and the curved spatula that grow out from the ethmoidal D2 20 MR. W. K. PARKER ON THE STRUCTURE AND cincture. On the outside the quadrate cartilage (q.) sends a process inwards which is tied by a ligament to the corresponding trabecular horn (c.tr.), and thus the circle is completed. From this part the cornua turn downwards in a gentle manner, as they diverge, and end in a broad, sub-emarginate flap. The sclerotics and nasal roofs, which were beginning to chondrify, having been removed, we have only the auditory sacs to describe in this stage. They are now thoroughly cartilaginous, and quite confluent with the cranial walls, roof, and floor. Moreover, they are swollen with the growth within of the three wide canals, and the sacculated base of the membranous labyrinth (q.s.¢., h.s.c., p.s.¢., vb-). A crescentic flap of cartilage grows outwards from the horizontal canal; this is the “teomen tympani,” under that roof we see the large fenestra ovalis, and the well- fitting oval stapes (vb., st.). The fore angle of the tegmen has a flap of cartilage growing from it, the result of coalescence ; this, the rudimentary “annulus,” will be described soon. Leaving out some of the smaller passages, we may refer to the double passage for the 9th and 10th nerves (fig. 6, [X., X.), the single passage for the 7th and 5th (VIL, V.) and the fenestra for the 2nd nerve (IL): the 8th and 1st are out of sight in this view. Each of the huge “mandibular piers” is two-thirds the length and two-thirds the breadth of the cranium proper. Besides the two earliest conjugations with the basal cartilage, each bar articulates by its ‘otic process” with the cartilage that projects from the tegmen tympani, and is tied, as above stated, near its distal end to the cornu trabeculee. It is thus swung at four points from the basis-cranii, but of these, the “ pedicle” is the true apex or dorsal end (fig. 6, pd.), it lies beneath the parting of the facial and trigeminal nerves (VII., V.), and runs parallel for some distance with the front face of the auditory capsule. Between the pedicle and the palato-pterygoid bar there lies a large subocular fenestra (s.o.f:), or membranous space, parallel with the skull ; and behind the pedicle a smaller oblique space is formed, bounded behind by the ear-sac. The large “suspensorium” is thick at its outer edge, and thin at its inner, which rises somewhat from its smooth, sinuous, upper face; on this the long temporal muscle rests, as it passes from its origin in the post-orbital region, to its msertion on the * coronoid” crest of MECKEL’s cartilage (m/h.). The suspensorium is terminated by a reniform condyle for the mandible (q., mk.), and opposite the ethmoid it has on its under surface, at the outer edge, a pyriform flattish facet for the hyoid plate (hy,f). Over this secondary condyle, which projects outwards, the suspensorium is developed into a large rounded leaf, with a broad adherent base, a ribbed edge, and a hollow upper surface—this is the “orbitar process” (o7.p.), which bends over the temporal muscle and is tied by its apex to the skull wall in front, DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 21 Its fore margin sends out a snag; another follows this near the condyle for the lower jaw, and opposite this is a third, which has been described as tied by a ligament to the cornu trabecule. The apex of the orbitar process rests upon a longitudinal crest of the palato- pterygoid band; this is the rudiment of the post-palatine (pt.pa.) ; it projects, now, into the subocular space. The whole of the structure just described is not merely an hypertrophied epi-visceral element, it has given rise to a large extra-visceral outgrowth. The “ cerato-visceral ” element is one-third the length and one-third the breadth of its own upper piece—but it is thicker ; this is the free mandible or Mrcken’s cartilage (mk:). This bar is strongly curved, having a long, terete, angular process behind its condylar face, which is a deep notch, like that of the human ulna. The shaft of the cartilage is both dilated and thick, and its subconcave end articulates with one side of the large sucking disk, the corresponding inferior labial (/./.). This huge, almost horizontal mandibular arch, with its free bars turned inwards at little more than a right angle, is flush, in front, with the trabecular end of the chondrocranium. The four points of attachment to the basis-cranii of this long suspensorium shows a modified condition of that complete fusion of the two regions seen in the Chimezeroids. In Dactylethra (ibid., Plate 56) this state of things is intensified, and in Pipa (ibid., Plate 60, fig. 3) the membranous space between the basal and lateral cartilages is reduced to a small crescentic slit. The larval Ranine skull may therefore be said to be, in this respect, a specialisation and dissection, so to speak, of the Chimeeroid type of skull; on the whole, it is much more like that of the Lamprey. This immense development of the first epi-visceral element is correlated with the suppression of all those that should succeed ; only afterwards, when this part has been relatively lessened, and greatly modified, does even that of the next succeeding arch appear, not then as a mere facial bar, but specialised to auditory purposes, as the “ columella.” But the “cerato-hyal” (Plate 2, fig. 7, c.hy.) is at present very large, several times larger than its counterpart of the mandibular arch (mk.). It has a dilated, fan-shaped distal expansion (fig. 7), a gently convex pyriform condyle (hy.c.), and a curved upper spike, the rudiment of the stylo-hyal (st.h.) or upper end of the bar. The basi-hyal (b.hy.) is still membranous, but it is followed by a solid basi-branchial (b.br.) ; these parts will be described in a corresponding larva of R. pipiens, and in a riper larva of this same species. The ossifications in this stage are, first, a roundish patch of bone on each side, formed in the substance of the cartilage, close inside the aperture for the 9th and 10th nerves ; these are the ex-occipitals (e.0., [X., X.). The next are formed in membrane, outside the cartilage; they are the fronto- 22 MR. W. K. PARKER ON THE STRUCTURE AND parietals (fp.), thin wedge-shaped shells of bone on the roof; and the parasphenoid (pa.s.),a large dagger-shaped bone lying under the floor ; its point is rounded and the median hind process is a low triangle ; the “basi-temporal wings” are broad and oblique at their outer end. These three are subcutaneous ossifications, but there is also a series of subcutaneous cartilages. The suctorial mouth is mainly formed by a coalesced upper pair and a distinct lower pair of these, here called “labials” (w.1., U./.). The upper piece shows where it was once in two distinct parts; it is a broad crescentic arched flap, with outer angles, an emarginate hind, and a round front, margin. Towards the sides there is a small fenestra, which is the beginning of a new sub- division ; the angles answer to the upper angular cartilages of the Lamprey’s mouth, and the main part to the “anterior dorsal cartilage” of that Fish. The “ lower labials” form together a horseshoe, and are fixed in between the mandibles; they have a thick lower edge and a concave inner face (fig. 2, /.1.). The wedge-shaped ray of cartilage, which is confluent with the “tegmen tympani,” and articulates with the otic angle of the suspensorium, answers to the mandibular ray (the “spiracular cartilage)” of the Shark: it becomes the “annulus.” The extra- branchials will be described in the next instance; part of the first (ex.br'.) is shown in fig. 7. Second Stage.—2. Tadpoles of Rana pipiens (A and B), from 33 inches to 5 inches long, with hind legs appearing. In the largest of these I shall show the details of the structure of the palate, mouth, and throat (Plate 1, figs. 83-5); and in the lesser specimens the skull and face (Plate 3, figs. 1-3). On the whole, the structure of the chondrocranium is very similar to that just described ; there are, however, some very instructive differences. The skull (Plate 3, figs. 1-3) is altogether rounder, less angular, and free from the projecting snags; the occipital arch is wider and flatter, the cranial cavity not so oblong, but is pinched in in the middle. The internal nostrils (7.7.) are oblique and subreniform, converging towards each other behind; thus the palato-pterygoid bar (p-pg.) is longer and more oblique, and its post-palatine rudiment ( pt.pa.) projects less into the larger and more oval subocular space. The cornua trabecule (c.t7.) have wider ends and with a more pronounced outer angle, the outer wings of the ethmoid (al.e.) are not so well developed, and the vertical septum (p.e.) is shorter. In the side walls (fig. 3) the cartilage is low, and the hinder roof-cartilage (“ tegmen cranii”’), with its two fontanelles, is seen to grow forwards from behind, and not to be continuous with the low side wall. In this stage, even, besides the centres of bone that grow inside the 9th and 10th nerves (¢.0., [X., X.), there is on each side a bony patch in front of the ear-capsules ; DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 23 they are crescentic and enclose the outside of each “foramen ovale;” these are the prootics (V. pv.o.). The fronto-parietals (f:p.) are less developed; each bone is a sharp style, growing broader behind, and shell-like. Also the parasphenoid, below (fig. 2, pa.s.), is altogether a smaller bone, with more slender processes; in front it stops short obliquely: here there will appear an additional and rare centre—the “‘ pro-parasphenoid” (see fig. 5, pa.s’.). The auditory capsules are relatively less than in R. clamata; the “tegmen tympani ” is a rounder, less pronounced growth, and the position of the stapes and its fenestra (st.) is more oblique and inferior in position. The narrow-waisted cranium, and the more out-turned suspensorium, make the subocular fenestra of a very regular oval shape instead of oblong. The orbitar processes (07.p.) are smaller, and the quadrate region in front of them shorter ; and the spur (pr.pa.) which by ligament is tied to the cornu trabecule is a mere blunt projection. The palatine ridge (“‘ post-palatine rudiment,” pt.pa.) projects backwards less, but is larger than the outer wing of the ethmoid (q/.e.). The arcuate form of these great suspensoria throws the quadrate condyles further (obliquely) inwards, and thus shorter free mandibles (mk.) are needed. Here the condyle for the hyoid (ly.f.) is only two-thirds the distance from the hinge for the lower jaw, as compared with the former instance. Except in Pseudis, to be afterwards described, there is no better kind of Tadpole than this of the American Bull-frog for showing the structure of the post-oral arches. The figure (Plate 2, fig. 8 ) was taken from a dissection of a larva 33 inches long ; it is drawn as it appeared from above, after the gill-tufts were cleared away, and the whole basket-work somewhat flattened out for display. The “cerato-hyals” (c.hy.) are roughly hourglass shaped; very solid in the middle, and twisted ; they flatten out into massive slabs at each end. The natural position of those bars is shown in Plate 3, fig. 3, articulating by a somewhat convex condyle with the flattish facet under the fore part of the suspen- sorium. , Outwards and backwards from the joint the “stylo-hyal” process (st.h.) forms a large irregular triangle ; below, the ‘‘ hypo-hyal” region is greatly outspread or pedate. Each lobe projects inwards at it foremost third, and the right and left lobes are united by an isthmus of simple (embryonic) cartilage—the “ basi-hyal” tract (b./y.). Wedged in behind this band of arrested tissue we see a large oval azygous cartilage with a rounded lobe growing from its hinder broad end: this is the Ist basi- branchial with a rudiment of the 2nd (b.br.). Overlying this rudiment, and articulating with the main piece, there is a pair of winged cartilages, closely applied each to each at the mid-line: these are the common “hypo-branchial ” plates (h.br.). 24 MR. W. K. PARKER ON THE STRUCTURE AND Broad in front, and rapidly narrowing backwards, they each have four small tinger- shaped cartilages attached to them, the hindmost of these is confluent with the great plate and the rest are merely articulated with it: these are the rudimentary “ cerato- branchials ” (c.b7!™,). These rays or rudimentary intra-branchial arches lie in the floor of the throat, above and within the branchial pouches ; but outside and below these there are four pairs of large extra-branchial cartilages (ew.br.) that form most of the framework of the pouches and lie close within the outer skin—are, indeed, subcutaneous cartilages, like the “labials,” and do not belong to the category of true ‘visceral arches.” The foremost of these remains free distally, articulating with the widest part of the hypo-branchial plate ; but the rest have become confluent with it : they all four coalesce with each other, above, like their counterparts in the Lamprey’s branchial basket-work. The jirst and fourth of these cartilages are bags or pouches folded into many hills and hollows; the two in the middle are widish bars. The first and second from their hind margin, and the other two from their fore margin, send inward tooth-like rudiments of branchial rays: these agree with the binding parts that make the bars into a basket-work in the Lamprey. In the Sharks (see “ On the Selachian Skulls,” T. Z. S., vol. x., plate 38, figs. land 2; and GEGENBAUR’S ‘Selachians,’ plates 11-20) we see that the extra-branchials are feeble, the intra-branchials typical, and the branchial rays distinct; in the Skates or Rays these fringing cartilages are much dilated externally, and these dilatations are articulated together and suppress the extra-branchials. The two upper labials (Plate 3, figs. 1-3, w./.) are but little united at the middle, and each has a fenestra in it; on the right side there is a lesser hole. The labials of the sucking disk (/.J.) are larger than in the last instance, and this corresponds with the fact that this species is very generalised and archaic. The spiracular cartilage (Plate 3, figs. 1, 2, sp.c.) is confluent with both the “ tegmen tympani” and the “ otic process” of the suspensorium. When these parts are studied, not only after dissection but also with all their enclothing structures, we find that the ordinary Batrachian larva is indeed a kind of “Cyclostomous” fish, and the Lamprey and this larva mutually explain each other. Palatal and lingual aspects and a longitudinally vertical section of the head of the largest of these Tadpoles (5 inches long) show many instructive things.* In the first view (Plate 1, fig. 3), the inferior arches have been removed bodily, and the palate turned upwards; we thus get a view of the horny jaw-plates, the oral papilla, the quadrate hinge, the condyle for the hyoid, the inner nares, and the whole of the palate up to the faucial region. The upper part of this Petromyzine mouth shows not ely the principal crescentic * Professor Huxiry (on “ Petromyzon,”’ Jour. of Anat. and Phys., vol. x., plates 17, 18, pp. 412-429) has served as “pioneer” in this part of the thicket; and a cleared space, such as he has made, makes my work comparatively easy. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 295 horny plate, convex outside and concave within, but also lesser denticulate horny pro- ductions (0.7.) on the outer folds of the lips, and numerous thick papillze (0.p.), laterally. The main horny plate (Plate 1, fig. 3) forms a sheath to the upper labials (w./.) ; behind this the mucous membrane is seen as a bilobate cushion, marked off behind by a deep sulcus which runs from one quadrate hinge (q.c.) to the other. Behind this suleus the mucous membrane is raised into a second cushion, which is concave in front, and which passes backwards into the large palatal roof. At the mid- line its foremost third is marked off by an “anterior palatal velum” (a.v.) (“median triangular papilla” of Huxuery, op. cit., p. 429); this is covered with small papille, is semi-oval in shape, and is curved forwards; its attached edge is crescentic, the convexity looking backwards. In front of this flap there are two long clefts (i.n.) which approach each other at the mid-line, and then greatly diverge, forwards, nearly reaching the exposed quadrate hinge (q.). Each cleft is lipped in front, and this lip runs back again along the cleft dividing it into two chinks; this apparently double slit is guarded by two large papille : it is the internal nostril. In the chondrocranium this passage appears as a large round hole between the trabecular cornua and the quadrate (Plate 3, figs. 1, 2, in., ¢.tr., g.) and only partially enclosed by cartilage in front. I have just described these parts. At the sides, opposite the “anterior velum,” the condyle on the suspensorium for the hyoid cornu is exposed (hy.f.). The interorbital region of the palate is lower than either the internasal or the inter- auditory ; it is one-fourth longer than broad, is covered with papille, and has the largest of these productions postero-laterally. The hind part of this palatal tract is, properly speaking, fawcial; it is beneath the ear-sacs, and reaches the basioccipital region. This is separated from the proper palate by two “ posterior vela” (p.v.); these are crescentic folds of mucous membrane, with their convexities forward, or opposite to that of the “anterior median velum ” (c.v.). The floor of the mouth is shown from above (fig. 4); the free mandibles (mk.) and the hyoid cornua (c.hy., hy.c.) have their upper or articular ends exposed. The inferior horny plate, with the outer denticles of the folds of the lower lip (Plate 1, fig. 4) are shown, as well as the right and left crops of labial papille ; the main horny plate is modelled on the horse-shoe shaped lower labials (/./.). In this view all the post-oral clefts are seen, for at present the pouch-shaped (blind) first cleft (c/'.) is low in position. I have found it open externally in no stage, although the skin over it is very thin at the time of hatching. (See my first paper on the “ Frog’s Skull,” Phil. Trans., 1871, Plate 3, fig. 10, c/'.) This low position of the first cleft is due to the extreme length and horizontal position of the “pier” of the mandibular arch (see Plate 3, figs. 1, 2), the hinge of which (fig. 1, q.) nearly reaches the upper lip; afterwards, in the adult, it is beneath the ear. This cleft (Plate 1, fig. 4, cl!.) is small, crescentic, and with its convexity looking MDCCCLXXXI. E 26 MR. W. K. PARKER ON THE STRUCTURE AND outwards and forwards ; all the other clefts (c/ *~*.) have the same form and direction, but they are larger, and open externally, although covered afterwards with the oper- cular skin, which only opens on the left side. The second cleft (c/*.) is between the hyoid cornu and the first branchial pouch ; it runs into the floor some distance from the hyoid facet. Between the articular surfaces of the free mandibles (mk.) the mucous membrane is raised into two crescentic folds ; these are the lateral rudiments of the tongue (tg.) ; if these lobes simply grew forwards the Frog would have a bilobate tongue with its Sree ends growing forwards—it would be protero-glossal. Between these lobes there are several large papillae; behind them the mucous membrane thinly veils the hyo-branchial arches (Plate 1, fig. 4, and Plate 2, fig. 8). At the mid-line there are, behind the papille, first the basi-hyal (b.hy.) and then the double basi-branchial (b.br.) ; whilst on each side are the broad hyoid cornua (c.hy.) and the sub-median and lateral elements of the branchial arches (/.br.) and its rays —all that exists of the endoskeletal framework of these parts. Thus in the larval condition the Frog has the dorsal portion of its hyoid and branchial arches suppressed : four large “extra-branchials” (Plate 1, fig. 4, br.p.; and Plate 2, fig. 8, eaw.br'~*) supplement the deficiency, and those bars, the first and last of which are cochleate, are homologous with half of the branchial basket-work of the Lamprey, namely, with the four bars of its framework from the first to the third pouch inclusive. In his valuable paper on the Lamprey (Journ. of Anat. and Phys., vol. 10, pp. 412— 429), Professor Huxiey does not notice the existence of arches homologous with the proper branchials of Selachians, Ganoids, and Teleostei; but his description of the cor- respondence of these parts with those of the Lamprey must be given here :— “Tn the present stage the branchize of the Tadpole are, as is well known, pouches, which present no superficial likeness to the branchial sacs of the Lamprey. A septum extends inwards from the concave face of each branchial arch, and the septa of the two middle arches terminate in free edges in the branchial dilatation of the pharynx. Vascular branchial tufts beset the whole convex outer edge of the branchial arch, and are continued inwards in parallel transverse series of elevations, which become smaller and smaller towards the free edge of each septum, near which they cease. “In the young Ammoceete the septa of the branchial chambers similarly bear vascular processes, which are first developed close to the external branchial aperture, and thence extend inwards transversely.* * In a note here, Professor Huxtny remarks: “If the first-formed long branchial filaments of the Ammocoete projected through the small gill-clefts outwards instead of inwards, they would resemble the first-formed ‘external gills’ of Elasmobranchs. And this difference of direction seems to indicate the solution of the difficulty, that external gills, which are so generally developed at first in Hlasmobranchii, Ganoidei, and Dipnoi, are apparently wanting in Muarsipobranchii.’ I may add to this note an observation of my own illustrative of the writer’s remarks, namely, that in the larva of Dactylethra there are no branchial filaments growing from the outer face of the branchial pouches (see “ Batrachia,” Part II., Pilate 58, fig. 1.) DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 27 “The recesses at the sides of the floor of the pharynx into which the interseptal clefts or internal branchial clefts open, answer, taken together, to the branchial canal of the Lamprey, which is not shut off from the cesophagus in the Ammoceete. The anterior boundary of each of these recesses is marked by a fold of the mucous mem- brane, the free edge of which projects backwards, and is produced into papilliform angulations so as to appear scalloped. The anterior face is concave. The inner angle of each fold passes into its fellow by a ridge, produced into one or two papilla, which is closely adherent to the median part of the floor of the mouth. The outer angle is continued into a more delicate fold of the mucous membrane lining the roof of the mouth, the free edge of which also projects backwards. It is plain that these struc- tures answer to the pharyngeal velum of the Lamprey.” The “inferior velum,” or membranous fold (/.v.) will be best understood by reference to both the figures of its natural condition (Plate 1, fig. 4) and also to those of its dissected framework (Plate 2, fig. 8). We see at once that this scalloped fold is formed upon the rudimentary arches that grow from the oblique sides of the hypo-branchial plates, and that there is the common pharyngeal covering of these rudimentary itra-branchial arches (‘ cerato-branchials’’). They are able to be the skeleton of a free, scalloped, papillated fold in virtue of their arrested condition ; if they grew upwards round the whole circle of the throat this structure could not exist. If each of these rudimentary arches was continued upwards, surmounted on each side by an “epi-branchial,” and this in turn by a “pharyngo-branchial,” we should have such arches as are seen in Selachians, Ganoids, and Teleosteans. If the lower, or outer, surface of these arches was beset with a double row of pecti- nated branchial folds, then the Ganoid and Teleostean type of gills would exist; in the Selachians (“branchiis fixis”) the gill-folds are formed in a double series of pouches. the common framework of which is formed of typical intra-branchial arches, each composed of nine pieces, on the inner side; the septa are strengthened by free “branchial rays,” and on the outside (in the Shark) there are distinct “ extra- branchials,” one to each pouch, which are pointed above and pedate below. The “branchial canal” of the Lamprey is correlated with suppression of the cerato-branchial ” rudiments, such as are seen in the Tadpole; but there is a hyoid arch. With respect to the branchial tufts that are so copiously developed within and at the edges of the gill-pouches of the “Phaneroglossa,” I find that my own views are in accord with those of Mr. Batrour.* There are true and false “external gills”: the first, only, are present in the “Urodeles,” and for a short time in the newly-hatched Tadpole (see Phil. Trans., 1871, Plate 3, figs. 2 and 10, br'., br*.); these are developed from the epidermis—are epi- blastic. But the tufts that break forth from the clefts in the more developed Tadpole, * T have recently had my mind set at rest upon this subject by my talented friend. E 2 28 MR. W. K. PARKER ON THE STRUCTURE AND and the long, precocious, filamentous gills of the embryo Selachian, are, like gills generally, of hypoblastic origin. The vertical longitudinal section (fig. 5) may be now described. The large upper labial (below c.tr.) is seen in section; it is the “anterior dorsal cartilage” of J. Mitier: below, the lower labial is seen,* where it joins its fellow to form the imperfect “annular cartilage” (see also fig. 4, /.l., mk.). Mecket’s cartilage is seen behind this in the section; it lies in the side of the mouth, and its main function is to serve as a suspensorium to the “ annular cartilage ” (the corresponding half of the imperfect ring) (see also Plate 2, fig. 6, /.1., mk.). The raised cushion behind these cartilages is the rudimentary tongue (fig. 4, tg.) ; in the back of this the basal cartilages are placed. In most things I agree with Professor Huxtey as to his harmony of the Lamprey’s and the Tadpole’s mouth. There is some reason to suppose that the antero-inferior median cartilage belongs to the mandibular arch. The basi-hyal of Fishes is always a “glosso-hyal;” it projects forwards beyond the arch to which it belongs, and so does the first basi-branchial. Any paired plate of cartilage lying in the same plane, but projecting outwards and backwards from the basi-branchial bar (or bars) would be “hypo-visceral;” any rods attached above these and lying in the lower half of the sides of the pharynx would be “ cerato-viscerals.” In the Lamprey (Huxtey, op. cit., plate 17, fig. 1) this anterior median cartilage, below, is outside the lingual rod, and is probably a basi-mandibular ; and the great lingual cartilage (k.) is an undivided piece, with neither osseous nor fibrous segmenta- tions, and is evidently an intra-visceral element. In the Tadpole (Plate 2, fig. 8, b.ky.) the anterior basal piece does not chondrify until metamorphosis takes place; it is a thick, narrow, conjugating band, composed of large cells of simple cartilage. But the basi-branchial of the Tadpole (6.br.) is an oval mass of cartilage, with a hinder bud that grows downward beneath the large hypo-branchial plates (h.br.) ; this bud does not become a separate second basi-branchial. These two semi-distinct rudiments do not represent more than the end of the long lingual cartilage of the Lamprey; and if that fish had paired cartilaginous plates lying on the same plane as the lingual cartilage and attached to its sides, behind, they would be hypo-branchial plates, for they would belong to the inner category of cartilages, and not to those forming the branchial basket-work. In the Tadpole the heart (fig. 5, h.) is roofed over by the “ hypo-branchial” plates, and the edge of the “inferior velum” (fig. 4, l.v.) is seen completing this roof and also covering the inner branchial openings in front. Where the right and left vela unite at the mid-line, there they are directly in front of the larynx (/z.); afterwards, the hypo-branchials become the “ thyro-hyals.” * The inner face of the lower labial is lettered msk., by mistake, in Plate 1, fig. 5. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 29 Third Stage.—1 (continued). Tadpole of Rana clamata (B) 3 inches 5 lines long; tail 21 inches long; hind legs, 13 inch long. In this stage (Plate 4, figs. 1-4) we see some metamorphic progress beyond the last ; the general form of this as compared with Stage A (same species) is not much altered (see Plate 2, figs. 5-7), but there are some new parts, and some changes in the oid. The time, indeed, has been short during which these changes have taken place, for these Tadpoles even at their height, as larvee, were beginning rapidly to be trans- formed. The exoccipituls (¢.0.) are larger, but the prootics are not visible ; outside the frontal region the premaxillaries (px.) have appeared in addition to the three investing bones seen in the last stage. But the thing of most importance is the rapid growth of the “intertrabecula ” as a vertical nasal septum (s.n.), and the solidification of the mem- branous roof of the nasal sacs into a pair of ear-shaped nasal cartilages (na.) whose antero-external notch forms the inner boundary of the outer nostril (na.,e.7.). These elegant shells of cartilage are becoming confluent with the top of the septum internally, and with the ethmoidal “ wing” (a/.e.), behind: they are true “ paraneurals,” and answer to the ‘“‘sclerotics” and “ periotics.” Another change is the retreating of the orbitar process away from the side of the skull, so that the “post-palatine” rudiment is exposed (0r.p., pt.pa.); the “‘pre-palatine” rudiment (pr.pa.) in front of the internal nostril (i.7.) is also further from the trabecula. The fore margin of the chondrocranium is very similar to what it will be eventually, but the rounded end of the intertrabecula is still distinct from the nasal roofs (s.n., na.), and the cornua trabeculee (c.tr.) are still undivided at their end. The great upper labial (w./.) has lost the fenestra that showed signs of division, right and left, and is now, more than ever, one large saddle-like plate.* The mandibular and hyoid arches are much in the same state as in the younger larvee (Stage A), but the branchials are altering preparatory to extensive absorption. Compared with those of the Bull-frog (Second Stage, Plate 2, fig. 8), it will be seen * Tn this and the last two stages (A and B) we have a very instructive analysis, so to speak, of the complex nasal capsule in the higher Vertebrates, ¢.g., of that in the Mammals. It is composed primarily of the trabecule with their lateral (antorbital) wings, where the cranium closes in in front of the vertical “ intertrabecular ” middle wall; of the nasal roofs which may grow down into side walls ; and of the rich outgrowths of turbinals that grow from the front of the antorbital wings of the ethmoid (upper and middle turbinals) ; and from the inside of the outer wall (inferior turbinals). Dther turbinals may spring from the roof (nasal turbinals of the Rodents), or inside the outer nostril (ali-nasal turbinals of Pig and Bird). Then the narial valves are formed of segments or subdivisions of the upper labials, the whole structure being roofed by the nasals, floored by the vomer or yomers and palatine plate of the maxillaries, and walled in by the maxillaries and premaxillaries. : Lastly this complex labyrinth of cartilage may, itself, become bony, to a greater or less degree, and even may coalesce, more or less, with the investing bones. Anyhow, the plan and pattern, and the nwmbered elements of this labyrinth, are to be seen in their first simplicity in the nasal region of the Tadpole. 30 MR. W. K. PARKER ON THE STRUCTURE AND that the first and second extra-branchials (Plate 4, fig. 3, ex.br'~*.) have lost much of their pouch-like shape, and are but little broader than the second and third; they all become very narrow bands, and thus vanish, with the exception of a common remnant that becomes fused with the permanent hyo-branchial plate. The third intra-branchial (c.br*.) is very long, as long indeed as the counterpart bars are in Selachians, Ganoids, and Teleostei; in its normal position, round the pharynx, this bar just reaches half way up the side, and in those Fishes is surmounted by the epi-branchial. In the Dactylethra Tadpole (Phil. Trans. 1876, Plate 58, fig. 1) the whole skeleton of the branchial pouches is one continuous structure; here the intrabranchials are nearly segmented off from the extra-branchials, but there is an isthmus of cartilage uniting them in Rana, as may be seen in the section (fig. 4, ¢.br., ew.br'.). I must pass again to Rana pipiens for my next stage; the reader will, however, easily eliminate the non-essential specific variations, at any stage, from those more important step by step processes that transform a Cyclostomous skull into one with a widely gaping mouth and jaw-hinge far behind, like that we see in certain aberrant Selachians—WNotidanus and Cestracion. Fourth Stage.—2 (continued). Tadpoles of Rana pipiens (C) with tail lessening and all the legs free. This stage (Plate 3, figs. 4-13, and Plate 9, figs. 1-6), of this species, shows that in the largest and most generalised of the genus the branchial apparatus remains in full development and function much longer than in the lesser Frogs. They are two or three years (according to Wyman) before they have finished their metamorphosis, and that excellent naturalist kept them (in captivity) six or seven years as Tadpoles. In R. clamata (Stage B) the nasal roofs just appear as the branchial pouches are dwindling away. In &. pipiens the whole nasal labyrinth is perfected before any sign of diminution appears in the gill-pouches. This indicates that the Tadpole of the Bull Frog is well fitted out with special sense organs long before it finishes its metamorphosis, and suggests the probability of a still slower metamorphosis in former epochs. Psewdis offers a remarkable confirmation of this view. Comparing this with the two last stages we shall see both the general progress and the specific differences that are found to be even in larve of the same genus. : The most striking change is the elongation of the palato-pterygoid band (Plate 3, figs. 4-6, p.pg.), thrusting the suspensorium outwards. Then the “waist” of the cranial box is narrower, and its sides more chondrified (fig. 6). The roof bones are segmented across (fig. 4, f, p.), and, contrary to wont, the frontals are only half the size of the parietals. Below, the parasphenoid (fig. 5, pa.s.) has become more developed, and foliated behind; in front, the membranous layer seen in the earlier larva (fig. 2) is now partly DEVELOPMENT OF THE SKULL IN THE BATRACHIA. Bl occupied with a new (and rare) bony centre—the “ pro-parasphenoid ” (pa.s’.) The endosteal tracts in front of and behind the auditory capsules (pr.o., ¢.0.) are enlarging, and the “ pedicles” (pd.) in front of the former are becoming very attenuated. The rest of the hind part of the cranium is but little altered, but in front the nasal capsule is almost finished (fig. 4). The nasal roofs (al.n.) are now completely confluent with the septum and the lateral wings of the ethmoid ; nevertheless the “ cornua trabecule ” (c.tr.) that form the floor of the nasal cavities are undivided in front. The ear-shaped nasal roofs are seen to have curved down, round the external nostril (e.n.), but the terminal halves of the trabecular cornua are interposed between the nasal capsules and the upper labials. These, however (figs. 4-6, w./.), are seen to be breaking up into an inner and an outer pair: these will disappear; the new labials and the newly formed premaxillaries have been removed. The inferior arches (Plate 3, figs. 5, 6, and Plate 9, fig. 1) are but little altered from what they were in the Second Stage. The drawing of the branchial arches was made from an outspread preparation which was figured from below ; they are seen to turn obliquely backwards as they ascend to the sides of the basis cranii. Tooth-like processes of simple cartilage are interdigated with each other in the large clefts; but the branchiz which grow on them, and which run inwards in rows, have been removed. But truly cartilaginous rays grow from the bars under these processes ; these are partly shown in Plate 9, fig. 3, br.r. The internal nostril (Plate 3, fig. 5; Plate 9, fig. 2, 77.) is a curious oblong slit, diverging forwards, and protected by one or more papillee. A series of transversely vertical sections (Plate 3, figs. 7-13; and Plate 9, figs. 4-6) show the solidity of this chondrocranium. The 1st section (Plate 3, fig. 7) is through the fore part of the nasal capsule, and catches the nasal roofs (al.sp.) both above, and where they turn over to form the narial rim, or ‘“‘alinasal” fold. They have coalesced with the intertrabecular wall, now the septum nasi (s.7.). A branch of the orbito-nasal nerve (5!) is seen emerging between the folds; the cavity between these folds externally is the outer nostril; another space is seen below this and above the thick cornua trabeculee (c.tr.), which runs inwards to the inner nostril. At present the trabecular horns are not modified into the flat nasal floor, but are attached to the structure above by fibrous bands. The 2nd section (fig. 8) is close behind the nostril, and behind the second fold of cartilage ; on one side the wide part of the roof has been (obliquely) cut through ; the septum (s.7.) is still some distance from the cornua (c.tr.; the cross band is coloured, by mistake). The 3rd section (fig. 9) is similar; it is also oblique, and the hinder part of the capsule is partly severed. The 4th section (fig. 10) is through the widest part of the nasal roof, where these 32 MR. W. K. PARKER ON THE STRUCTURE AND cartilages (fig. 9, al.sp.) have coalesced with the ethmoidal tract (al.e.); here the septum (s.7.) as it becomes “ perpendicular ethmoid ” (p.e.) is confluent below with the coalesced trabeculze (t7.). On one side the section is seen to pass into the palato- pterygoid band (p.pg.). The 5th section (Plate 9, fig. 4), rather more than half of which is shown, is through the ethmoidal part of the skull proper, where the nasal cavities (7.p.) are closing in. The perpendicular ethmoid is seen to be very thick, and the trabecula (tr.) to pass into the solid root of the palato-pterygoid bar (p.pg.) from which ascends, suddenly, on the outside, the strong arcuate “ orbitar process ” of the suspensorium (o7./p.). The 6th section (Plate 8, fig. 11) is through the solid ethmoid where it closes in the skull, leaving only the olfactory nerve-passages (1); here the skull is as solid as that of a Chimera, The palato-pterygoid bars (p.pg.) are cut through, outside the “ post- palatine rudiment” (see Plate 3, fig. 1, pt.pa.), whose form, as a gentle elevation, is there shown. The 7th section (Plate 3, fig. 12) is behind the palato-pterygoid, and through the fore part of the skull-barge, where the olfactory crura (1) are given off. The cavity is elliptical, the wide wall thick, the roof (anterior “ tegmen,” t.cr.) and the floor are gently concave. On the roof we see the fore end of the frontals (f), and under the floor the parasphenoid (pa.s.). The 8th section (Plate 3, fig. 13) is behind the “tegmen” and through the fore part of the “ fontanelle,” covered by the widest part of the frontals (f); the bone below is widening, it is the parasphenoid (pa.s.). The elliptical cavity is deeper, the walls thinner, and the floor is now convex. The concavity, below, in the two last sections, is due to the drawing in of the trabecul, and the high, vertical form of the “intertrabecular” tract in the fore skull, is now a bulging floor. The 9th section (Plate 9, fig. 5) is through the middle tegmen, between the great median and the lesser paired “fontanelles,” and also through the widening parietals (f, by mistake for p.). Here the roof passes into the wall, downwards; then there is a large fenestra optica on each side, with the optic nerves (II.) emerging; and then the rounded trabeculze, with the bulging “ intertrabecular tract,” floored by the para- sphenoid (t7., pa.s.). The 10th section (Plate 9, fig. 6) is through a lateral fontanelle, the hind part of the parietal (p.), and the auditory capsule, confluent with the basal plate (iv.). The anterior and horizontal canals («s.c., h.s.c.) are cut through, and the general cavity of the vestibule (au.) laid open. The “tegmen tympani” (¢ty.) is seen projecting from the outside of the capsule, beyond the horizontal canal ; and the basal plate, or “investing mass,” is seen to grow outwards as a floor to the tympanic cavity (ty,f.). This plate is deticient, more or less, behind, the capsule resting on an imperfect floor below, and projecting into a fenestra in the wall of the skull. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 33 fifth Stage.—1 (continued). Advanced Tadpoles of Rana clamata (C), all limbs free ; 31 inches long; tail, 144 inch long. If this (Plate 4, figs. 5-7) be compared with the Third Stage, the second larva of FR. clamata, (figs. 1, 2) we shall see what approach is made towards the adult condition. The angular and almost oblong form of the whole structure is changed for a short oval ; and the facial outworks are only half the real and relative width they possessed then. The auditory capsules are of greater extent, antero-posteriorly, the ex-occipitals (¢.0.) are larger, and the prootics (p7.0.) have begun on the outside of the foramen ovale (V.). The form of the cranial cavity is narrower in front compared with the hind part, and the whole ethmo-nasal structure in front of the cavity is shorter, and much modified. Besides the septum nasi (s.7.) the whole intertrabecular space is filled in in front ; the “cornua” (c.t7.) are only half as long, and are bifurcated, the outer fork has become the large arcuate angle of the ‘‘subnasal lamina,” and the lesser (inner) fork is the hooked “ pro-rhinal” (p.rh.) which lies inside the premaxillary (p«.), and curves backwards. The internal nares (i.n.) are now relatively much forwarder, are more elongated, and are placed quite transversely. The front wall of the nose is completed by fusion of the roof with the septum, and a growth of cartilage over it to the lower face of the labyrinth. The external nostrils (e.n.) are now also much nearer the front of the head, for the large labial plate has been absorbed and two pairs of small new plates that form the outworks of the nose have appeared. The inner pair (w.l".) are applied to the inner face of the apex of each nasal process of the premaxillaries (px.); whilst the outer forms a shell-like valve to the external margin of the nostril (w./?., e.7.). The nasal bones (a film over na.) have appeared like small shells. The condyle of the quadrate (g.) is now opposite the hinder margin of the nasal roof, instead of the front margin ; it is twice as far from the roof, and its condyle looks obliquely outwards, and not inwards as before. The elongated and oblique palato-pterygoid bar now has the “ post-palatine rudiment ” as a ridge looking outwards and backwards, and the “ pre- palatine spur” (pr.pa.) nearly touches the tip of the horn of the trabecula (angle of “sub-nasal lamina,” c.tr.). Tracing the narrowing and retreating suspensorium backwards, we see that the orbitar process, and condyle for the hyoid (or.p., hy.) are now opposite, not the antorbital region, but the middle of the subocular fenestra. Further back, the true “otic process” is small and rounded (fig. 5, of.p.), and the spiracular cartilage, now detached from the tegmen tympani, applies itself to the side of the otic process (sp.c., ot.p., t.ty.); it is now a trifoliate rudiment of the “annulus.” Below (fig. 6, pd.), the pedicle is forming its condyle, to glide upon a pre-auditory process of the basal plate, whilst the dorsal part of the pier is now an attenuated thread of cartilage running under the prootic to the front of the foramen ovale (V.). MDCCCLXXXI. F 34 MR. W. K. PARKER ON THE STRUCTURE AND A remnant of the notochord (nc.) still persists; it lies on a thin junctional tract of the investing mass through which it is very visible. The fronto-parietals (fip.) are but little changed, and are not yet segmented. The parasphenoid (pa.s.) has enlarged in its basi-temporal wings, in correspondence with the enlarging ear-capsules. There is only a stapes in the middle ear; no epi-hyal element has appeared. In accordance with the out-turning of the quadrate condyles, the lower jaws (fig. 7, mk.) are longer ; but the whole arch is largely extended by the strangely altered form of the original suctorial “ horseshoe” formed by the “lower labials” (/./.). Now, they are nearly straight, and stand across between the ventral ends of the mandibles. These latter bars are invested on their inner and lower face by the rudimentary “articulare” (ar.). The maxillaries (mz.) also have appeared above. Sixth Stage.—3. Half metamorphosed larva of Rana tail, + inch; all the legs free. 2 sp. (India) ; 14 inch long ; The less important specific modification here seen (Plate 4, figs. 8-10) is the flat, wide cranium; the more important morphological changes are self-evident. The transversely extended auditory masses, and the wide flat occipital arch, are now largely becoming ossified by the two pairs of perineural centres (pr.o., ¢.o.) that gradually more or less occupy the occipital, periotic, and post-sphenoidal regions, protecting those parts of the head where the 5th to the 10th nerves emerge. The rest of the cranial “barge” is well chondrified, but has no bone as yet; the roof bones (f-p.) are still large; the floor bone (pa.s.) is a short, blunt, dagger, with a very wide “ guard,” and a very short handle. Over the nasal roof the nasal bones (n.) are now crescentic shells of bone ; the pre- maxillaries and maxillaries (px., ma.) are now well developed, but the vomers are not apparent. A thin squamosal (sq.) like a “ pre-operculum” runs down the suspensorium, and the dentary (fig. 10, d.) has appeared on the outside of the mandible, distally ; the coalesced “lower labial” is becoming ossified, and the dentary is grafting itself upon the endosteal patch, so as to form the “ mento-Meckelian bone ;” the “ articulare ” (ar.) is lengthening with the jaw. The fore parts of the trabecule, now completely confluent with the nasal roofs, are broken up into a remarkable cervicorn structure. The inner angle of each is now the finger shaped, curved “ pro-rhinal” (p.rh.), and the outer angle, which finishes the nasal floor in front, is divided into three lobes, one upper and two lower, that turn outwards, and are imbedded in the fore end of the maxillary. The two pairs of upper labials (u.J1.u.2.) are quite normal now, perfecting the antero-external edge of the nostril (fig. 8). Behind the pro-rhinals a pair of ‘‘ fenestrae” (fig. 9, 1.7.) are seen. The “ pre-palatine ” (p7.pa.) has now escaped far away from the trabecular angle in front ; it is a sharp faleate process at the fore end of a narrow band of cartilage which runs backwards until it is now opposite the prootic bone; there it ends in the suspen- sorium, which has become a stunted, triradiate cartilage, one-third of its former length. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. a) The antorbital pedicle (“ethmo-palatine”) (e.pa.) is now a flat tape, narrowest in the middle; the “post-palatine” (pt.pa.) is a similar tape passing insensibly into the still narrower “pterygoid” (pg.), which latter passes into the suspensorium where it subdivides into its three forks. The lowest part ends in the reniform quadrate condyle (9.); and the position of the hinge is opposite the prooticand the foramen ovale (pr.o., V.), instead of being in its old place opposite the front of the nasal capsules. This is, truly, only part of the way back of the condyle (see in the adult Rana pipiens, Plate 8, where it gets some distance behind the occipital condyle), but this is a good distance to be travelled whilst the legs have been growing and emerging, and the lungs gradually rendering the gills unnecessary. The spiracular cartilage (sp.c.) has begun to take on its crescentic form, and is now an evident annulus tympanicus, The epi-hyal has not yet appeared, ready to become the columella; the cerato-hyals (fig. 10, chy.) are dislocated from the suspensorium, the hinge having become absorbed ; they lie behind the first pair of clefts, which, of course, are between them (on each side) and the corresponding suspensorium. This latter part, the mandibular “ pier,’ has lost all its dorsal end, the narrow upper tract of the “ pedicle” having been completely absorbed. The stunted, ampu- tated part has now a flat condyle on its end, which glides on the facet formed for it over the front of the ear-capsule below. Above, the “ otic process” (or third part) of the suspensorium (ot.p.) has crept close to the fore edge of the tegmen tympani, ready for fusion ; it is already invested with the squamosal (sg.), which lies in front of a gentle ridge—all that remains (now) of the “orbitar process.” The arches, pouches, sub-basal and basal plates of the branchial cartilages are reduced to a lozenge-shaped plate, ending in a pair of diverging rods, and ready to unite, in front, with the attenuating cerato-hyals (fig. 10, c.hy., b.hbr., t.hy.). I shall next describe the condition of the skull in young Frogs, when that which was left unfinished in the cranium, on their assumption of terrestrial life, will be seen to have gone on unto perfection ; and the “headstone” brought on to this graceful piece of vegetative architecture. SKULLS oF YOUNG AND ApbuULT “ RANID.” Seventh Stage.*—4. Skull of Rana palustris (Cambridge, Mass., U.S.). Young recently metamorphosed ; 11 lines long. ges show the endo- The figures of this (Plate 5, figs. 6-10) and the next (figs. 1-5) sta cranium with the outer bony laminz removed from one side, and retained on the other ; and these, whether they are bones that are permanently, or only for a time, distinct. * Tf the three early stages of the larval skull (in Bufo vulgaris) were added to these, this would be the Tenth Stage, 4 36 MR. W. K. PARKER ON THE STRUCTURE AND All the structures are now rapidly passing into their permanent form, although the individuals are, at this time, very small. The endocranium is here of the typical form ; a flat-bottomed “barge,” with one large anterior, and two small posterior membranous spaces in the “ deck.” The larger lateral expansions behind are connected with the lesser lateral expan- sions in front, by an elegant bow of cartilage, the pterygo-palatine (pg., e.pa.) ; the hinder expansions are caused by the impaction of the ewr-masses into the sides of the skull, and the expansions in front are due to the nasal growths; from the open spaces between, on each side, the eye-balls have been removed. The two pairs of ossifications in the endocranium, the prootics and the ex-occipitals (pr.0., €.0.), are considerably larger than in the last stages; the hinder pair have quite encircled the twin passage for the 9th and 10th nerves ; a trace of the notochord (nc.) still remains. There is no girdle-bone yet ; thus, with the exception of the “centres” just men- tioned, the endocranium is a “chondrocranium.” The form of the contents of the ear-capsule is well seen through the semi-transparent cartilage, and the form and extent of the originally separate nasal roofs, as distinguished from the pre-cerebral region of the cranium, are well seen. Also below (fig. 7), the manner in which the “intertrabecula” has filled in the space between the “cornua trabecule,” and the sub-division of the root of the “horns” into a pro-rhinal hook, and a subnasal outer angle (p.rh., s.n.J.), are clearly seen; also the divisions of the upper labials (fig. 6, u.l'.u.E.), and their relation to the outer nostril (e.7.). The “ethmo-palatine” bar (e.pa.) is lobate; the pre-palatine is spiked, and the post-palatine is a gently lessening bar, which with the pre-palatine region in front and pterygoid region (pg.) behind, forms a most elegant subocular “ bow.” Behind, this bowed bar ends in three lobes ; the quadrate with its reniform condyle (q.¢.) running outwards, and downwards, and backwards ; the “pedicle” (pd.) pedate, with a flat inturned facet below; and the “otic process” above, clamping the outer and front end of the “tegmen tympani.” For articulation with the quadrate we see the cylindroidal condyle of the mandible (fig. 8, a7.c.); the under and inner face of the long arched rod is invested by the “articulare” (ar.) nearly to the fore end; the ‘“ dentary” (d.) runs along the distal half on the outside, and it is grafting itself, near its lower end on the lower labial, which is not quite confluent with the mandibular rod (mk.), and which is itself ossifying, to become the “ mento-Meckelian” bone (m.mk.). The obliquely semi-oval stapes (figs. 7 and 10, st.) has now wedged in, between its fore margin and the ear-capsule, the proximal end of the “ columella” (epi-hyal element). This structure is not yet finished, and has a subdivision in it very rarely seen in adults, but which is normal in certain fishes (e.g., the “ Acipenseride ”), These two parts are, morphologically, the proximal or “hyomandibular,” and the distal, DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 37 or “symplectic.” Asa rule a short segment is cut off close to the stapes, the morpholo- gical ‘ pharyngo-hyal ;” this has not taken place, at present, in this young Frog. The proximal piece is the medio-stapedial (fig. 10, m.st., 7.st.), and it is largely ossified ; the distal piece is not ; it is the “ extra-stapedial ” (e.st.)—a pedunculated sub-circular disk ; as yet, there is no supra-stapedial band growing up from the inner side.* The “ cerato-hyal” (figs. 7, 9, st.h., c.hy.) is not yet as narrow as it will be, but from hanging by its primary joint under the antorbital region, it now is articulated to the tympanic floor under the columella and behind the Eustachian opening (ew.), to which it now forms the normal boundary as part of the hyoid arch, that is to say, it is behind the “1st visceral cleft ;” over that cleft, outside, the “spiracular cartilage” is closing round the membrana tympani so as to form the “annulus” (fig. 6, «.ty.). The cerato-hyal band still retains an auriform hypo-hyal lobe (fig. 9, c.hy., h.hy.), but the basi-hyal conjugation (b.hy.) is still a tract of simple cartilage running into the dilated basi-branchial plate (b.br.) ; attached to this we see on each side the remains of the branchial pouches, and another remnant is seen also on each side of the paired hypo-branchial plates (h.br.), which now end in long sigmoid rods (¢.hy.), the still unossified “ thyro-hyals.” This is the most instructive intermediate stage of these parts I have as yet succeeded in dissecting out; if the figures of these parts be compared with what is seen in the larva on one hand, and in the adult on the other, the value of this stage will be self-evident. The investing bones, the true “ parosteal” plates, are nearly all present; those of the lower arches have been already described. The frontals (fig. 6, fp., by mistake) are now, for a while, distinct from the parietals ; the latter lie over the paired fontanelles, and the frontals cover the larger space (/fo.), The nasals, premaxillaries, and maxillaries (7., px., mx.) are now quite normal, the quad- rato-jugal (q.j.) is appearing, and the squamosal (sq.) is acquiring a supra-temporal plate. Below, the parasphenoid (fig. 7, pa.s.) is quite normal, having all the hinder processes, and there is no pro-parasphenoid. The “ vomers” (fig. 7, v.) are thin crescentic plates notched in front. The palatine and pterygoid bones (pua., pg.) are now quite separable from the subjacent cartilage, but that is only a temporary state of things in most cases. Eighth Stage.—5. Skull of Rana halecina (N. America) ; 13 inch long. This young Frog was about twice the size of the last, but the species are quite distinct ; this is the most welcome intermediate form I have found between the typical R. temporaria and the gigantic R. pipiens. The cranium is much more advanced, but it is still unfinished ; in this respect it is a good link between the young &. palustris and the adult &. temporaria. In some * In Pseudis (“ Cystignathide”) I shall be able to show two earlier stages of the columella than this (Plates 11 and 12). 38 MR. W. K. PARKER ON THE STRUCTURE AND respects it corresponds with the Oriental Frogs, and differs from most of those of the Nearctic and Palearctic regions; this is seen in the presence of an “anterior super- orbital” expansion of the chondrocranium (Plate 5, fig. 1). This is an unusually narrow and elongated skull, and thus differs from that ot R. pipiens which is short and wide ; the last was elongated but did not narrow in so much in front. This sub-triangular form, the relatively elongated nasal region, and the gradual, but great, widening from before backwards of the cranial “ barge,” are very characteristic things in the skull of this species. Add to these the rudimentary super- orbital plate in front at the narrow end of the skull, the wide temporal region, the heart-shaped great fontanelle, and the small lesser fontanelles, and we get a number ot things worth notice. Yet these are of secondury morphological importance, their value is largely Zoological and Taxonomic. In the nasal region we see that the outer angle of the “subnasal” cartilage (s.n./.) has formed a retral lobe—a part very distinct in the adult R. temporaria. The pro-rhinals (p.rh.) are rather retral than out-turned ; the upper and lower cartilaginous lamin are curiously alternated as to their wide and narrow ends; the upper is narrow in front, and the lower wide, and vice ver'sd. The bending of the palato-pterygoid “bow” is greater here than in the last, and the proper suspensorium is modified by the more backward position of the quadrate condyle (q.c.), by the fusion of the otic process with the “tegmen tympani” (fig. 1, between the letters «.s.c. and h.s.c.), and the greater perfectness of the articular facet of the pedicle ( pd.). The ex-occipitals (e.0.) are now large, and leave only a narrow oblong basioccipital space, and a wider triangular superoccipital tract of cartilage; they have risen over their own roof, and up the inside of the “epiotic” region, partly walling in the posterior canal (p.s.c.). The prootics (p7.0.) are now typical, they enclose the great foramen ovale (V.) for its outer half below, and run round the front from thence, above; further out they have climbed up on to the auditory capsule, flanking the anterior canal on its outside, and covering the ampulla of the horizontal canal (a.s.c., h.s.c.). A pair of “sphenethmoidal” centres have appeared in the chondrocranium under the superorbital cave, and behind the ethmo-palatine bar (e.pa.); these are the (a3 symmetrical rudiments of the “girdle-bone;” they are “ ecto-ethmoids” now, and are the proper side-wall bones protecting their own nerves (the 1st), just like the prootics of the Frog, the alisphenoids of Teleostei and higher types, the orbito-sphenoids of many types, and the ex-occipitals here and everywhere. They are separated by a tract equal to their own width below.* * There is much variety in the formation of the girdle-bone, but these are its most essential parts; it may have, however, a median element below (as in Pseudophryne Bibronii), or an azygous plate above (Rana temporaria, Rappia (Hyperolius) bicolor). In Dactylethra (“ Batrachian Skull,” Part 2, Plate 59, fic. 1, s.eth.) this is a large [T-shaped bone, it is free from the cartilage below it, which is not ossified so far forwards, DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 39 The outworks of this chondrocranium are being converted into bone at various points; the quadrato-jugal (q,j.) has grafted itself on to the quadrate (distal part of suspensorium), the long faleate palatine (pa.) is grafting itself on to the ethmo-palatine bar, and the pterygoid (pgq.) same to the “ pterygo-quadrate” region. In the Batrachia generally, the pedicle ( pd.) is also more or less ossified by the proper osteo-pterygoid; but here, as in Rana temporaria and R. pipiens—that old fish- bone—the “metapterygoid” (mt.pg.) breaks out again; the possession of this centre a long sigmoid bone with an internal snag—is doing the is, perhaps, the only blot in the escutcheon of our native species,—in all other things the Batrachia may be said to “set their clocks” by that of Rana temporaria. The mandible (fig. 3) has all its parts perfect or typical. The columella is much more developed than, but is extremely similar to, the last ; it has the distinct distal segmentation (Plate 5, fig. 5), but the distal piece is now contracted, like a “spatula,” and has sent from its inner face a supra-stapedial band of cartilage upwards to coalesce with the under surface of the tegmen tympani (¢.st., s.s¢.). Also the large solid mass of cartilage at the stapedial end of the bony medio- stapedial (m.st.) has become a separate “inter-stapedial” (¢.st.), articulating with the stapes (sf.). The “annulus” (fig. 1, a.ty.) is a perfect ring of cartilage as in other North American kinds, and as in many of those from the Oriental region. The “hyo-branchial” apparatus (fig. 4) is now quite normal; all the parts are fused together, are quite chondrified ; and the “thyro-hyals” are now (typically) ossified ; the hyoid bar (st.h.) is distinct from the ear-capsule. The nasals and premaxillaries, and their contiguous labials (n., px., u..u.P.) are quite normal, so also are the maxillaries (ma.); but I can find no “septo-maxillaries”—small, variable, and inconstant bones; the quadrato-jugals (q,j.), as I have just mentioned, have united with the quadrate cartilage ; and the squamosals (fig. 1) are truly elegant Batrachian bones, with a twisted sub-falcate upper, and a flat sigmoid lower, part. The “vomers” (Plate 5, fig. 2, v.) have their characteristic spikes, fore, middle, and hinder; the two latter enclosing the inner nostril (e.7., by mistake) largely; and a postero-internal, thick, dentigerous lobe. The parasphenoid (fig. 2, pa.s’., pa.s.) like that of Rana pipiens (Plates 3 and 8) shows the very rare condition of a distinct bony centre for the point of this little dagger. These are the conditions of the growing skulls in this genus; afterwards I shall show that many of the dwarfed and more or less generalised types are arrested, in some things, at certain stages that correspond with growing stages of the typical kinds. 40 MR. W. K. PARKER ON THE STRUCTURE AND Skulls of adult individuals of the genus Rana. a. European species. 6. Rana temporaria. Ihave just referred to the characters of this (typical) species, described and figured in my first paper (Phil. Trans., 1871, Plates 3-10, pp. 137-211). I shall use it as a norma, after eliminating its one aberrant character, viz.: its sub-distinct “meta- pterygoid.”* 7. Rana esculenta. For a description of the skull of this species the reader is referred to Professor Hux ey’s article on the ‘ Amphibia” (Encye. Brit., vol. ix., pp. 750-771).+ b. Oriental species. 8. Rana gracilis—Adult male ; 14 inch long. Ceylon. The skull in this small species (Plate 6, figs. 6-10) differs but little from that of the Common Frog (op. cit., Plate 9); its endocranium is more ossified, but its mandibular hinge does not reach so far back: the general form is very similar, being half a rather long ellipse. Between the oval occipital condyles there is a large right-angled space (figs. 6, 7), and the outline of the foramen magnum (/f.7.) above is a large crescentic emargination. The superoccipital soft tract is twice as wide as the basioccipital, but they are both of small extent. The whole occipito-otic mass is a large transversely-placed oval, the ends of which are nearly ossified above, but with a large tract of cartilage below, which runs across from side to side between the prootics and ex-occipitals (pr.o., e.0.). The passages for the main nerves (II., V., [X., X.) are large and very clearly seen in the lower view (fig. 7). The optic nerve passes through a large membranous fenestra, which has a narrow tract soft behind it, and a broad tract in front; half the inter- orbital region is unossified, but the foramen ovale (V.) lies in the centre of the large postorbital face of the prootic. Above, from beyond the horizontal canal to the edge of the tegmen tympani, there is a soft tract; but from thence inwards to the lesser fontanelles all is bony, and the prootics and ex-occipitals are quite blended, walling in the canals and covering most of the hind archway of the skull. The fore half of the interorbital region is occupied by the common “sphenethmoid,” or girdle-bone (eth.), which does not take up all the ethmoidal region ; it projects in front above, but is emarginate below. The fore part of the ethmoid, all the nasal region, and the transverse ethmo-palatine bars * Tn my figures of the nasal region of this species (‘‘ Batrachia,” Part 2, Plate 54, figs. 1, 2) the 2nd upper labial is figured too high up, and the pro-rhinals are lettered c.tr. + This species has the pro-rhinals distinct, as Professor Huxtry’s figure (p. 755, fig. 9) correctly shows. This distinctness is seen again in Dactylethra (Phil. Trans., 1876, Plate 59, fig. 6, w.l°.); it is there lettered w.l°. by mistake. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 4] are unossified ; the floor is wide in front and narrower behind, whilst the contrary is seen in the roof. The subnasal angles are well developed; the pro-rhinals (p.rh.) are large, elegantly pedate, and turned outwards and downwards: there is a distinct prenasal rostrum more than half their length. The cake-shaped inner, and the shell- like outer, upper labials (w.l’.w.2.) are well developed. The large palato-suspensorial arch is but little affected by the bones investing it,—the palatal and the pterygoid (pa., pg.); but these are normal and well-developed ; there is no metapterygoid, and the part above the quadrate hinge is largely ossified by the quadrato-jugal (¢., q.J-)- The hinge reaches to opposite the stapes ; its stem is strongly clamped on the inside by the pterygoid, and on the outside by a well-developed squamosal (sq.), whose supra- temporal and postorbital regions are rather larger than in R. temporaria. Over its descending part is the “annulus” (a.ty.), which is also larger than in the type, and it is also completed into a ring above. The Eustachian opening (ew.) is large, and the stylo-hyal end of the hyoid bar (st.h.) is confluent with the lower part of the ear- capsule behind this passage. The stapes and columella are large and well developed, but the solid inter-stapedial mass of cartilage is not segmented off from the medio-stapedial bar (fig. 10, i.st.,m.St., St.) 5 the extra-stapedial (e.st.) is not spatulate, but orbicular, and the strong supra-stapedial (s.st.) is confluent with the auditory roof. The mandible (fig. 8) has the dentary broad and ascending behind the mento-Meckelian ; the hyo-branchial plate (fig. 9, h.hy.) has a broader hypo-hyal lobe than in the Common Frog. The fronto-parietals (fp.) are rounder and thinner, the parasphenoid relatively larger, but the preemaxillaries, maxillaries, and nasals agree very closely with those of the type; the quadrato-jugal, however, differs : it is largely grafted on the quadrate cartilage (figs. 6, 7, fip., ., px., Mx., Pa.s., YJ. J-)- This kind differs from 2. temporaria in a few points, viz. :-— . It has a prenasal rostrum. No septo-maxillaries. The supra-stapedial is confluent with the “tegmen.” . The inter-stapedial is not distinct. . The stylo-hyal is confluent with the ear-capsule, . It has a crested dentary. . Quadrate partly ossified. 9. Rana cyanophlyctis.—Male ; 13 inch long.* Ceylon. This is another smaller kind of Indian Frog ; it is, according to Dr. GUNTHER (ibid., * Dr. Giinrner (“Reptiles of British India”) gives the following measurements (from snout to vent) of the largest specimens known to him of the Indian species of Rana here to be described :—a. FR. Kuhli, 4) inches; b. R. hewadactyla, 5} inches; ¢. R. eyanophlyctis, 14 to 2} inches; d. R. tigrina, 6 to 7 inches; e. R. gracilis, 13 inch. R. pygmea (see Gintuer, P. Z. 8., 1875, p. 568) measures (adult female with ripe “ ova”) only_25 millims. long, or one inch. MDCCCLXXXI. G 42 MR. W. K. PARKER ON THE STRUCTURE AND p. 406), much like the large R. hevadactyla, but always remains small: its skull (Plate 10, figs. 7-10) does not suggest a specific identity with that type, of which it might be mistaken to be the young. A young individual of the large kind of this size would have a very similar skull, yet there is no reason to suppose that the two are identical ; nevertheless, this is an arrested form. Besides the external characters by which this kind differs from the last—R. gracilis— the skull itself shows a difference of character, although on the whole the variation seen is very much what would be found between an adult and a young of this somewhat larger species. The occipito-otic regions (Plate 10, figs. 7, 8) are here less solid than in R. gracilis and much less ossified, for much of the horizontal, and nearly all the posterior canals (fig. 7, pr.o., e.0.) are uncovered by bone, so that the four “centres” are distinct, even above. The cranial “barge” does not widen out so much towards the postorbital region, and the girdle-bone (eth.) is of less extent. In front there is a “prenasal rostrum,” but it is shorter, and the pro-rhinals are shorter and broader ; among the outworks of the nose, viz.: the upper labials (w./'.w.P*.) we find a small sigmoid “septo-maxillary” (s.m.a.). The other more constant, investing bones are very similar, but the parietal region is narrower, and the parasphenoid (pa.s.) is less elegant. The squamosal (sq.) covers very little of the “tegmen tympani ;” the “annulus” (a,ty.) is even larger, and a perfect ring; the quadrato-jugal scarcely affects the quadrate at all, and the hyoid cornu (c.hy.) keeps free, above. The stapes (fig. 10, st.) is less, and the “columella” (Plate 6, fig. 12, and Plate 10, fig. 10) 1s very inferior in development to that of 2. gracilis, yet its inter-stapedial segment (i.s¢.) can be seen to be distinct from the medio-stapedial (m.st.). But the form of the main rod is more simple and undeveloped ; the extra-stapedial (e.st.) is a mere rounded, soft end of the rod, and there is no “ supra-stapedial” band. The mandible is quite normal, but shows the crest on the dentary near the chin (Plate 6, fig. 11, d.). The hyo-branchial apparatus (Plate 10, fig. 9) is also quite typical, it shows, how- ever, some difference in this, that the lobes which grow out from the cerato-hyal and the base, are smaller. There are some curious points of difference between the skull of this little Oriental Frog and that of a young Rana temporaria of the same size; the following are noticeable :— 1. A definite prenasal rostrum. Smaller septo-maxillaries. A more arrested squamosal. A crested “dentary.” 5. An arrested columella, without an ascending part, and with the distal part a ae ese mere bud. Minute, but measurable, differences are to be found between species and species, DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 43 everywhere in this outwardly similar, but inwardly most variable, “Order.” The determination of these will, I am satisfied, help the Zoologist in his work of deter- mining species and classifying the groups. 10. Rana pygmea—Adult male; 2 inch long. Anamallays Mountains, Malabar, 8. W. India. IT shall, in turn, describe the skull of a true Frog as small as this from the Ethiopian, and of a sub-typical form from the Neotropical region; they will appear like the young of larger species; this is manifestly old and its skull is intensely ossified (Plate 5, figs. 11-15). This is a very short skull; in R. gracilis and R. cyanophlyctis the width of the skull across the quadrato-jugals is exactly the same as the length. In this species the breadth is to the length as 4 to 31. In R. cyanophlyctis the condyle of the quadrate ends opposite the exit of the 9th nerve ; in /2. gracilis it ends opposite the same point ; in R. pygmea it does not reach further back than the front third of the stapes—two-fifths of the distance between the 5th and 10th nerves, at their exit. The extremest case of wide gape is in R. pipiens, where this joint is far behind the occipital condyles (see Plate 8). This is almost the only sign of avrest in the skull of this small, stout species,* but there are some things in it which have a look of generalisation about them, as if this were an old offshoot from the stock. The occipital condyles (0c.c.) are oval, small, and very wide apart ; the emargination of the roof, over the foramen magnum (fm.) is slight; that of the floor is a mere gentle arcuation of the transverse margin of the skull-floor. The whole cranium, proper, is flat and wide, only gently narrowing in at the temples, and a little more in front of the eyes; thus the ear-masses are wide apart, and with their drums outside and a little in front of them, have a very elegant appearance (Plate 5, fig. 11). These four regions are nearly of the same size. The common bony occipito-auditory mass is cleft in its fore margin below (fig. 12), and bi-emarginate in its fore margin above (fig. 11). This widely extended post- cranial region is capped with cartilage at three places on each side. These unossified tracts are—a, the occipital condyles (oc.c.); 6, the edge of the tegmen tympani (fig. 11, t.ty.); and c, the facets for the “pedicle” below (fig. 12, pd.). The roof- cartilage has evidently been absorbed considerably, between the temporal regions, thus the three fontanelles all run into one, which is an oval, with rounded lobes behind. The alisphenoidal region is ossified half way to the optic opening (II.) ; the girdle- bone (eth.) just reaches that passage in front; thus there is a half belt of cartilage in * T am of opinion that there are forms which, although highly specialised, are arrested as compared to their large nearest congeners; this must be due to dwarfing, and is to be distinguished from archare, or truly generalised conditions. G » 44 MR. W. K. PARKER ON THE STRUCTURE AND the orbito-sphenoidal region, and this grows backwards as a sharp wedge, between the right and left bony tracts below (fig. 12). The “girdle-bone” runs well into the true fore edge of the ethmoidal region, up to the autogenous nasal roof, and outwards into the “ethmoidal wings,” where they pass into the palatal pedicle of the face (e.pa.). There is no “ prenasal rostrum” formed by the intertrabecula in front, but the pro-rhinals (p.r/.) are good-sized hooks, and the upper and lower laminz of cartilage are broad and well-developed. The out- works of the nose, the two pairs of upper labials (w./!.u./.), are quite normal. The outer less, and the inner more, bent arches of the face are very beautifully formed ; they are strong, but not thick and solid. The palatine (pa.) and the pterygoid (pg.) have affected the cartilaginous bar very little; the former is falcate, with an unusual curve; the latter is short, both in its process for the ‘‘ pedicle” and that for the quadrate region. The bones that form so perfect a semi-oval outline are quite normal also (pza., mzx.,q.j.); the hindermost or quadrato-jugal has largely infected the quadrate with bony matter. The Eustachian tube (ew.) in the round notch of the pterygoid (pg.) is a short oval with its greatest diameter parallel with the axis of the skull; it is generally oblique in position (see Plates 6 and 7). Over the semi-osseous suspensorium the squamosal (sg.) binds by a very sigmoid, broadening stem : above, it is small on the tegmen tympani, but has a long, out-turned, postorbital snag. It is enclothed with a large and very broad “annulus,” which, however, differs from the Oriental Rane generally, but agrees with the English kind, by being imperfect above ; its crura are unusually wide apart. The stapes (Plate 5, figs. 12 and 15 st.) is large, oblique, sub-uncinate behind, and has a sinuous front margin. It has the intercalary inter-stapedial (7.st.) very dis- tinct, large, and ovoidal ; the medio-stapedial bone (m.st.) is pistol-shaped, with a very large handle ; the extra-stapedial (e.st.) is small and orbicular, and its supra-stapedial process (s.st.) is free above, as in /2. temporaria, but it is here a short rounded bud. Still there is in this kind a very highly developed “ middle ear,” and it is not in any sense a low kind of Frog. The nasals (n.) and the fronto-parietals ( fp.) are well-developed, but not very thick ; the parasphenoid (paz.s.) is, like the skull it forms a floor to, broad and strong; it is trifurcate in front, sinuous laterally, and notched twice on each side behind: has a triangular “handle” at its end, and a wide rounded “guard”; it nearly reaches the foramen magnum, behind, but in front it reaches only half-way along the girdle-bone. The vomers (v.) are quite normal, but very wide apart. The mandible (fig. 13) is perfectly normal, and does not show any dentary apophysis. The hyo-branchial plate (fig. 14) is like that of the two last species, and differs but little from that of the common kind; the dorsal end of the hyoid band does not coalesce with the auditory floor ‘behind the Eustachian tube (fig. 12, ew., st.h.). As compared with that of 2. temporaria, this skull differs very little in essentials. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 45 Leaving out of question mere size, form, and degree of ossification, we have only the following things really worthy of notice, viz. :-— 5 by? 5) 1. There is no division of the upper ‘‘ fontanelle.” PP . There are no ‘‘septo-maxillaries.” . The “quadrato-jugal” largely ossifies the lower part of the suspensorium. . The “supra-stapedial” is very short. y . The horns of the “annulus” are wide apart. of CO LP 11. Rana tigrina.—Adult female ; 54 inches long. Ceylon. This was a large specimen, but not equal to some examined by Dr. Ginrner.* In coming next to this kind I have passed from the smallest to the largest of the Frogs of India. As in the last, I have to be careful to distinguish those cranial characters that are due to mere size from those that lie deeper, and are of more impor\ance. Of this I am certain, viz. : that although there is a special density and strength in the bony elements of the larger skulls, yet that is not necessarily connected with an extensive and generalised ossification, which, in some cases, shows most in the dwarfed kinds. The last species is an instance at hand, for its endocranial landmarks are much more obliterated than in this gigantic species. In the largest specimens the skull cannot come far short in size of that of Rana pipiens ; it is a stronger structure, and the cranium proper is much larger in proportion to its facial outworks (compare Plate 6, figs. 1-8, with Plate 8, figs. 1-4). The two differ, however, in a much more important point, for this is a severely typical kind; the American Bull-frog, on the other hand, is very archaic or generalised. Although I have taken the medium-sized Common Frog as the best typical form, I do, nevertheless, consider that in Rana tigrina and in the helmeted Frogs (e.g., Ceratophrys and Calyptocephalus) we come across the most perfect examples of Batrachian cranial architecture ; moreover, they are not deficient in respect of morpho- logical characters that are deeper than, and almost independent of, mere size and bulk. The general outline of this skull (Plate 6, figs. 1-3) is more than half of an oval, rapidly narrowing towards either end, as the outline of the Hen’s egg does towards one end. The whole form is broad behind, but narrowing rather rapidly forwards ; the length of the skull itself as compared to its greatest breadth is as 73} to 9, but if the measurement be made from one quadrate condyle to the other, as 8 to 9, for these hinges lie some way behind the occipital hinges. The ovoidal occipital condyles (oc.c.) are well seen in both the upper and lower aspects, for they ave large and turned over the edge of the occipital floor, so as to fit deeply into the atlantal concavities—a thing answering to the great strength of this Frog. These condyles are separated by an emarginate tract less than their own width ; * He gives 6 to 7 inches, mine was therefore only three-fourths the size of the largest. 46 MR. W. K. PARKER ON THE STRUCTURE AND this is greatly in contrast with what is seen in 2. pygmea, for it has, like many other small Batrachians, very little condyles, very wide apart. The lower face of the occipito-auditory mass runs across at exactly a right angle to the axis of the skull, but when its whole extent is seen from above its outer angles are seen to be bent forwards.* This lower face, from one “ stylo-hyal” to the other, is a broad tract, one-third longer and one-third broader than the straight and almost evenly oblong interorbital region. ‘The prootics and ex-occipitals (pr.o., e.o.) of the same side are confluent, but there is a distinct synchondrosis both above and below (figs. 1, 2). A slight edge of cartilage remains to the floor of the tympanum, and a larger tract to its “tegmen” (fig. 3); nevertheless the ossification is intense, and reaches in front within a short distance of the optic foramen (II.). Also the girdle-bone is more than twice as extensive as in the medium type; it abuts upon the optic passage hehind, and besides the whole ethmoid, with its outer wings, runs half way along the proper nasal roof, floor, and middle wall (figs. 1, 2, and 3). A small superorbital projection can be seen over the inner canthus; these, and the wings of the ethmoid, up to the point where segmentation takes place in Bufo, are ossified., As to the fore and middle parts of the endocranium, we see some curious results of overgrowth. First, the cranial “ barge” appears to be very small and contracted as compared with the huge arches of the facial outworks; and, in the second place, the nasal labyrinth is greatly exposed by the retirement from it of the facial arches, in their wide sweep outwards. In the smailer kinds (see Plate 6, figs. 6 and 7) the nasal labyrinth is carefully packed between the laminz of the premaxillaries and maxillaries. In the upper view (fig. 1) we see the nose lying naked, for the most part, between the bony boundaries ; the outer angles of the floor (c.t7.) lying on the palatine lamina of the maxillaries, but not hidden by the upper plate. The ascending part of the outer angle of the floor (fig. 3, ”.w.) is considerably ossified ; it reaches the roof (a/.n.) above, and coalesces with it. There is a “prenasal rostrum” (in front of sv.) of moderate size, and large subretral pro-rhinals (p.rh.); the labials (w./., ul.) also are well developed round the circular aperture (e.7.). The cartilaginous palato-quadrate arch, on each side, is almost eaten away by the ectosteal palatine and pterygoid (pa., pg.), the one thick and spatulate, the other a strong triradiate bone.t * In the lower view (fig. 2) I have only figured the parts that come immediately under notice; the distant upper parts are left ont, as they are not necessarily seen by the eye when discriminating the lower surface. The upper view must be compared with this to get a complete idea of the form. + In this and in a large number of Batrachia these two bones are as truly ectosteal as the sheaths of the columella or of the thyro-hyal, and become truly homologous with the perichondrial lamin of the prootic exoccipitals and sphenethmoid. I therefore colour them, throughout, as endoskeletal bones. In the ‘« Shoulder. girdle” we see the same temporary distinctness of the perichondrial, from the endosteal tracts of bone, in the “ pro-coracoid”’ and ‘‘ supra-scapula.”” (See my “ Shoulder-girdle and Sternum,” Plates 5-7.) DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 47 Above (fig. 1), and below (fig. 2), we can trace here and there some remains of the cartilage ; above, in the ethmo-palatine, the pre-palatine, and the post-palatine, on the supero-external edge of the pterygoid ; below, on the end of the “ pedicle,” its articular facet; and above and below the graft of the quadrato-jugal; the latter being the coating of the large bilobate hinge (q.c.). These inner works of the skull, so largely converted into solid bone, are built over by very strong membrane-bones, which give the finish and the beauty to this style of cranial architecture. Below, in relation to the skull proper, the parasphenoid (fig. 2, pa.s.) is of the usual relative size, but looks small as compared with the large bones of the face ; it is typical in form, but unusually strong. The vomers (fig. 2, v.) here attain a very large relative and real size, for they meet at the mid-line (compare them with those of R. gracilis, fig. 7, v.) ; their snags are large and strong, and the oblique dentigerous lobe of each is of considerable extent. Above, the fronto-parietals (fig. 1, fp.) form a strong roof with a notch in front, the remains of the frontal suture, but are wholly coalesced beyond this ; they end, behind, in two broad wings which spread over the hinder region of the cranium, almost to the end. At first, hollow in the middle, in the postorbital region they develop a sagittal crest, which opens out into two temporal wings. The temporal part dips into the orbit (fig. 3, fip.) and then rises over the ear-masses, moulding itself on to their sinuosities. The sides are notched, and the end has a concave margin. The nasals in front (figs. 1 and 3, 7.) form cross bars which are nearly as straight as the wings of the parasphenoid ; they meet along the middle, run their pointed ends far forward, are triangular bebind, where they overlap the fronto-parietals, and with them cover the girdle-bone—all but the superorbital projections and the middle part in front of the fontanelle. The fontanelles are presumably like those of the lesser kinds, but they are quite covered over. The investing bones, outside, differ from the type in relative size and density; other- wise they are quite normal: there is no septo-maxillary, the palatal ingrowths are well developed, the nasal process of the premaxillaries and of the mailluries (px., mx.) is strong; the quadrato-jugal (q,j.) is grafted on to the quadrate, and the three regions of the squamosal (sqg.) are extremely well developed and give the highest idea of the Batrachian form of these bones. So, also, the various condyles—of the occiput, the pedicle, and the mandible—are well and typically formed. The mandible (fig. 3) is crested both in its dentary and coronoid regions. The stapes and its additions are of an average size (figs. 3 and 5); the stapes (st.) has its fore margin oblique; jammed in between it and the fenestral edge of the capsule there is the large seed-like inter-stapedial (7.st.) ; then, distinct from it, the pistol-shaped medio-stapedial (m.st.), whose unossified part passes into the broad spatula of the extra-stapedial (e.st.), which sends upwards its supra-stapedial (s.st.) ; this is ligulate, and confluent above. The stylo-hyal end of the hyoid tape is 48 MR. W. K. PARKER ON THE STRUCTURE AND confluent, but the main part of it has been absorbed (fig. 4), otherwise, the basal plate is typical. The “annulus-tympanicus” is large and a complete ring (fig. 1 a.ty.), the Eustachian openings (ev.) are large, oval, and oblique. The greater part of the difference between this skull and that of the Common Frog depends upon the intensity of ossification and the large size of the investing bones— all correlated to the great size of this type. Other more important differences there are, Viz. :— 1. A definite prenasal rostrum. . No “septo-maxillaries ;” the nasal cartilage ossified in that region. . A crested dentary. Much of the cerato-hyals absorbed. Stylo-hyal end of these bars confluent, above. Supra-stapedial confluent, above. Quadrate partly ossified. . Rudiment of superorbital. ON Ao RP wD 12. Rana hexadactyla.—Adult female ; 5} inches long. Ceylon. This specimen was a little less than that of R. tigrina, and the skull (Plate 7, figs. 1-5) is slightly longer, but its greatest breadth is only as 8 to 83; the quadrate condyles only project one millimetre, or about half a “line” beyond the occipital condyles ; in &. tigrina they project twice as far backward. In outline, besides its narrower form, the nasal end is broadly truncate, and these two modifications cause a third, viz. : the unusual straightness of the sides of the skull. It is altogether a feebler skull, with less intense ossification, as well as being straighter and narrower; moreover, it is not quite symmetrical ; yet its facial plates (fig. 3, ma., sg.) are deep and well developed. But these are mere superficial differences between the two; there are others that lie deeper down. Gentle enough are these morphological variations, and no bar to the supposition that the two species originally sprung from one common stock, yet they are not to be lightly passed over, for they are full of interest. The occipital condyles (oc.c.) are rather larger and more under the skull than in the last ; also the emargination between them is deeper, these are due to some differences in the working of this hinge. The whole occipito-auditory mass is less oblong, its front faces being oblique as they pass into the interorbital region, exposing the foramina ovalia (V.), and the occipital condyles and epiotic eminences (ep.) over the posterior canals project more backwards : moreover, the bones of the two sides are confluent over the foramen magnum (f-m ). Yet there is more cartilage between the 2nd and 5th nerves (II., V.), and also a more DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 49 evident “ pterotic” eminence projecting behind the squamosal, at the end of the tegmen tympani (fig. 1, sq.). The form of the canals within is less obscured by the bony growths and coverings than in the last, so that the general surface of this part is more uneven. The girdle-bone (eth.) reaches from the optic fenestra to the middle of the septum nasi (s.n.); it is therefore less in anterior extent, but it is visible between the vomers (fig. 2, v.), in front of the descending bar of the nasal (figs. 1 and 3, .), and also behind that bar, ossifying the wing of the ethmoid up to the ethmo-palatine suspensorium. The whole interorbital part of the cranium is narrower and more pinched in the middle, and the relative narrowness of this trough approximates to what I shall show in my next instance but one, viz.: in Rana pipiens (Plate 8). Here, again, a small superorbital projection (s.ob.) is seen outside the meeting of the nasals and frontals, as in the last. In front, the prenasal is a mere bud, little more pronounced than in FR. temporaria; the pro-rhinals (p.rh.) are much like those of the last kind. The outer angle of the subnasal lamina (fig. 1) is more extended and quite exposed, and this (trabecular) plate also projects outwards further back, where the girdle-bone ends; it joins the roof above by an ascending plate (v.w.), which is ossified for some extent at its root. The labials are very similar in both species; the upper is hidden behind the premaxillary. 7 In R. tigrina (Plate 6, fig. 2, pa.) the palatines turn forwards and outwards ; in this kind they run straight across; they also come nearer together ; but on the whole these bones and the pterygoids (pg.) are very similar in both species : there is, however, more cartilage left unossified in this. The gliding joints of the pedicles (fig. 2, pd.) are nearer together in this narrower skull, and the Eustachian opening (eu.) is thrown obliquely backwards instead of forwards, outside, by the straighter hind process of the pterygoid. The hinge of the quadrate (q.c.) is less deep; its substance is more ossified by the quadrato-jugal (¢,7.). Altogether the parasphenoid (pa.s.) is slenderer and more elegant; it is equally sub- carinate, and has the blade and handle longer in proportion to the guard than in R. tigrina: here we see the most perfect form of this bone in the Batrachia, the basi- temporal processes being slender at first and then dilating outwards. The vomers (Plate 7, fig. 2, v.) are not so large as in the last, do not come so close together, and their inner edge is sinuous, not arcuate. The fronto-parietals (fig. 1, fp.) are less dense, are distinct in their fore half, and the temporal fossee are bounded, above, each by its own parietal ridge; the hinder spreading part is altogether less. The nasals (n.) are larger, broader before and behind, and their facial process (fig. 3) forms a more perfect suture with the maxillary. MDCCCLXXX1. H 50 MR. W. K. PARKER ON THE STRUCTURE AND The nasal processes of the premaxillaries (fig. 1, pz.) are less, but the palatal (fig. 2) are larger, and the right bone is much the larger of the two where it joins the palatal part of the maxillary; there is a small septo-maxillary also on the right side (fig. 1, s.mz.). The maxillaries and quadrato-jugals (mz., q.j.) are much alike in both species; the squamosal (sq.) ends abruptly over the auditory mass, and not as a triangular process (see also Plate 6, fig. 1); its postorbital process is deeper; the mandible (fig. 3) has a crest to the dentary, and the coronoid process of the articulare is high. The “annulus” (a.ty.) is large and complete. The parts of the “middle ear” are similar to those of the last, except that the inter-stapedial (fig. 5, 7.st.) is larger and is well ossified, the medio-stapedial (m.st.) straighter, and the extra-stapedial (e.st.) broad and orbicular ; the supra-stapedial (s.st.) is coalesced above. The stylo-hyal ends of the hyoid bands are confluent above ; the whole bar is perfect and normal on the left side, but on the right it is absorbed from the hypo-hyal nearly to the stylo-hyal regions (Plate 7, fig. 4, c.hy., h.hy.). The differences to be seen between this and the last are largely due to diminished size and strength; these variations, as compared to what is seen in the lesser types, are largely due to their greater bulk. But there is a residuum of variations that cannot fairly be put down to these causes. There are in this species :— 1. A small but definite prenasal rostrum. 2. A small septo-maxillary on the right side only, but the nasal angle in its ascent has a solid bony mass formed in it. 3. A crested dentary. 4, The right cerato-hyal absorbed. . Stylo-hyal confluent, above. . Supra-stapedials confluent, above. . Inter-stapedial well ossified. . Quadrate partly ossified. . Rudiment of a superorbital plate. oonN GO 13. Rana Kuhli.—Male ; 23 inches long; two-thirds grown. Ceylon. This large, but not full grown, Frog belongs to the most aberrant species of its genus, and its peculiarities are the stronger expression of what is more feebly seen to particularise the other Oriental kinds of Rana. We shall see in this type how near the characters of a species of this genus may come to those of types which lie at the furthest distance from the model-form ;— borderers and mixed breeds, so to speak, that help to break the isolation of the main group, and to show its affinities to the groups that encompass it round about. As to general form, this skull (Plate 7, figs. 6-10) is half a moderately long ellipse, and has very neat outline. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 51 Its regions, fore, middle, and hinder, are in contrast with what is seen in the two Bull-frogs just described and also in the lesser species, for the middle region is long, relatively, beyond that of any kind I know, each orbital region being large enough to ensocket a pair of eye-balls twenty times the size of those that do lodge in them. This middle region shows a cranial trough, not high (fig. 8) but narrow, and with an approach to the outline of an hour-glass, The length is a fraction less than the greatest breadth, and the quadrate condyles (q-c.) just reach to a supposed line running across the double hole for the 9th and 10th nerves (IX., X.). Thus the gape is not that of a very characteristic Frog, but rather that of an arrested or a somewhat generalised type. Laterally seen (fig. 8), the skull is arched, having strong deep curved planks of bone built round and over it: this is a correlate of its powerful mandible with its mimetic canine tooth (d.). Everything in the side view speaks of strong pterygoid and temporal muscles. The outer bones are unusually strong, so are the intermediate pterygoids (pg.), but the ossification of the endocranium is exactly like that of an adult Common Frog, and very inferior in degree to what is seen in the two large kinds, and in the dwarft species (R. pygmea, Plate 5, figs. 11, 12). The occipital condyles (0c.c.) are large, low in position, near together, and with the short interspace straight. There is a considerable basi- and supraoccipital tract unossified, and the prootics and ex-occipitals (pr.o., ¢.0.) are divided below, and only slightly confluent above. There are the three normal fontanelles above, and the side walls (fig. 8) are well built up to the roof, under the edges of which there is on each side a very definite ‘ wall- plate.” The arched form of the skull is combined with considerable overlapping of the roof and floor (figs. 7, 8, f.p., pa.s.), the investing bones being applied above and below very closely round the endocranium (eth., 0.s.) so as to leave the interorbital wall uncovered to an unusually small extent. The short and not very broad nasal region lies well within the outer bones; there is no “‘rostrum,” but the pro-rhinals (p.vh.) are well developed and unusually long and projecting ; the labials (w./’.u/.) are normal. The “oirdle-bone” (eth.) reaches only to the front of its own region and half-way back to the optic nerves; it does not ossify the very definite angular supraorbital projection (s.ob.), which here attains to a distinctness almost equal to what it has in the “ Hylidze.” The more rounded form of the great orbital space in front, enclosed there by the ethmo-palatine bar, gives rise to a sickie-shaped palatine bone (pa.), and this is followed at the outer edge by a very remarkable pterygoid (pg.). The processes of this bone that enwrap the pedicle (pd.), and bind upon the inside of the suspensorium to the hinge (q.c.), are short but normal, but the imtra-jugal portion of the bone is of great depth (fig. 8, pg.), and strongly inbent. In the “axil” of the pterygoid the Eustachian passage is small and round; the stylo-hyal end of the hyoid (st..) has coalesced with the tympanic floor, and bends H 2 52 MR. W. K. PARKER ON THE STRUCTURE AND strongly back to keep to the margin of this passage as the hinder half of its natural skeleton. The quadrate region of the suspensorium (q.) is not ossified by the “ quadrato-jugal ” (7.j.)—shaped like a Serpent’s tooth ; the saddle-shaped condyle (q.c.) is very large, but does not reach to the extreme angle of the suspensorium (figs. 7 and 8). The “annulus,” like the Eustachian passage, is small, the band itself is wide (fig. 6, a.ty.), and it is a perfect ring; the whole tympano-Eustachian cavity is small, but the columella and stapes (fig. 10) are of the average size; the latter (s¢.) is almost a long triangle, but the posterior and inferior edges are rounded: the anterior margin is con- cave, fitting into the inter-stapedial. This latter part (7.st.) has a deep saddle-shaped condyle for the stapes ; it is unossified. The medio-stapedial (nz.st.) is nearly all ossified ; it is gently arched, and not much expanded where it articulates with the inter-stapedial. The extra-stapedial (e.st.) is very large, elegantly heart-shaped, peltate in arrange- ment—its cartilaginous “handle” passing from the medio-stapedial bone into the inner face of the main plate obliquely ; and it covers half the concave, wide “ annulus.” There is no supra-stapedial band—a part which is seen even in the peltate extra- stapedial of Bufo vulgaris (Phil. Trans., 1876, Plate 54, figs. 7 and 8); but here we have, suddenly, as it were, the columella of Pipa and Dactylethra over again, with the difference of a short and soft inter-stapedial (Phil. Trans., 1876, Plates 59 and 62). The extra-stapedial part is mostly orbicular in the Oriental Rane, but not large. The mandible (fig. 8) is normal, with the exception of the fore part of the dentary (d.); the crest which grows up from that bone in front in other species is here a high strong spur—an imitation, in bone, of a large canine tooth: in old age (according to Dr. GtyrHer) it becomes capped with tooth-substance. It “cuts the gum” early, and soon dents the side of the palate, which becomes hollow to receive it. Moreover, the bones all round this excavated part are greatly modified; this is partially seen on one side in R. hexadactyla (fig. 2), with its much smaller dentary crest.* Where the palatine plates of the premaxillaries and maxillaries meet (fig. 7, pa., mz.), their processes project far inwards; this is to leave room for the socket of the quasi- canine ; the hole is bordered on the outside by the dentary edges of these bones at their junction. Not only so. Where they meet on the outside (fig. 8), there they rise high at their junction, and leave an angular space or gap in the toothed margin of the fore face. There are no septo-maxillaries; both the quadrato-jugal (q,j.) finishing the cheek, and the squamosal (sq.), in the temporal region, are very strong and elegant bones, but the latter is very remarkable, even for a Frog. The bone seen from above (fig. 6, sq.) is a sickle, its rough notched handle lying on the tegmen tympani; running obliquely * Dr. Giinruer looks upon this quasi-canine as a special thing—a tendency, so to speak, to produce a tooth, here; my own mind leans to the opinion that it is not a rudiment, but a remnant; nearly all the Indian kinds show it more or less, and their Common Parent may haye had large genuine teeth in the front. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 55 into it, we see the postorbital process which is the curved blade; together, they are half as long as the whole skull (see fig. 8). Also the angle at which the stem runs down over the massive quadrate and its condyle is very remarkable (compare figs. 3 and 8, sg.), for the axis of the postorbital process is almost coincident with it, and between the stem and the supra-temporal plate the space is only half a right angle. The roof and floor have been wrought over the inner skull, above and below, so as to half hide it; thus the parasphenoid and the fronto-parietals are bony troughs or hollow splints. The latter (/:p.) are quite distinct, right and left, very long, narrow, narrowest in the middle, not very wide in the temporal region, where their edge is emarginate, curving with the canal below; whilst in front they are overlapped by the nasals (v.) and hide the girdle-bone: outside they do not hide the superorbital cartilages (s.ob). The nasals form a pair of large wings and are highly arched and strong (fig. 8); the outer part is an elegant facial hook, binding on the upper edge of the maxillary ; the two bones meet all along, forming a nasal suture. The parasphenoid (figs. 7 and 8, pa.s) is a very long trough of bone of the usual shape, with well-formed basi-temporal wings that are trilobate. The hyo-branchial plate (fig. 9) is quite normal in form ; there is a slight fissure in the substance of the “ hypo-hyal” lobe (h.hy.) ; a division of the hyoid cornua into an outer and an inner tape not uncommon in the Batrachia: the outer is a remnant of the pectinate inter-branchial cartilage such as we see in the Chimeroids and their kindred. Compared with the type, Rana Kuhli shows the following divergences of character :— 1. No septo-maxillaries. 2. Dentary with a very large tooth-shaped process. 3. Articulation of pre-maxillaries and maxillaries very wide and high, to admit of the process of lower jaw. 4. Stylo-hyal confluent, above. 5. Extra-stapedial a very large cordate disk, peltate, and without supra-stapedia. 6. Superorbital cartilaginous, and very distinct. 7. Eustachian openings very small. I have already mentioned the peculiar form of the skull, and of certain individual bones, but the above are the most important morphological variations. This skull is altogether very instructive and suggestive ; &. Kulli is certainly a “borderer,” and I suspect it has retained some very archaic characters that have been more or altogether obliterated in the other species of its genus and its territory. ce. North American (“ Nearctic”) species. 2 (continued).* Rana pipiens, Harn.—(R. catesbiana, Suaw ; R. mugiens, MERR.). This specimen was considerably larger than those of the Indian Bull-frogs just described. * See p. 22 where the larva of this species is described. 54 MR. W. K. PARKER ON THE STRUCTURE AND I have already shown the characters of two kinds from this “region” (2. palustris and FR. halecina), but they were young; yet some unexpected and very important points were elicited, and will be referred to in the description of this large and most characteristic kind. In this species we see those peculiarities of the Batrachian skull which set it by itself, making it to differ from other kinds of skulls, carried to an extreme degree of development. These things bespeak a highly specialised type, and indeed this American Bull-frog is a Frog of the Frogs ; his shape, dress, voice, and carriage, all combine to make him the representative of his group or “ Order.” On the other hand, like all giants, there is much in him that bespeaks an ancientness (as if he only remained of the remnant of the giants), and was somewhat out of place among the more proud and elegant dwellers in the marshes and miry places of this, the newest, epoch. In outline this skull (Plate 8) is half an ellipse and very regular; its greatest breadth is, as to its axial length, as 74 to 53. If the length were measured up to a line passing from one quadrate condyle to the other, then the skull would be one-tenth longer than when measured up to the convexity of the occipital condyles. In my specimen the former reach one-fifth of an inch further back than the latter. In R. tigrina (Plate 6, figs. 1, 2) this distance is only one-sixth of an inch; relatively, however, the distance is the same, for the Indian skull was smaller; in R. hexadactyla (Plate 7, figs. 1, 2) the distance is only half as much (one line). This extreme extension of the gape back wards, during metamorphosis, is in remarkable contrast with what I find in my oldest specimen of the adult of a small toothless Australian Batrachian, viz.: Pseudophryne Bibronii. In this kind, in which the female is one inch long, reaches very little more than half the relative distance attained to in f. pipiens and and the male three-quarters of an inch, I find that the quadrate condyle R. tigrina, or only two-thirds as far back as the occipito-atlantal hinge, measuring from the front of the snout, backwards. In that kind the whole suspensorium is arrested in its backward movement, when it forms a right angle with the axis of the skull: this is very similar to what is seen in the skulls of ‘‘ Caducibranchiate Urodeles,” whose gape is so much less than in the “ Anura.”* The next thing that strikes the eye, after the great extent of the gape, is the very small size of the cranial ‘‘ barge” as compared with the huge facial outworks ; its average width is a quarter of an inch, scarcely more than that of the Common Toad, with a head little more than half the length of this Bull-frog. The occipital condyles are large, near together; more shown below than above, and with a gentle emargination between them. The whole occipito-auditory region, right and left, is marked by great hills and hollows, and jutting snags, very unlike what is * Tt is a general rule in this ‘ Order”? that the gape is relatively, as well as really, larger, the larger the species becomes ; and in very dwarfed kinds the contracted gape is a correlate of several other arrests in the development of the parts of the skull. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 55 seen in the skull of small kinds, and bringing into prominence parts that also project largely in osseous Fishes. The epiotic region burrowed by the posterior canal (ep.) stands out, divergingly, on each side of the great archway, with its transversely oval entrance (fig. 4). The anterior canal sweeps round the front; outside it there is a great hollow, and then the periotic mass rises over the horizontal canal, this it does still more at the edge (tegmen tympani). The outer margin grows backwards into a large terete unossified “ pterotic” ridge (pt.o.)—the part that is ossified by the pterotic bone in Teleostei. The whole structure is, on the whole, bony up to the fore margin of the foramen ovale, and these right and left masses are confluent over the foramen magnum, but not below (fig. 2); there is a small tract of cartilage in the basioccipital region. Below, the twin post-aural nerve passages (fig. 2, IX., X.) are wide apart, and the antero-external margin of the ear-capsule is bevelled away and covered with a plate of cartilage for articulation with the pedicle (pd.). The ear-masses and intervening hind skull together make only a third of the whole width at this part; above (fig. 1) the pterotic crests (pt.o.) stand further out, where they pass beneath the squamosal (sq.). Suddenly in the temporal region the skull is compressed to two-thirds its average width, then becomes of a fuller form, and gently narrows again before it spreads out into the wings of the ethmoid, opposite the closing in of the cranial trough. The optic fenestra (II.) is moderate ; it is margined by cartilage in front, above, and below, and the girdle-bone (eth.) has not much more interorbital space than this tract, which lies in front obliquely over the bone. There is a small ossified superorbital lobe (s.ob.). The cruciform bony mass only reaches in front to the proper morphological edge of the ethmoidal territory ; all the true nasal region is unossified. The whole nasal roof is a winged sub-pentagonal tract, not ending in more than a bud of the “ prenasal,” and with half its septal part and a headland of its roof uncovered by the nasals (fig. 1, s.v., 7.7., ”.). The outer nostrils (figs. 1, 3, ¢.) are protected by the usual inner and outer upper labials (w./.w.7.). The roof runs forward, narrowing; the floor (fig. 2, s.n.) is sub-quadrate, and is finished antero-externally by the curious cervicorn angles (Plate 8, fig. 2, and Plate 9, fig. 7, s.n./.). The septum (s.7.) below is very thick in front, and narrows backwards towards the ethmoid bone; the angles of the subnasal lamina behind pass into the large, widely-extended ethmo-palatines which are curiously covered with splints, above and below. On each side of the bud-shaped end of the septum there is a rather large fenestra (Plate 9, fig. 7, s.n.), and in this enlarged figure of part of this region we see the pro- rhinal (p.rh.) ; it is very large, sub-flabelliform, and turned (contrary to rule) inwards, In this same figure the angle of the nasal capsule, a part derived from the trabecule, is seen to give off a curious retral process (al.n., s.n./.): this is quite normal. The whole palato-quadrate arch (pa., pg.) is an immense structure; measured from the pre-palatine spike to the end of the condyle, it equals the entire length of the 56 MR. W. K. PARKER ON THE STRUCTURE AND skull, measured from the snout to the occipital condyles, yet in front of the pre-palatine we have the fore part of the maxillary and the whole of the pre-maxillary. All at once in this part we come across a number of bones that have no representa- tives in the Batrachia, generally, yet some of them correspond to familiar bones in the Teleostei. I shall name these latter in accordance with my description of the ‘“ Salmon’s Skull” (Phil. Trans., 1873, Plates 6-8). As compared with what is seen in the typical kind, the palato-quadrate arch is scarcely more ossified, but it is relatively, as well as really, very much larger. Instead of a single palatine bone running across under the right-angled ethmo-palatine carti- lages, there are several separate pieces.* On the left side (fig. 2, pa.) the large sub-faleate ridged palatine is single up to the handle or inner end; there, however, three at least small sesamoid centres are seen. On the right side the handle is entire, but the blade is composed of two unequal additional bones (pa’.); these are reversed in the figure. The huge sigmoid pterygoid bone (pg.) is single from its fore end near the outer part of the palatine to the end of the quadrate condyle (¢.c.); but the thick remnant of the dorsal end of the suspensorium—the “pedicle”—is invested with, and more or less ossified by, a group of bones. Two of these, on each side, are the familiar “ meta- pterygoid” and “mesopterygoid” (Plate 8, figs. 2-4, mt.pg., ms.pg.) of osseous Fishes. The metapterygoid (Plate 8, fig. 2, mt.pg.) ossifies most of the pedicle (fig. 1, pd.), leaving the thick ovoidal condyle soft; but where it is passing inwards, at a right angle to the pterygoid bone, there in front of its outer end is a bone one-third its size: this is the mesopterygoid (ms.pg.). The large out-turned, almost vertical quadrate region is largely fenestrate (Plate 8, figs. 1-4, and Plate 9, figs. 12-14); this fenestra is bounded above by a bridge of cartilage, which passes to the “otic process” (cut through in these larger figures of the details). That bridge is largely ossified by an additional centre (pd’.) or supernumerary “metapterygoid.” The fenestra is mostly occupied by three long splints—*inter- suspensorial”” membrane bones; one is sub-falcate and twice as long as the others; their form is sub-oval (Plate 9, figs. 12-14, a.sp.). The hinder part of the proper pterygoid is itself fenestrate on its inner face (Plate 9, fig. 13, pg., sp.). On the outside this arch is well invested by the three normal subcutaneous bones, viz.; the squamosal, quadrato-jugal, and maxillary. The first of these (sq.) is remark- able for the forwardly-bowed form and oblique position of its stem (Plate 8, figs. 1 * This specimen, the gift of Dr. Murtr, came to me in a perfectly uninjured state; it had not been caught (and possibly bruised) in capture, but it had lived in the gardens of the Zoological Society for some time before it died. I often find specimens of Batrachia with some of these bones fractured in capture, but I have learned how to distinguish them from supernumerary bony centres formed naturally in the animal. There are several kinds that show additional bones, but none at all comparable to this species. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 57 and 3, sq.), for the descending position of the post-orbital process, and for the bifurca- tion of the supratemporal plate, embracing the “ pterotic” cartilage. The quadrato-jugal (q,j.) grafts itself very little; the maxillary (ma.) has a very distinct ascending facial process below the nasal; it does not, however, articulate directly with that bone, for there is a small squarish pre-orbital (p.ob.) between them. Behind this (Plate 8, figs. 1 and 3, /.) there is an oblong bone, also attached to the descending crus of the nasal (n.); this is the “anterior suborbital” or lacrymal ; this is a familiar bone in the Teleostei, and the other is common in one “ Family ” of that “ Order,” namely, the “ Siluroidei.” But these are not all the preter-normal ossicles; there is no functional septo- maxillary, as in 2. temporaria, where it lines the nasal passage, but there are a number of generalised bony points close to where that bone appears, below, in the Common Frog. On the left side (Plate 8, fig. 2, and Plate 9, fig. 7, p.mz.) these are close to the palatal junction of the pre-maxillaries and maxillaries, some larger, and others lesser, “ palato-maxillaries ;” on the right side (Plate 8, fig. 2) there are two narrow bones, larger than the largest on the left side. These do not make up the whole tale of the additional bones; the parasphenoid (Plate 8, fig. 2, pa.s.) has at its apex a separate “ pro-parasphenoid (pa.s’.) as I showed in the larva of this species (Plate 3). This bone is about a quarter the size of the blade of the large bone, which is split where it underlies the new centre. The vomers (fig. 2, v.) are large and quadrilobate, yet they neither meet at the mid- line nor reach near to the maxillaries; for the nasal region, although large really is small relatively to the huge face. The fourth or dentigerous lobe of each vomer is rounded, the two intermediate lobes that enclose the inner nostril are sharp, and the front lobe is crescentic. Above (figs. 1 and 3, .) the nasals are broad, roughly convex, in contact in their hinder part, do not reach the maxillaries externally, and together form a stb-pentagonal roof. Between them, behind, there is an emargination which passes into the notch between the fronto-parietals ( fp.) and exposes some of the girdle-bone (eth.), whose superorbital lobe (s.0b.) is also uncovered externally. The frontals are not confluent with each other, but with the parietals ; these are completely anchylosed together (fig. 1, fp.). The sides of this roof are almost parallel for two-thirds of the interorbital region ; then the bones suddenly widen, then become pinched and bi-cristate, and then double their breadth over the inter-auditory region, where their outer margin is emarginate, and their surface sinuous over the double “ canal.” The two crests are divided by a fossa ; they are first temporal, and then sagittal. This compound roof ends, behind, in a transverse, dentate, squamous edge. leaving the narrow superoccipital region uncovered. Laterally (fig. 3), the bony roof and the bony floor are seen to be modelled over and around the endocranium, hiding half of it. MDCCCLXX XI. I 58 MR. W. K. PARKER ON THE STRUCTURE AND The middle and outer ears (Plate 8, figs. 1-4, and Plate 9, figs. 8-14) are in a high state of development ; externally, the annulus (a.ty.) is almost an inch across, and is complete ; its greatest width, below, is fully a quarter of an inch; it is thick-rimmed, unusually wide at the junction of its “horns,” and altogether unique in size and finish. The “columella” also is unique as to its development, having an osseous shaft in its extra-stapedial (Plate 9, figs. 8, 9, 10, e.st., e.st’.), which is not seen in other Batrachia. The stapes and the dorsal part of the columella lie in a deep hollow, in the hinder part of which is the fenestra ovalis (Plate 9, fig. 8). The stapes (st.) is oval, obliquely truncate in front, and thick and convex externally. Against it, and somewhat within, lies the thick, short inter-stapedial (7.st.), whose distal half is bony; it is quite segmented from the medio-stapedial (m.st.) which is long, phalangiform, thick proximally where it is not quite ossified, and narrow and arcuate further forwards. Then comes a short cartilaginous tract which is followed by the extra-stapedial (e.st.) whose proximal part is ossified as a styloform shaft bone (e.st’.), the cartilage within being also considerably ossified beyond the ectosteal tract. On the inner side, near the dilated end the supra-stapedial (s.st.) is given off at a sharp angle, it passes upwards and backwards and is confluent with the cartilage lining the tegmen tympani; it is nearly as broad as the main bar, and is entirely unossified. This division of the columella into three segments is of great interest, and can be understood only by comparison with the state of things seen in the upper hyoid region of Fishes in various groups. In the Chimeeroids (see Huprecut, fig. 2, hy’’.) there is a “ pharyngo-hyal” above the “epi-hyal” (hy’.), but this topmost piece, so constant in the branchial arches, proper, is not found in the hyoid of other Fishes. Much lower down, towards its distal end, the epi-hyal (‘ hyomandibular”) is cut off as a short distinct cartilage in the Sturgeon’and Paddle-Fish. In osseous Fishes this cartilage acquires an additional bony sheath—the “symplectic” —but is not cut off. I am now perfectly satisfied that this lower subdivision of the “ epi-hyal” element is a secondary segmentation, such as is seen in the “pharyngo-branchials” of the Sturgeon, where each cartilage is sub-divided into two pieces.* In illustration of this we see that the ‘“‘ hypo-hyal” of osseous Fishes has two bony centres ; in the Menopome it breaks up into three patches of cartilage. Therefore we see that the middle ear of this Frog has parts that correspond to the following morphological elements of the upper part of the hyoid arch in the Fishes :— 1. Pharyngo-hyal=inter-stapedial. 2, Epi-hyal, subdivided into—a, hyo-mandibular=medio-stapedial ; and b, sym- pletic=extra-stapedial. * Mr. Howes pointed out this to me, as shown in his excellent dissections at South Kensington. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 59 The supra-stapedial is a secondary process of the distal part, and very inconstant ; the segmentation of the extra-stapedial from the medio-stapedial is complete in Rk. palustris and R. halecina (Plate 5), and I doubt not is the same for a time in AR. pipiens.* The mandible in this species is chiefly remarkable for its great length (Plate 8, fig. 3). All its parts are normal; the coronoid process of the “articulare” is well developed. The hyo-branchial plate (Plate 8, fig. 5) is also quite normal; it differs very slightly indeed from that of the Common Frog, and as in that kind the stylo-hyal end of the hyoid articulates with the floor of the tympanic cavity. As compared with the skull of the ‘‘norma” we have many differences that are due to their size, such as the distance of the facial bars, laterally, from the nasal labyrinth, the small size of the cranial trough, and the huge suborbital spaces. The endocranium is not so much hardened as might be expected in so large a kind ; the bony masses behind are still distinct below, but not in the superoccipital region. The investing bones generally are quite normal; the roof in the parietal region has its two sides anchylosed and crested : this is simply due to the extent and strength of the temporal muscles. The cartilaginous foundations of the great palato-suspensorial arches is quite normal, only they are very large, relatively. There are two groups of modified parts, namely, the various bony ‘‘ centres” and the elements of the middle ear. The modifications may be classified as follows :— 1. Groups of small ossicles in a generalised state below the junction of the pre- maxillaries and maxillaries, instead of a functional septo-maxillary. 2. Ectosteal bones applied to the “ palato-suspensorial arch” numerous, and in the palatine region unsymmetrical ; several of those in the pterygoid region being referable to normal zchthyic centres. 3. Further additions of bony centres in the pedicle and quadrate, both membrane- bones and endosteal patches. 4, The quadrate region of the suspensorium fenestrate, below the division into “pedicle” and “ otic process.” 5. A sub- and a pre-orbital bone on each side. 6. Parietal region anchylosed and crested. 7. A distinct pro-parasphenoid. 8. Inter-stapedial semi- osseous. 9. Extra-stapedial half osseous, with distinct, but apposed, ectosteal and endosteal tracts. 10. Supra-stapedial confluent above. * This Acipenserine subdivision of the epi-hyal element is to be seen also in the embryo of Chelone viridis.—‘ On the Skull of Chelone viridis,” ‘Challenger Memoirs.’ (‘‘ Zoology,” vol. i., part 5, plate 10, fig 8, 1880.) I 2 60 MR. W. K. PARKER ON THE STRUCTURE AND d. Ethiopian species. 14, Rana.—-—? sp.—Adult female, 1 inch long. Lagos, W. Africa.* The largest of the specimens examined by me was one inch long, and was a female with developed ovaries. Dr. GGNTHER was doubtful as to the exact species of Rana to which it belonged; it was possibly immature, but evidently belonged to a small and very slender kind, with proportionally extremely long hind legs: its tongue was deeply bilobate, and one lobe was much shorter than the other in this specimen.t My dissections show this kind to have a skull closely corresponding to that of a Common Frog of the first summer, or about four months after metamorphosis ; it is instructive any way, whether it be young or arrested, for it shows how close this Ethiopian species comes to our native kind, in all essentials of cranial structure. That it was a mature individual I have no doubt, and it came in the same bottle with adult specimens of a species of Rappia, the largest of which was only five-sixths of an inch long. The general form of the skull (Plate 13, figs. 7, 8) is half a long ellipse ; and the length is equal to the greatest breadth—a proportion which gives a rather long skull for a Batrachian. The occipital condyles (Plate 13, figs. 7, 8, oc.c.) project very little, are not so wide apart as in some small species (e.g., Rana pygmea, Plate 5, figs. 11, 12), and the emargination of the basal plate between them is slight; the roof also (fig. 7) is not cut away much above. The auditory capsules project moderately in front, but they have an arrested, smooth appearance ; their inferior surface is nearly as great as the superior; the roof of cartilage is complete up to the post-orbital region, so that there are no lesser fontanelles behind the main space (fo.). The optic fenestra (II.) is small, and very near the foramen ovale (V.). The ex-occipitals and prootics (e.0., pr.o.) are but little advanced ; the former leave a wide basioccipital, and a still wider superoccipital, tract soft; the former ride over the anterior ampulla (a.s.c.) above, and are seen outside the foramen ovale (V.) below. The interorbital region of the cranial “barge ” is widish, and lessens very gently up to the antorbital region ; there is no ossification there ; and if this be, what I am satisfied it is, namely, a mature specimen, here is an instance of the entire absence of the ‘ girdle-bone.” I shall soon corroborate this view by showing that another small kind of Frog is equally devoid of this bony tract. The ethmo-nasal region is nearly as wide as the occipito-auditory ; the roof and floor (s.n., 8.2.0.) are broad, and nearly equal. The angles of the latter, and the pro-rhinals * Collected by R. W. Wauker, Esq.; specimens of the large R. Bibronii and R. Grayi, brought me from S. Africa by the Rey. Amprose WitsoN, came too late for this memoir. + This, according to Dr. Ginrunr, is artificial, and due to contraction by the alcohol. This was, however, supposed by Hattowe tt to entitle it to a distinct generie name, viz.: Heteroglossa (see Hatio- wet, Proc. A. Philad., 1857, p. 64; and Ginruer, “ Batrachia Salientia,” p. 26). The latter author assures me that this is a normal species of Lau. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 61 *fig. 8, p.rh.) are well developed, but there is a small prenasal rostrum ; the bulbous end of septum nasi is somewhat emarginate. The fore part of the palato-suspensorial car- tilage is thick, and forms a triangular pre-palatine ; its narrow ethmo-palatine pedicle has on its postero-inferior surface a slight palatine ossification (pa.). This cartilage in its pterygoid, quadrate, and auditory regions, is slender ; the pterygoid bone ( pg.) is, however, well-developed—not like that of a young Frog—and only leaves the upper and outer surface of the cartilage naked (fig. 7). Each pedicle (pd.) is small, and they ure wide apart; so also is the quadrate region, but it ends in a large condyle (q.c.) that reaches as far backwards as to the middle of the stapes; this is quite an average retreat backwards. The Eustachian openings (ev.) are of medium size, oval, and transverse, and the space of the middle ear is well developed. The annulus (a.ty.) is large, and its horns nearly meet; the stapes (fig. 11, st.) is large and oval, with a moderate anterior scooping for the columella, This rod is well ossified in its main part, and has no appearance of belonging to a young individual ; its interstapedial segment (7.st.) is small, pisiform, and lies between the large unossified lobes of the long, arcuate medio-stapedial (m.st., m.st’.). The extra- stapedial (e.st.) is a small, long-oval lobe of cartilage, and has no supra-stapedial process. The mandible (fig. 9) is quite normal. The styloid end of the hyoid bar (st.h.) is articulated with the ear-capsule ; the bar itself (fig. 10, chy.) is narrow, and has no hypo-hyal lobe ; the two side lobes of the basal plate (b.4.br.) are small; the thyro- hyals (¢.hy.) are normal. The narial valves (fig. 12, w.J'../7?.) are well developed and perfectly normal. The investing bones lend no support to the opinion that this is anything but a “mature individual : they are well developed; those on the upper surface differ from their counterparts in a full-grown Common Frog, by the nasals (n.) being relatively much larger and nearer together, and the fronto-parietals (/-p.) completely anchylosed along the middle—all save their pointed front ends, where a little of the fontanelle (fo.) is exposed. This common sheet of bone well covers the roof behind, and although the bone is thin, it is highly developed as bone tissue. The nasals (n.) touch the septum nasi (s.v.) and curving elegantly outwards in front, overlie the large sub-tubular external nostrils (e.n.) which are very wide apart. The fore part of the endocranium is well wrapped in the nasals, the pre-maxillaries (px.) and the maxillaries (mx.); these outer bones are but little developed in the palatine region, but are large and foliaceous on the outside. The jugal part of the maxillary is joined to a quadrato-jugal (qj.) which is well grafted on to the quadrate: a good sign of maturity. The upper part of the squamosal (sqg.) is narrow, but all its regions are well developed. The parasphenoid (fig. 8, pa.s.) is well developed and large, but is not so elegant as in the Common Frog, its angles behind being attenuated and not outspread ; the vomers (v.) are small crescents, widest in front and forming a mere hook behind ; 62 MR. W. K. PARKER ON THE STRUCTURE AND they are toothless—a state of things commonly seen in small types of tooth-bearing* Batrachia. My specimen of this kind was equal in size to many other dwarfs in this group of Vertebrata, and the evidence in favour of its maturity greatly outweighs that which seems to tell of immaturity. Taking its skull then as that of an adult, and comparing it with the “norma,” I find the following discrepancies, viz. :— 1. No septo-maxillaries. . Fronto-parietals of each side coalesced. . Quadrato-jugal ossifies part of suspensorium. . No teeth on vomers. No “ girdle-bone,” even in rudiment. . Inter-stapedial very small. No supra-stapedial, and extra-stapedial arrested. mI a om FR ww bo . Lesser fontanelles absent. Skulls of “ Ranwde” not comprised in the typical “ genus.” 15. (A) Tomopterna breviceps.—Half-grown female ; length, 14 inch. 8. India. This and the next kind are both put into the same genus in GUNTHER’S ‘ Reptiles of British India’ (p. 411); but more recently the author has put this short-headed, high-faced species into the genus Vomopterna, removing it from Pyxicephalus. This is as it should be. The two (this and P. rufescens) differ quite enough to entitle them to be placed in distinct generic groups; this species comes nearest to the Rane, the other is a more aberrant form. The outline of this skull (Plate 14, figs. 5, 6) is half a short ellipse, and the width is to the length as 8 to 63; the general outline is very regular. A full grown individual would have shown a more irregular outline and denser bony centres, but the skull is fairly finished at this stage; size and strength are all that is gamed afterwards, The occipital condyles (Plate 14, figs. 5, 6, oc.c.) are rather wide apart, and project but little. The auditory capsules carry the hind skull out considerably, opposite the quadrate hinge ; and this hinge is opposite the middle of the stapes (q.c., vb.) as in the last. The suborbital spaces are very large, and each forms two-thirds of an almost regular circle. The interorbital region of the cranium only lessens very gradually forwards, and it swells out gently in the middle. The cartilaginous roof of the skull is complete up to the orbital region; there is only the main fontanelle. ce The whole fore skull is extremely like that of the “norma,” but the prenasal (fig. 9, p.v.) is a more distinct bud, and the pro-rhinals (p.rh.) are larger hooks ; also the outer angles of the subnasal laminze (s.7./.) are simpler. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 63 The hinder bony centres (e.0., p7.0.) are just such as would be found in a Common Frog two-thirds grown ; uow, a considerable amount of roof, floor, and periotic capsules are soft. So also the “girdle-bone” (eth.) is not of great extent, it hardly reaches half way to the large optic fenestra (II.) ; from this space to the foramen ovale (V.) the wall is also unossified. The palato-suspensorial arch, and its ectosteal plates (pa., pg.), are quite normal; the condyles of the quadrate are very large, and the cartilage at that part is but little affected by the quadrato-jugals (q,j.). The pedicles are wide apart, and well developed ; the Eustachian passages (eu.) are large and oblique. The annulus (a.ty.) is moderate and complete. The stapes (Plate 14, fig. 10, st.) is large, thick, and oval ; the medio-stapedial (m.st.) fits inside it by a decurved unossitied lobe ; but the inter-stapedial is not distinct. The medio-stapedial is therefore large behind ; it is straight, and joins on to the narrow stalk of a broadly spatulate extra- stapedial (e.st.), which becomes double on its inner face, giving off a strong ligulate supra-stapedial (s.st.) which is confluent with the “tegmen” above. The stylo-hyal part of the hyoid band (st.4.) articulates with the floor of the tympanium, and turning round, borders the Eustachian opening (ev.). The investing bones, one and all, are extremely like those of the type; there is a septo-maxillary on each side, but it is small. The mandible shows in the dentary (fig. 7, d-) very little of the crest; yet in old individuals the mandible is high at that part, and also somewhat hooked at the synchondrosis, like an old male Salmon. The hyo-branchial plate (fig. 8) is normal, but its lateral lobes are badly developed. 15 (continued).—(B) Yomopterna breviceps.—Adult female, 2 inches long. Ceylon. The distinction between the skull of the half-grown young of this species and the young of a Common Frog of the same age, is slight as compared with what is seen in their adult condition (Plate 15, figs. 1-4; and Phil. Trans., 1871, Plate 9). The facial outline now forms half a rather short ellipse, and the greatest breadth is nearly one-fourth more than the length; moreover, this is one of the highest (or deepest) of the skulls in the whole Order. Indeed, this Frog is an isomorph in respect of its short deep head and its thorough want of neck of the most remarkable Toads of the same (the Oriental) territory, e.g., Callula, Diplopelma, Cacopus. This, again, is an instance of what is seen in the main geographical territories, viz. : that some particular modification characterises the members of very different “ Families” of the same Order, as if the Anurous type had, in each territory, broken up into groups isomorphic of, but not immediately related (genetically) to, those of other terri- tories. Thus from a generalised root-form, in territories wide apart, there may have sprung, in each place, independently, Frogs with teeth and Zoads without teeth; 64 MR. W. K. PARKER ON THE STRUCTURE AND types with sharp, or broad toes; with narrow or wide backs; with or without neck glands; with soft or stony skulls. For this tribe is notable, perhaps above all others, for ready response to surroundings; here is a swimming creature who learns to crawl, to walk, to leap, to climb, and to whom (in a rare instance) even a sort of flight is not denied. Moreover, the slow metamorphosis of the larva may take place in these types in the water, in the damp floor of the jungle, in the foliage of trees, or ina large bag, or in little pockets, on the back of the mother. This Frog, which is about equal in size to our native species but much stouter, has its skull even less ossified than the common kind; this is true both of the endocranial “centres” and the investing plates. In this it differs from most of the species of Rana of its own territory, but agrees with them in other things, rather than with that familiar species, whose distribution reaches even to us. The nasal and auditory regions are about equal in axial extent ; the orbital region is one-third longer; the hind skull is of extraordinary breadth. The occipital condyles (0c.c.) are large, reniform, and postero-inferior; they are sepa- rated by a shallow notch less than their own breadth, and the foramen magnum ( fm.) is large and oblique, the roof retreating forwards. The double canal (q.s.¢., p.s.c.) is large and projects beyond the less evident horizontal canal (h.s.c.); the capsule pro- jects as an obliquely oblong tract. The ex-occipitals (¢.0.) do not reach to the stapes, right and left, and only slightly overlap the parasphenoid ; the prootics (pr.o.) guard the outside of the foramina ovalia (V.) and mount up over the two front ampulle ; thus there is a very large cruciform tract of cartilage on the floor of the hind skull. There are no lesser fontanelles, and the main space is only half as long as the cranial roof, and is long and emarginate behind ; it is partly uncovered in front (fig. 1, fo.). The interorbital region is narrow, lessening steadily forwards; the roof is enlarged a little in front by the small oblong superorbital tracts of cartilage (s.ob.). The long, oval, optic fenestra (II.) lies in the middle of a large unossified tract, which leaves the girdle-bone (eth.) only two-fifths of the orbital space; that is, however, compensated by the bone running almost to the pro-rhinals, below, and along the hinder third of the septum nasi (s.v.), above. The nasal roof is only two-thirds as large as the floor (s.n., s.n.l.), the front of it is elegantly sinuous, for the thick septum projects as a small subconical prenasal rostrum. The angles of the floor (s.7./.) project well into the fore part of the wide face, and the pro-rhinals (p.rh.) are large, long, and uncinate. The rising wall (al.n.) grows up behind the nostril from the surface of the angle of the floor, but does not form a perfect ring by coalescing with the roof as in the larger Oriental Frogs. The narial valves (w./'.v.2?.) are large and normal. The palato-suspensorial arches are slender behind, but in front they are thick at their root (e.pa.), and have a large pre-palatine spike (in front of pa.), and continue wide inside the maxillary until they end as a lobe (pt.pa.) from which the narrower ptery- goid cartilage grows. The pedate pedicles ( pd.) are sharp and well formed, and reach DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 69 almost to the prootic and parasphenoid. The angle of the fork is somewhat rounded, so that the large Eustachian opening (ev.) is sub-triangular. The bony plates (po., pg-) are normal, but thin and lathy. The quadrate region or outer fork is very long, and retreats as far back as the middle of the stapes : this is one of the deepest cheeks to be found in the group. The cartilage is very slightly ossified by the quadrato-jugal (q.j.); the condyles (q.c.) have the peculiarly elegant form seen in the Oriental Rane, viz.: a very large, long, postero- internal trochlea, and a rounded, small, antero-external convexity. The annulus (fig. 3, @.ty.) is rather large and perfect, as in the congeners of this species ; the stapes (st.) is large, oval, and umbonate. The medio-stapedial (m.st.) has no proximal intervening segment cut off, the extra-stapedial (e.st.) is spatulate, and the supra-stapedial is confluent above. The mandible (fig. 3) is normal, the condyle (ar.c.) is long and subreniform. The stylo-hyal (fig. 2, st.h.) is narrow at first and confluent; it widens gradually up to the hypo-hyal bend below (fig. 4, c.hy., h.hy.) ; over the curve there is now a small thin extra-hyal (e.hy. for ex.hy.). The notch in front of the basal plate is wide, and the whole structure is normal; the front lateral lobes are now fan-shaped, and the hinder ligulate. The thyro-hyals (t.hy.) are large and bent outwards. The investing bones are very thin and splintery; the fronto-parietals ( fp.) are feeble and arrested in front; the nasals (v.) are feeble and ragged, like the newly-formed osseous centres of a young specimen ; the premaxillaries (pz.) are wide and well developed as to their processes ; there is a pair of small seed-like septo-maxillaries (s.ma.). The maxillaries (ma.) are high, but thin; they are notched in front where they overlie the angle of the nasal floor, and have a bilobate ascending plate; they stretch along more than half of the temporal space behind. The dentiform quadrato-jugals (q,j.) are only slightly grafted on to the quadrate ; the squamosals (sq.) lie well over the narrow tegmen, and have a shortish diamond-shaped postorbital process. The descending bar (fig. 3, sq.), is of great length, and widens gradually downwards. The parasphenoid (fig. 2, pas.) is quite Ranine, and its basi-temporal plates are very large and bilobate externally. The vomers (v.) are Cystignathine ; of the three spurs the foremost is twice the size of those which fence in the inner nostril (7.n.) ; the body is oblique, arcuate, almost reaches the middle, and is armed with an almost straight crest of teeth; the two crests form somewhat more than aright angle, but their sharp outer end is strongly turned forwards. As compared with that of our Native Frog, this skull is— . Much broader and deeper. . The hinder centres of the endocranium are smaller, and the girdle-bone is larger. . There is only the main fontanelle, which is rather small. m Oo tb . The parotic wings are extremely outstretched. wo . The interorbital region is very narrow in front, 6. There are distinct superorbital “ eaves.” MDCCCLXXXI. K 66 MR. W. K. PARKER ON THE STRUCTURE AND 7. There is a short prenasal rostrum. 8. The pro-rhinals are unusually long and folded over. 9. There is no inter-stapedial, and the supra-stapedial is confluent above. 10. The suspensorium of the jaw is of great depth. 11. There is a small extra-hyal on each side. 12. Most of the investing bones retain their early condition as thin ragged lamine. On the whole this instructive skull shows that Tomopterna is a Frog close akin to the species of Rana, but leaning towards the Cystignathide ; but it does not lie directly between the two families, for it has been modified, as « Frog, much as those neckless types Cacopus, Diplopelma, and Callula have been modified, as Toads, Here we have a problem relating to the influence of territory; that which gave the apoplectic look to the toothless kinds, doubtless thrust together the head and shoulders of this tooth-bearing type. 16. Pyaicephalus rufescens.—Adult male ; 1 inch 5 lines long. India. This a much slenderer, smaller, and more warty kind of Frog than the last. Its toes are much more perfectly webbed in my specimen than in several specimens of Tomopterna breviceps in my collection. Dr. GinrHer says that the interdigital membrane is equal in both kinds (ibid., p. 412). The outline of the head is so elongated and pointed as to be almost triangular— a great contrast to the last. Dr. GUNTHER says that the eyes are much smaller than in the last kind (p. 412); this is a correlate of the narrowing of the skull forwards. The length is to the greatest breadth as 8 to 83; for although narrow in front, this skull is very wide behind. The occipital condyles (Plate 14, figs. 1 and 2, 0c.c.) are large, reniform, inferior, and wide apart; an evenly emarginate tract of cartilage, two-thirds their own width, separates them. This basioccipital cartilage is twice as broad as the superoccipital synchondrosis, and the bony masses formed by the extensively spread prootics and ex-occipitals (pr.o., e.0.) are much nearer together above than below; there they deviate more and more from the foramen magnum to the foramina ovalia (V.), which they reach below. Above (fig. 1), there is a large transversely oval space unossified, and with small secondary fontanelles; the prootics meet in front of this, and are far extended in the roof. Laterally, the prootics and ex-occipitals are confluent, but only affect the posterior, and the arch of the anterior, canals (fig. 1, av.); the rest of the divergent ear-mass is soft: below, an oval floor is left soft to the vestibule, externally (fig. 2, vb.). The actual cranial cavity is three times as wide in the post-orbital as in the ant- orbital region ; far from being an extremely wide skull, it lessens forwards, to become one of the narrowest known. Its lateral margins are sinuous, the middle bellying a DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 67 little. The narrow part opens out elegantly into the ethmoidal wings; up to them the girdal-bone (eth.) reaches, and no further. Above, this bone reaches half way to the extended prootic tract ; below, two-fifths of the walls are soft, and this unossified tract has in its middle a very large optic fenestra (II.), and behind, it encloses most of the foramen ovale (V.). The nasal region is quite unlike that of a typical Frog; above (fig. 1), the nasal roofs are very narrow, and merely form a crescentic selvedge to the septum nasi (s.7.), and on to the fore margin of the ethmo-palatine band. The narrow roof curls round the front of the outer nostrils (¢.7.), which are very near each other—only one-third the distance apart that we have just seen in the Frog from Lagos (Plate 13, fig. 7, e.n.); behind the opening the nasal wall (n.w.) is thick and crescentic. This narrowing of the nasal end of the skull is made more remarkable by the development of a very distinct and pointed prenasal cartilage (p.n.).* But below (fig. 2, s.n./.), the trabecule have united to form a very wide elegantly winged tract, which passes between the laminz of the maxillaries (im.), externally : behind, it is narrowed, for each margin is cut away by a semi-circular notch, through which the internal nostrils (7.n.) pass. The “ pro-rhinals” (p.rh.) are very long, slender, and bent back upon themselves. These processes (Plate 15, fig. 5, p.rh.) are impacted between the two lamine of the premaxillary ; they are equal to the prenasal in thickness. The nasal valves (w./?.) are outside the nostrils ; they are, in form, quite normal, but very small. The whole of the palato-suspensorial arch is quite normal, and rather slender; the same is true of the bony plates (pa., pg.). The pedicles (pd.) are wide apart, for the auditory masses are relatively large, and widely outspread. The condyles of the quadrate (q.c.) are a little further back than in the last, are opposite the end of the stapes, and are large and reniform ; there is no erafting of the quadrato-jugal over them. The Eustachian openings (fig. 2, ew.) are very large and are turned obliquely backwards, outside ; the annulus is very large, and like that of the Bull-frogs (Eastern and Western). The stapes (fig. 4, st.) is large and like that of the ‘norma ;” the columella is very generalised ; there is no inter-stapedial segment at the scooped top of the club-shaped medio-stapedial (m.st.) whose interstapedial end (i.sf.) is not ossified. By a short stem, the bony tract is connected, in front, with a small extra-stapedial (e.s¢.), “annulus,” below, is more which has no secondary process ; here, the breadth of the than twice as great as that of the circular extra-stapedial : this is the converse of what is seen in Rana Kuhli and Dactylethra. The investing bones are about as strong as those of the skull of “the type,” when * The right dotted line from p.n. in fig. 1 points to the nasal process of the right premaxillary ; behind its apex the first upper labial is hidden. K 2 68 MR. W. K. PARKER ON THE STRUCTURE AND half-grown—or equal to the size of the full-grown individuals of this species. The parasphenoid (pa.s.) is very large, both in its main part and wings; this form is like that of the Common Frog, but the point is longer and sharper. The vomers (v.) are typical ; the septo-maxillaries extremely small 7f present, and the bones of the outer face normal, but slight. The nasals (7.) are long crescentic shells of bone ; the fronto-parietals (fp.) are large and out-spread over the hind skull and then become gradually fine sharp styles in front, leaving part of the girdle-bone and a little of the fontanelle (fo.) uncovered. ’ The “ frontal suture” is permanent ; the “sagittal” is filled in with bone; the two sides being continuous over the inter-auditory region. The mandible (fig. 3) is quite normal. The stylo-hyal end of the hyoid band (fig. 2, st.h.) is distinct and pointed as it passes behind the Eustachian opening ; this band widens to double its breadth (fig. 3, chy.) in the lower half, and then turns suddenly round to join the basal plate (h.hy., b.h.br.). Round the outside of the broad distal half there is another band of cartilage as wide as the upper half of the main band. This is an oe extra-hyal” element (ea.hy.); it passes over the hypo-hyal loop as a short hook. Other Oriental Batrachia show this, but not so distinctly, e.g., Tomopterna, Callula, and Diplopelma, and the Australian Tree-frogs have a rudiment of it: here it is most largely developed. If this cartilage be compared with the pectinate “inter-branchial” of Chimera (Husrecut, fig. 2), it will be seen to correspond with the base of that comb-like cartilage. In Tadpoles the extra-branchials send pectinate processes inwards, but they show no separation of the extra-branchial bands from these “rays” we see in Sharks, for the branchial apparatus of the Tadpole is as highly generalised, as that of the Shark is intensely specialised. The “spiracular cartilage” belongs to the suspensorium (of the 1st arch), and becomes utilised as the annulus tympanicus ; this rarer ‘‘extra-hyal” and the four extra- of the septa, such as branchials all belong to one category. This infero-external element of the hyoid arch, like the supero-internal (epi-hyal or “columella ”), does not appear until after the metamorphosis of the Tadpole; that is utilised as part of the apparatus of the fast-improving ear; the extra-hyal merely serves as an additional platform for the thin fan-like muscles of the throat. The rest of the hyo-branchial structure is quite normal ; the lateral lobes are highly developed, and the thyro-hyals strong (fig. 3, t./y.). Besides the remarkable shape of this skull, so wide behind and so pinched in in front—necessarily modifying the form of the investing bones—we have the subjoined differences from the typical form, viz. :— DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 69 . The septo-maxillaries are suppressed (?). . The fronto-parietals confluent behind. . The pro-rhinals doubled over as hooks. Co h) 4. A distinct sharp prenasal. 5. Very narrow nasal roofs, 6. No inter-stapedial segment. 7. No supra-stapedial band. 8. A large long uncinate ‘ extra-hyal.” The folding over of the pro-rhinals, the length of the prenasal, the generalised state of the columella, and the additional hyoidean element, indicate a very marginal posi- tion for this type in the Family “ Ranide.” Second Family. “ CysTiGNaTHID”.” First genus. Psevdis. 17. (A) Skull of Pseudis paradoxa.—First larva ; total length, 10} inches ; head and body, 33 inches; tail, 7 inches; greatest width of tail, 4 inches; depth of body, zg inches; hind legs, } inch. §. America. A side view of this, the youngest and largest Tadpole of Pseudis (natural size) (Plate 1, fig. 1), shows to what a magnitude the larva of a medium-sized Frog may grow ; for the old individuals are but little larger than our native Grey Frog. As to actual length, more than two-thirds of this larva is a temporary structure, and belongs to the tail; the bulbous fore part, below, is half pharyngeal and half abdominal. Supposing a measurement by vertical lines across the length of the creature, the dis- tance between the anal aperture and the occipital hinge is one-third the length of the essential animal; the single gill-opening (left side) is at the middle ; and the eye-ball two-fifths from the front. Altogether, the outlets and inlets to the essential animal are five - three azygous—the oral, anal, and pharyngeal; and two paired, the narial ; these latter are above and behind the upper lip, and are small and rounded: they open into the palate almost vertically. The mouth is small but complicated (Plate 1, fig. 2), the upper labials (Plate 2, figs. 1, 2, w./.) arching over the horseshoe-shaped lower labials (Plate 10, fig. 5, /./.) and the small oral passage opening being between the halves of the horseshoe : the skin over these labials is formed into a horny plate, and the plice of the lips are developed into horny rasps. Much fibrous tissue—some watery, some strong and tough—encloses the skull and the outworks ; above, the head is marked by a median, and a pair of lateral, ridges ; the facial cartilages are spread out on each side and lie ona plane but little below that of the cranial cavity (see Plate 2). The “chondrocranium” at this stage is one of the best for comparison with that of 70 MR. W. K. PARKER ON THE STRUCTURE AND the Lamprey, but it has already gone beyond the suctorial type in having both paro- steal and ectosteal plates ; it is profitable also for comparison with that of the larvee of Dactylethra and Pipa. The free growth of cartilage, here, gives us a character in this type which I have not found in others, but which Gorre describes in Bombinator, namely, that the noto- chord (nc.} is ensheathed in an azygous, tubular cartilage, as in the Salachians.* Hence the parachordals are very narrow bands between this tube and the ear-sacs ; and very short also; for the trabecular apices embrace the fore end of the notochord, The cartilage, right and left, has formed a perfect arch above, and the flat roof runs up to the post-orbital region (Plate 2, figs. 1 and 3, s.0.). Behind (Plate 2, fig. 3), the occipital condyles are forming; they look directly back- wards, and the bony matter (e.0.) runs into them. The auditory capsules were not relatively large at first, but they are so largely developed outwards now as to seem of unusual size, being winged above and below. By the study of the growth of the skull in many types I have satisfied myself that the upper wing (“tegmen tympani”) has not the same morphological import as the lower, or floor of the tympanium, both of which are enormously developed. The basal plate, formed by the trabeculee and parachordals, grows round the base of the ovoidal ear-sacs, and appears outside, especially in front : hence this copious growth of cartilage, forming elegant lower wings to the ear-masses, and serving as a shelving tympanic floor (fty.). This undergrowth showing itself outside the ear-sacs, appears in many forms; it is a manifest striving of the basal plate to meet the arches that belong to it, this junction being impeded by the huge sense-capsules impacted in at this part. ” But, as in the “ Aglossa,” the “tegmen tympani” (¢.ty.) is a superficial growth. of cartilage applying itself to the capsule, in the margin of which is imbedded the hori- zontal canal; this roof-plate is very large, still larger than the floor (Plate 2, figs. 1-3). This tegmen is very thick as well as wide; it is widest in front, and there it over- laps, and is confluent with, a large process from the elbow of the suspensorium (ot.p.). Here the tegmen cranii, and the tegmen tympani, of each side, remain soft, but the labyrinth is almost entirely enclosed in a generalised occipito-petrosal mass (e.0., 7’.0.). This is an undistinguishing spread of bony matter over and under these regions, a division into “ periotic” and “ex-occipital” being attempted later in the growth of the head, It is very solid bone, and leaving the azygous cartilage untouched below, and a similar breadth of roof above, it builds a side wall to the foramen magnum, forming a solid rough kind of masonry, and enclosing most of the “canals” and part of the “ sacculus.” The cartilaginous stapes (Plate 2, fig. 3; Plate 10, fig. 5, st.) lies in its fenestra just under the middle of the “tegmen,” and above the deep shelving floor ( ty.) * The outer sheath of the notochord is extremely thin in most of the Batrachia, om DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 71 Already the passages for the 5th and 7th, and the 9th and 10th nerves (V., VIL, IX., X.) are being enclosed in bone. The trabeculz and “intertrabecula” form the whole skull from the notochord to the upper labials, the hinder half being cranial and the fore half nasal. The fontanelle (fo.) is not large, and it appears of a pyriform shape, because of the fronto-parietals (fp.) which partly cover it; they are strong wedges of bone close together behind, where they are broad, and wide apart in front, where they are narrow. There is an enlargement of each bone in the temporal region, and from thence the uncovered skull rapidly widens into the auditory masses. In front of that point it is almost oblong, only gently widening towards the front, where it is continuous with the large facial bars. The floor is flat—scarcely convex—and above the edge of the floor the sides are scooped, and shelve inwards, so that the flat top is narrow, especially in front, where the ‘“‘tegmen ” (¢.cr.) reappears. The tegmen is continued in front of the closed cranial cavity, whose only fore outlet is through the olfactory foramina (I.); therefore between these holes there is a part answering to the “perpendicular ethmoid” (p.e.). This is continued forwards as a crest growing up from the line of junction of the trabecule ; it is the “septum nasi” or fore part of the intertrabecula. Beyond this wall, with a slight interruption, the trabecule are united for a space, and then are free, and diverge at their end. These “cornua” are convex above and concave below, with a thick outer and front edge (Plate 2, figs. 1, 2, c.tr.); and this part, ending in a flat facet, articulates with a like facet on the upper labial (w./.). The coalesced part of each cornu is somewhat constricted, and a similar lunate outline exists in the bar outside. Thus the inner nostril (7.7.) is made ; it is finished in front by the short pre-palatine ligament (pr.pa.), both cartilages approximating. Nearly the middle third of each trabecula is continuous with the outer band ; this is in the ethmoidal region, which is very extensive. There are in reality two bands—one outer very large, and a smaller band lying between this and the trabecula; this lesser part is the “ post-palatine” ( pt.pa.), and the huge bar outside is the suspensorium of the lower jaw (p.pg-, sp.). The inner bar is half the length and one-third the width of the outer; its hinder third is free, projecting backwards and hooked a little outwards between the trabecula and the suspensorium. Its fore part is continuous with the inner edge of the suspensorium, where it bounds the inner nostril (7.n.), and ends in the inturned pre-palatine spike. The middle part, which is the longest, is the conjugational tract binding together the two main cartilages, outer and inner (¢.¢7., sp.). The upper part of the ethmo-palatine is strongly crested on its inner side, and this crest meets the lateral ethmoidal wall at an acute angle behind the inner nostril (Plate 2, fig. 1, p.pg., un). 72 MR. W. K. PARKER ON THE STRUCTURE AND . The nasal capsules are membranous as yet; and this crescentic ethmoidal wall, which articulates with the “tegmen,” bounds them behind. The dorsal element of the mandible, or its “‘suspensorium,” is the most extraordinary thing in this remarkable “chondrocranium.” Each bar is equal in size to the combined trabeculze cranii, and to each is suspended three other cartilages, and a process which becomes a free ray (ot.p.). The dorsal end of this bar early coalesced with, or rather grew from, the trabecula, beneath the emerging trigeminal nerve (V.); that part is now narrow, but terete; an elegant crescentic margin to the subocular fenestra (Plate 2, figs. 1, 2, s.o,f) is formed by this “ pedicle” as it passes outwards and forwards, and becomes the main bar; that bar is becoming rapidly widened, so as to be five or six times the breadth of the dorsal end. A rounded notch exists between the ‘ pedicle” and the spiracular cartilage (ot.p.), and this cartilage is continuous with the tegmen tympani at its distal end, and with the “elbow” of the suspensorium at the proximal end. The “elbow” of the suspensorium passes outwards into the cheek as a sides projection, with a rounded outline; the bar is then bent in a faleate manner, so as to run into the face where the ethmo-palatine projects. ‘In front of this, opposite the middle of the large conjugational tract, there is the pyriform, gently concave condyle for the cerato-hyal ; it is just beneath the edge of the bar (Plate 2, figs. 1 and 2, hy). Over this part the larval Batrachian “ orbitar process” grows upwards and inwards as a sessile, semi-oval leaf, with decurved edges and a swollen base; it is attached to the post-palatine crest by a short ligament, and is not confluent as in Bufo vulgaris. The decurrent enlargement of its edge runs backwards insensibly into the main bar as a thickened margin; in front, it projects over the edge of the bar as a free point, and then runs along the rest of the bar as a selvedge (Plate 2, fig. 1, o7.p.). From thence the anterior fourth of the suspensorium is a many-sided flap, not sensibly lessened in width, and having on the outside a snag, and in front the convex sinuous condyle of the mandible (q.), which looks a little inwards, and is only a trifling distance behind the cornu trabecule. That ending, however, is but the beginning of the proper mandible, which is a stout, transversely-directed, sigmoid bar (Plate 10, fig. 5, mk.) ; it is twisted and notched (like the human ulna) to roll upon the quadrate. To its flattened inner (distal) end is attached the lower labial (Plate 10, fig. 5, 1.2), a thick, short, arcuate bar, lessening in size downwards, where it is attached by a strong ligament to its fellow to form the horseshoe, or imperfect suctorial disk. The elongated angle of the upper labial (w./.) overlaps the double jaw-piece (labio-mandibular), and is represented by a separate cartilage in the Lamprey, and in the Tadpoles of many kinds of Batrachia. The hyoid (Plate 10, figs. 5 and 6, chy.) is a large irregular lozenge of cartilage, growing towards the mid-line below, and connected to its fellow by simple cartilage (c.hy., b.hy.). Externally, or above (Plate 10, figs. 5 and 6, c.hy., st.h.), it sends out a DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 73 stout angular styloid process, on the upper surface of which there is a large, oval, convex condyle, which rolls in the shallow cup of the suspensorium. The sinuosities of the hyoid bar are filled in by the convexities of the branchial apparatus. In the middle, behind the soft basi-hyal, there is a large pyriform basi- branchial (Plate 10, fig. 6, b.br.), composed of hyaline cartilage ; a short process on the postero-inferior surface of this is the rudiment of a second basi-branchial segment. Outside and behind this median piece there is a pair of flat lozenge-shaped cartilages, the hypo-branchials (/.br.) ; these grow outwards and are connected with cartilages above (inside) and below (outside) ; these latter are the branchial pouches. Two of the small upper cartilages are distinct from the hypo-branchials, but are partly confluent with the large outer bars (¢.br1., c.br*., ex.br.); the two hinder rudi- ments (c.br* *.) are continuous with the hypo-branchial plate, and the space between the two is filled in with cartilage. The first and fourth outer bars (ex.br'., ex.br*.) are pouch-like, the others are thin broad bands. The rudimentary inner arches (c.br.) are less differentiated than in the larva of the species of Rana and of other kinds (Cyclorhamphus, Calyptocephalus, Cystignathus, &e.) that I have worked out. These parts are at their fullest development at this stage. I have already mentioned the roof-bones in relation to the fontanelle; the para- sphenoid is a dagger with a guard, but without a handle. It seems small, yet it occupies the same place, and has precisely the same relations as in the adult (Plate 2, fig. 2; and Plate 10, fig. 2, pa.s.); it is one-third longer than in the old male. Even now it is split in front: a character which is retained throughout life. 17 (continued).—(B) Second Tadpole of Pseudis paradoxa.—7 inches long; tail, 43 inches; greatest width of tail, 2 inches; hind legs, 3 inches long; fore legs hidden. In this stage the legs are six times as long as in the last, and the tail two-thirds the length and half the width; here the chondrocranium is but little more than half as broad across the suspensoria and only two-thirds the length.* There are many things to be noticed in this stage besides its lessened size and more oblong general shape. And first, the Selachian character of a huge notochord enclosed in a tubular azygous cartilage is now as diflicult to find as in most Tadpoles; it has become part of the basal plate by coalescence with the “ parachordals” and trabeculz, and the gelatinous axis is * If I had only found these discrepancies in size in two or three, I should have thought it accidental ; but my specimens are too numerous, and run over too many stages, for there to be any mistake: the skull in the third stage (C) is but little more than one-third the length of that of the youngest (A). MDCCCLXXXI,. L 74 MR. W. K. PARKER ON THE STRUCTURE AND very much shrunken. The wild and general growth of bony matter, without division into periotic and occipital regions, is now still more remarkable, and, above, the right and left tracts are rapidly coalescing, so that there is no superoccipital cartilage (Plate 11, fig. 1). This bony matter reaches forwards to the great fontanelle (fo.) and laterally up to the inner margin of the horizontal canal (inside t.ty.). Below (fig. 2) there is a wide basioccipital tract (iv.) unossified, and the bony matter only skirts the floor of the vestibule ; it runs, however, well round the pre- and post-auditory nerve-passages (V., [X., X.). The steeply-sloping floor of the tympanic cavity (Plate 11, fig. 3, fty.) has become much reduced in size, and thus the oval stapes (st.) is seen clearly from below. The fontanelle ( fo.) has the same relative size, and so have the fronto-parietal bones (f-p.); but the upper part of the skull in the ethmoidal region has altered greatly ; ready to alter still more as the tail, and indeed the whole creature, keeps lessening in size. The ethmoidal roof (Plate 11, fig. 1, p.e.) has grown over the growing septum nasi, but the nasal roofs are still membranous, or only composed of soft cartilage. The back wall of the nasal fossze is not so distinctly articulated to the edge of the tegmen cranii; but, gently shelving down, the median part passes into a concavity which lies between the roof and the post-palatine (pt.pa.). The cornua trabeculee (c.tr.) have largely united together, but a small hole some distance behind the notch may be still seen : this soon fills in, it represents the large open space in the “ Urodeles.” The post-palatines (pt.pa.) have been, as it were, moulded into a more solid, but altogether a rounder and smoother, structure; the sharp crest has become a neat longitudinal ridge, and the hinder process is not a flap but a geniculate projection into the fore part of the suborbital fenestra. Laterally (Plate 11, fig. 3, pt.pa.), it appears as a gently concave plate, like a “ post- zygapophysis ;” the erbitar process playing against its scooped face. The fore part of the palatine has chondrified the pre-palatine ligament, and the apex of this bar abuts against the trabecular cornu (pr.pa., ¢.tr.); thus the internal nostril (i.n.) is fairly enclosed by cartilage. Beneath (Plate 11, fig. 2), the trabecular, palatine, and suspensorial regions all pass gently into each other; this, however, is a very temporary condition. Besides the general lessening of the cartilage, the suspensorium is not now so bent outwards, and the “ pedicle” has bent itself into a sigmoid form for want of room (fig. 2, pd.). The space between the “elbow” of the suspensorium and the spiracular cartilage or otic process (of.p.) is larger, and that band seems to belong equally to the tegmen tympani and the suspensorium; the ridged edge of the orbitar process (o7.p.) runs along to the tegmen, strengthening the whole band ; this “ extra-suspensorial” tract has no counterpart in the Lamprey. The quadrate region (q.) stands out further from the trabecular cornu, and its DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 75 condyle looks more directly forwards, and not so much inwards; the orbitar process is more elongated. The lower arches and labials (Plate 11, figs. 3-5, mk., LL, c.hy., b.hy., b.br., h.br., c.br., ex.br.) are but little changed. 17 (continued).—(C) Tadpole of Pseudis paradoxa, with all the legs large and free ; tail, 5 inches long ; its greatest width 14 inch. The chondrocranium (Plate 11, figs. 6, 7; and Plate 12, fig. 1), although relatively wider on account of the throwing out of the suspensorial bands, is only half the length it had, even in the second stage (B), and a third the length of that of the first (A). This temporary transitional form is very extraordinary ; I have caught no stage quite like it in any other species. The third Tadpole of Rana clamata (Plate 4, figs. 5, 6) comes nearest to it. The ethmo-nasal region, which was one-half the length of the skull in the last stage, is now less than one-third ; thus the orbital and auditory regions are now very large, relatively, and with them the cranial cavity is equally increased in size (relatively). The occipital condyles (Plate 11, figs. 6,7, oc.c.) are now wider apart ; but the ossifi- cation here is but little increased in extent. The bony substance has united over the occipital roof, and a definite tract of cartilage lies in front of the bone. There are no “secondary fontanelles,” and only a wedge-shaped tract of the main fontanelle ( fo.) is uncovered by bone (fig. 6). The notochord (nc.) is still present as a fine thread, and the basioccipital synchon- drosis is unaltered; moreover, the floor of the vestibule below, and the outer part of the capsule with the tegmen tympani, continue unossified. The semi-circular canals stand out strongly, and along the ridge of the posterior canal and part of the anterior (pr.o., au., ep.), the bone has undergone dehiscence, exposing the cartilage within, and partly separating the prootic from the ex-occipital on each side (fig. 6, pr.o., €.0.). In like manner, over the skull, the large frontals are partly severed from the small parietals (f, p.); the parasphenoid (pa.s.), below, has become smaller (actually), and more elegantly formed. The unossified auditory floor (fig. 7) is now a more transverse, sub-oval mass, and the projecting, descending lip of the floor is very small (Plate 11, figs. 7 and 8 ; and Plate 12, fig. 1, fty.). The upper part of the auditory capsule projects beyond the vestibular pouch, very considerably. The cranial cavity lessens in width gently to the fore end; the “tegmen” is of moderate extent over it, and part of it passes into the nasal region, as the roof of the ethmoid. The optic nerve (Plate 11, fig. 7, II. ; and Plate 12, fig. 1, II.) passes out of the hinder part of a large fenestra ; the roof-bones have a down-turned orbital edge. ») I, 4 76 MR. W. K. PARKER ON THE STRUCTURE AND The ethmoidal wall behind the nasal sac was articulated to the “ tegmen cranii” (Plate 2, fig. 1); that joint was lost in the last (Plate 11, fig. 1); now a new articulation has appeared, namely, of the post-palatine crest with the ethmoidal wall (Plate 11, fig. 6, pi.pa.); this, however, is merely at its upper part; the bar is continuous with the skull below (Plate 11, fig. 7, p.pg.). The direction of the endoskeletal palatine (“ post-palatine”) was directly forwards (A); the pre-palatine point being slightly turned inwards to join the trabecula by a ligament (Plate 2, fig. 1, pt.pa., pr.pa.). Afterwards (B, Plate 11, figs. 1-3), the pre-palatine point was well developed, and turned directly inwards to join the trabecula, but the main piece was parallel with the skull; the sharp crest had become an even thick ridge. Now (C, Plate 11, fig. 6; and Plate 12, fig. 1, pt.pa.), this part has behaved like a railway signal: it has turned outwards, almost directly, but is bent a little forwards ; the pre-palatine has again become reduced to a mere point carrying the ligament that runs to the cornu trabeculee. The flat, but crested, bar (A, Plate 2, fig. 1, pt.pa.) has become a massive, rounded rib of cartilage, standing high above the thin pterygoid plate that unites it with the suspensorium (Plate 11, fig. 6, p.pg.); thus that outer bar now stands off a good distance from the skull. The tegmen cranii has grown well over the perpendicular ethmoid, and the septum nasi has become surmounted by a pair of thick crescentic folds, lying back to back (Plate 11, fig. 6; and Plate 12, fig. 1, a/.sp.); from these the cartilage grows to sume little distance over the nasal sacs, but I cannot find, either now or afterwards, any such clearly distinct nasal roof as is seen in the species of Rana. The anterior part of the skull has become reduced in size and modified to a much greater extent that the outer bars; thus the quadrate hinges (q.) are now a sensible distance in front of the cornua trabecule. These latter parts are now very small and bifurcated ; the inner lobe is a short hook, (c.tr.) pointed inwards ; this will be the pro-rhinal, it lies in front of the nasal sae; the outer lobe is a long “ear” of cartilage (s.n./.) directed outwards and little forwards, and attached by ligament to the pree-palatine spur (pr.pa.). The position of the pre-palatine ligament shows the amount of lessening the trabecular cornua have undergone, for the ligament and this owter cornu, together, form the fore boundary of the internal nostril (i.n.). The last stage (Plate 11, figs. 1, 2, ¢.t7.) shows how much of the cornu there was then in front of that passage. The ethmo-palatines meet below (fig. 11, p.pg.); there was no such appearance in the last stage, before these bars had turned round (fig. 2); here, if the segmental nick (fig. 6), which is seen above, had extended between the post-palatine and the trabecula, it would have severed it from the skull, beneath which the right and left bars are confluent. The sub-cranial confluence is peculiarly Petromyzine. The huge suspensorium (sp., g.) is being folded up that it may be changed; huge DEVELOPMENT OF THE SKULL IN THE BATRACHIA. ie indeed now, it will soon become less than one-fourth its present size. Even now it is less than a fourth the size of that of the young Tadpole (A, Plate 2). Relatively to the cranium, it is immense, yet, and retains all the larval characters, being still confluent with the chondrocranium at three places. The pedicle (pd.) is now a very narrow band, and like the pterygo-palatine, is at right angles with the main bar. The “elbow” of the suspensorium has developed a new “ otic process” (ot.p.), and the spiracular band has become a mere thread (Plate 11, figs. 6, 7, sp.c.). As in the “ Urodeles,” the permanent “ otic process ” (ot.p.) mounts up against the fore edge of the tegmen tympani and the swellings caused by the ampullee : it reaches inwards, as in them, to the anterior ampulla, and even a little further. Above (fig. 6), it is a thick, rounded mass, but below (fig. 7), it is flat; it has, as it were, been thrust back against the auditory mass, and cleaves to, and lies outside, as well as in front of it. The greatly developed ridge on the edge of the suspensorium, instead of passing in a gentle are forwards to become the thickened edge of the orbitar process, turns suddenly as a round loop, and then, at an acute angle, bends back again round the front margin of the orbitar process, scarcely dying out as it approaches the condyle of the quadrate, in the front of the face; all this growth is ready to vanish away. This swollen upper selvedge is thickest behind, then narrows up to the apex of the orbitar process, and keeps its breadth until it dies out in front. The quadrate region and condyle (Plate 11, figs. 6, 7; and Plate 12 fig. 1, q.) is very broad and also thick, especially at its outer edge. In conformity with the divergence of the quadrate bars the mandibles (Plate 12, fig. 1, mk.) are considerably longer; but the lower labials (/./.) are full sized, as yet, and the condyle for the hyoid (hy.f.) is perfect. So also is the bar itself (c.hy.), but one can see that the styloid region (st.h.) is elongating. In the Common Frog and Toad there is no upper hyoid element until about three months after the loss of the tail; in Pseudis the tail has lost only two-sevenths in length, although much narrower, when that element (the epi-hyal or “columella ”) appears. Here, at any rate, this rod appears much earlier, relatively, than in the common kinds; it is possible, however, that the Tadpole, at this stage, may be several years old. In these large Tadpoles it is as easy to show that the early “stapes ” belongs to the periotic capsule as that this Jate segment does not. There are two cartilages, besides the stapes, in certain “ Proteidea” (e.g., in the Menopome), both inside the facial nerve. The upper piece sends its narrow proximal part to the stapes; the lower cartilage, manifestly part of the epi-hyal, becomes partly confluent with the hind margin of the suspensorium. Here things take a very different course, for the cartilage (Plate 11, figs. 8, 9) is undivided, and forms a rudimentary epi-hyal. 78 MR. W. K. PARKER ON THE STRUCTURE AND This new cartilage (co.) lies inside the nerve, as it forks to join the glosso-pharyngeal (VII., IX.), and to supply the region of the first cleft, fore and aft. This rudimentary columella appears as a solid bud of cartilage abutting against the fore edge of the large, almost circular, stapes (st.). A section of these parts (Plate 11, fig. 9, st., co.) shows that, according to the rule in the Batrachia, the dorsal end of this cartilage is wedged in between the antero-superior margin of the stapes and the auditory capsule; it is already constricted into two regions; the massive “inter-stapedial” part (a distinct segment generally, but not in Pseudis (see Plate 10, fig. 4, m.st.), is already marked off from the main rod (Plate 11, fig. 9, co.). Below the stapes (Plate 11, fig. 8, fty.) the tympanic floor, once so large as a down-growth of the actual skull-base, is now a mere thickening of the bulbous vestibular region. The premaxillaries (Plate 12, fig. 1, px.) have been added to the bony plates, and they fit against a small inner (anterior) cartilage (w./.); a crescentic valve to the nostril is also found outside this; the large primary upper labials have been removed ; they were ready to be absorbed. If the metamorphosis of this skull stayed here, we should have much to contemplate that is both striking and instructive ; this stage, however, is scarcely mid-way to the actual end of this changing skull. 17 (continued).—(D) Fourth larva of Pseudis, rapidly acquiring the Frog-form, but with a tail still 3 inches long and 7 lines wide. In this species the tail co-exists for a long while after many of the characters of the adult skull have appeared (Plate 12, figs. 2-7). At first sight this larva appeared to be not very different from the last, but it had lost the contracted swctorial mouth, and had gained the characteristic open gape of an adult Frog. This had been done by the rapid enlargement of some parts, and the rapid lessening of others—an interstitial change which, appears the more marvellous the more it is contemplated. In the broad basioccipital region the notochord (Plate 12, figs. 2, 3, ne.) still persists as a small thread; and a three-rayed tract of cartilage has escaped the ex-occipital erowths (¢.0.); the lateral extensions of cartilage are the occipital condyles (oc.c.), which are large and wide apart. The cranial cavity has become still wider than in the last ; it lessens gently up to the ethmoidal region. The parietals are now fairly distinct from the frontals (p., /), and the hinder part of the “tegmen” has no uncovered cartilage, the confluent ex-occipitals having the narrow part of the subquadrate parietals lying on their fore margin. These hinder bones send out a temporal angle behind the widest part of the frontals, and then the latter bones continue this down-turned edge (orbital plate) up to their narrow apex (fig. 4, f.). DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 79 Thus in a side view (fig. 4) the wall is about half of it bony, but this is due to the curling over of the roof; the ‘“sphenethmoidal” wall itself is quite unossified. The orbito-sphenoidal region is largely fenestrate (0.s.f.), as in Lizards, but in a very different manner. In this stage the fontanelle (fig. 2, fo.) is seen in front between the diverging ends of the frontals, and behind, as a lozenge-shaped space (fo’.) between the frontals and parietals. Laterally, the ear-capsules are very large relatively (compare Plate 11, figs. 1-3; and Plate 12, figs. 2-4); and the tegmen tympani with the enclosed horizontal canal (t.ty.) projects much in front; this part is not ossified. Below (fig. 3), the two unossified vestibules, embraced at their inner margin by the large grooved ex-occipitals, resemble a pair of symmetrically imbedded acorns, the rough “cup” being bony, and the smooth “ fruit” cartilaginous. Each of these capsules is fenestrate, and operculate, obliquely; the opercular stapes (st.) looking outwards and somewhat backwards; the rough “cup” two places for the glosso-pharyngeal and vagus nerves (IX., X.). In front, the bony matter scarcely reaches the 5th nerve (V.); the optic nerve (II.) passes out of the hind part of the orbito-sphenoidal fenestra, and the olfactory escapes through the closing is perforated in wall in front. The fore part of the tegmen cranii is rather long; laterally, it is articulated to the ethmo-palatine (e.pa.) ; in front, it passes into the two narrow crescentic nasal roofs (fig. 4, al.sp.), which are partly covered by the small ear-shaped nasal bones (n.). The nasal floor (fig. 3, s.7./.) is of the normal breadth, and in front it is now a finished structure, elegantly crenate, with seven lobes. The outer of these are the outer angles of the trabecular cornua, now finishing the nasal floor in front. Next to these, but on a higher plane, are the horns of the nasal crescents (s.7., al.sp.). Near the mid-line we see a pair of short inturned pointed “pro-rhinals” (p.rh.); and now, at the middle, there is a little bud of cartilage, the prenasal (p.n.). Besides the nasal roof-bones (n.), there are now, below, two oval dentigerous plates, placed transversely and some distance apart ; they touch the inner nostril (7.7.) by their outer edge ; these are the vomers (v.). The semi-oval outline of the face is now nearly finished by the dentigerous pre- maxillaries and maxillaries (fig. 4, pw., mx.) ; most of the maxillary bone is moulded on a cartilaginous bar, which is quite unlike anything seen in the last stage. The solid ethmo-palatine bar (Plate 11, fig. 1; and Plate 12, fig. 1, pt.pa.) is now (Plate 12, fig. 4, e.pa., pt.pa.) a slender rod, flatter below than above, and instead of being curved forwards outside, it is turned suddenly backwards. The prepalatine spike (fig. 3, pr.pa.), which turned directly inwards in the second stage and touched the trabecula (Plate 11, figs. 1-8) now has its outer margin coincident 80 MR. W. K. PARKER ON THE STRUCTURE AND with that of the whole suboccular bar (pterygo-palatine), whilst the point looks directly forwards. It forms now a pointed wing to the fore third of the crescentic palatine region ; the upper process of which (e.pa.) is half the length of the post-palatine (pt.pa.) which ends opposite the middle of the optic fenestra (IL, 0.s,f), whereas in the last stage it ended opposite the middle of the internal nostril (Plate 11, fig. 6). The last fourth of the pterygo-palatine bar (Plate 12, figs. 2-4, pg.), is the thin pterygoid foregrowth of the suspensorium, which was in the antorbital region in the last stage. The parasphenoid (Plate 12, fig. 3, pa.s.) retains its shape, but its point is more split, and these two spikes end in a sulcus caused by the meeting of the ethmo-palatines (e.pa.) beneath the ethmoid : their commissure is elegantly bowed forwards. A thin ectosteal plate (Plate 12, fig. 3, pg.) has applied itself to the “chondro- pterygoid ;” this latter part, although it is not segmented off from the post-palatine as in the genus Bu/o, is yet quite distinct from it to the eye, it being suddenly compressed ; the part in front retaining its thickness up to the meeting of the two regions. The suspensorium (e.pa. to q.c.) has shrunk to one-fourth the size it had in the last stage; it has lost the large leafy orbitar process; and the quadrate condyle which reached to the frontal wall of the head, now reaches nearly as far back as the “ fenestra ovalis.” . The hinge for the hyoid has also gone, and that bar (c.hy.), which did hang opposite the ethmoid, is now tied by a ligament (hyo-suspensorial), so far backwards that its styloid apex is directly below the middle of the auditory capsule. This position is attained by the Axolotl when it measures 43 lines—that is, soon after hatching (see “ Urodeles,” Part I., Plate 22, fig. 3, q., c.hy.). In all this change of size, form, and relation, the enlarged “ elbow” or otic process (ot.p.) has been the fixed point, and the quadrate has been thrust back by the reversed curve and increased length of the ethmo-palatine. In the last stage (Plate 11, fig. 7, pd.) the pedicle was becoming a mere thread of cartilage, and the inner and posterior edge of the lateral bar was becoming’ solid and rounded in front of the groove where the facial nerve emerges. Now (Plate 12, fig. 3, pd.), the attenuated pedicle has become fibrous proximally, and the outer part has developed itself into a thick pedate mass, the concave side of which fits against the convex fore face of the unossified wall of the vestibule. This ‘condyle of the pedicle” is a process growing from the wnder, as the otic process is an outgrowth from the upper, part of the suspensorium. Over the outer surface, from which the orbitar process has vanished, a thin straggling tract of membranous bone now lies; this is the squamosal (sq.), with all the characters, as yet, of a “ preoperculum.” The little bridge of cartilage which arched over the facial nerve (Plate 11, fig. 8, VIL.) is now once more free (Plate 12, figs. 2, 4, 6, 7, a.ty.); it is now an elegant DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 81 ear-shaped flap, notched below, and in front sending down a membranous tract, which will soon become cartilage. The broad part of this rudimentary “annulus” lies on the swollen unossified tegmen tympani, covering the ampulla of the horizontal canal (figs. 6, 7, h.s.c.), and forming a bridge over the 7th nerve and the tympano-Eustachian cleft, which opens inside, between the suspensorium and the condyle of the pedicle (Plate 12, fig. 3). In the recess beneath this double roof the epi-hyal element (columella) has now become almost normal in form, but is not yet ossified (figs. 4, 6, 7, co.). Although relatively sc much smaller, it is no hard task to harmonise this element with the hyo-mandibular of a Skate, which is developed quite independently of the cerato-hyal, and perchance a little after it in point of time. There is but one piece of cartilage, for the inter-stapedial segment is not distinct ; this part is broad, oblique, and wedges in between the ear-sac and the stapes (st.), now a large lozenge-shaped cartilage. The extra-stapedial region is bent downwards, it is ligulate, and retains the breadth of the medio-stapedial region ; there is no supra-stapedial process. The mandible has more than doubled the length it had in the last stage (Plate 12, fig. 1, mk.) ; the angular process is reduced greatly, aud the loosely swinging hinge is quite unlike that of the larva. The inferior labial has coalesced with the distal end of MrckeEt’s cartilage (mk.), but the dentary (d.) has not yet converted it into a ‘mento-Meckelian” bone. The proximal broad end of the lower jaw lies in a thin bony trough, the “articulare” (a7-) ; this is developed most on the inside. The cerato-hyal (c.hy.) is now attached by the “hyo-suspensorial ligament ” (fig. 4, h.s.1.) to the inside of the suspensorium at its lower third; its styloid region (fig. 4, st.h.) is somewhat uncinate, and its hypo-hyal region (4.hy.) decurved; the basi-hyal has vanished. A single rudiment (fig. 5, ex.br.) on each side remains of the four inner and four outer branchials ; this plate embraces the basi-branchial (b.br.) which is wider than long, and has a crescentic emargination in front. The pair of hypo-branchials (h.br.) are not much altered, but their hinder part, which sends out rudiments of a third and fourth cerato-branchial (Plate 11, fig. 4, ¢.br., 3, 4), has now become a solid “ horn,” ready to ossify as a “ thyro-hyal” (é.hy.). This is a very instructive stage of the hyo-branchial apparatus ; if it be compared with the larval structures (Plate 10, fig. 6; and Plate 11, fig. 4), and with those of an adult Frog (see in Rana pipiens, Plate 8, fig. 5), we shall see what becomes of the copious cartilaginous growths of the larval branchial arches. Before leaving this stage [ have to remark upon the evidence these changing larval skulls supply of the existence of a pre-oral visceral arch in the ethmoidal region. Years ago my study of the development of the skull in Birds and osseous Fishes, besides what I saw in various kinds of adult Birds, left me in a state of mind that scarcely admitted a doubt upon the subject. MDCCCLXX XI. M 82 MR. W. K. PARKER ON THE STRUCTURE AND The skull of the Tadpole, evidently a most generalised structure, covered the whole question with a cloud of doubt. Since then, whilst wavering, I have collected copious evidence of what seems to me to be a true second pre-oral rudiment in the antorbital region, and a first or terminal rudiment in the front of the face. I have already shown (“ Batrachia,’ Part II., and ‘‘ Urodeles,” Part I., and in my papers on the “Selachian and Avian Skulls”) the very constant occurrence of an ethmo-palatine in several large natural groups of Vertebrata. Here, in the larva of Pseudis, 1 have shown this part partly segmented off, and when not segmented off, it is yet very distinct, and not to be misunderstood; it assumes, in fact, four positions, each of which is like what is seen in the members, generally, of some large group or groups. This large and most remarkable conjugational bar is parallel with the skull in the two first stages (A and B, Plates 2 and 11). So it is in the Siluroid Fishes (eg., Clarias capensis), where the palatine bone formed by the ossification of an autogenous cartilage (in these and other Teleostei) remains distinct, but has no “ ethmo-palatine” or ascending process, as in the Salmon. In the third stage (C) it is bent outwards and forwards; this is the natural form and position of its independent homologue in a large number of the “ Urodeles.” It afterwards turns round; but in doing this there is a time in which it is directly transverse, or at a right angle to the skull, as in Menobranchus. In Protews anguinus, where it is not drawn forwards by a chondrified nasal capsule, it turns backwards, and so it does also in Notidanus, in the Skates generally, and in many Birds. In the Salmon, as in the adult Toad, the ethmo-palatine has its three regions well developed—an ascending, an anterior, and a posterior part; all these are well seen in the stage of Pseudis just described. Such a rudiment of the ethmo-palatine as exists in the adult of many Urodeles, where it has coalesced with the back and lower part of the nasal capsule, is seen also in the “‘Sauropsida” (e.g., Chameleo, Dromaus, Casuarius, and Struthio); whilst in many Birds it is separately ossified, and forms, as the ‘‘os uncinatum,” a most characteristic endoskeletal bone in the fore-palate. The foremost pleural rudiment is less widely distributed as the “ pro-rhinal,” or “recurrent trabecular cornu.” I should not be surprised if it turned out to be the serial homologue of the mandibular suspensorium, of the epi-hyal, and of the epi- branchial elements.” * T have already mentioned (p. 18) that Professor Huxtey and Mr. Batrour do not take the same views of these parts as Professor Mityus Marsuatu and I do. = a a DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 83 17 (continued).—(E) Skull of adult Psewdis paradova.—Old male ; 24 inches long. Surinam. (Hyrrt’s prepn. in Mus. Coll. Surg., Eng.) The specimen whose skull is here figured was evidently a very old individual ; it is therefore of great value in this series.* It is at once seen to be a very generalised type (Plate 10, figs, 1-4), sharply separated from a normal Frog, such as Rana temporaria. The skull is one-half the length and a little more than half the breadth of that of the perfect larva. The skull of the nearly metamorphosed larva (D) is five-sivths the length of that of the adult : so that after lessening much, it enlarges again, a Little. About the size of the skull of a large Common Frog, it is more elegant in shape, being a very perfect semi-ellipse in outline. The well-bent bow formed by the two series of cheek bones has its “horns” meeting at a sharpish angle, for the parts in front are modelled on three trabecular outgrowths, the “pro-rhinals” and a small “ prenasal.” The complex structures that run across, behind, from one cheek to the other, are also very beautiful in their construction (Plate 10, figs. 1-3). The hind part of the skull is densely ossified from the occipital condyles to the optic foramen (oc.c., II.), and in front the ethmoidal girdle-bone (eth.) occupies the hind half of the nasal capsule below, and all but the edge in front; the parts in front of that, and the anterior sphenoid (0.s.), are unossified. The occipital condyles are rather wide apart, being separated by a large semicircular emargination ; they are hemispherical and subpedunculate, as in those Urodeles that have an intercalary “ odontoid” vertebra. Below (fig. 2), the auditory masses seem to stand out in a directly transverse line ; but above (fig. 1), they are seen to turn forwards as well as outwards, and the well- marked semicircular canals (between pr.o. and e.o.) are some distance from the edge, which is flanked by the squamosal (sq.): the intervening part is the ossified “ tegmen tympani.” The gently rounded floor of the vestibule (fig. 2) shows no trace now of that Selachian development of the basal plate, outside the capsules, which formed the tympanic floor. The main nerve-passages (fig. 2; II., V., 1X., X.) are well seen from below ; the two hinder passages are separated by a bony bar ; the ‘foramen ovale” is enclosed in bone, and the optic fenestra is partly margined, also, by the extended alisphenoidal wing of the prootic (pr.o.). The foramen magnum has less roof than floor; neither of these plates has any divisional line as in a normal Batrachian. * It is a most fortunate thing for me that our College has possessed itself of this invaluable specimen ; it is not an easy thing to get an adult. M 2 84 MR. W. K. PARKER ON THE STRUCTURE AND Nor has this generalised growth (both of cartilage and bone) allowed any deficiency of roof below the parietal region ; there is only one, the main, fontanelle : this space is barely covered by the shortish frontal end of the roof-bones (/-p.). The roof-bones are completely re-united, and the frontal suture is partly closed, as well as the whole of the sagittal; the skull is sub-suleate towards the mid-line (fig. 1.). These roof-bones leave much of the endocranium uncovered behind as well as in front ; a little temporal wing, and a moderate orbital plate is formed by them. The parasphenoid retains the form it had in the large larva (Plate 2, fig. 2, pa.s.) ; and it is very characteristic of Pseudis, being a dagger, gapped at the point, with a wide elegant “guard,” and no “ handle.” The large girdle-bone (eth.) extends its osseous substance half across the cheeks, half along the internasal region, and it trespasses on the orbito-sphenoidal headland. Together, these elements make the interorbital part of the cranium (its anterior two-thirds) a boat-like structure, gently narrowing forwards. Then, on a sudden, it widens again, passing out into the palatal and nasal structures. Above (fig. 1, .), the elegant conchoidal nasals cover much of the roof, and run down over the ethmo-palatine wings, but they do not hide the middle over the front extension of the creeping ossification from the girdle-bone ; above (fig. 1), the soft pre- nasal (p.n.) finishes the face. The ectosteal palatine (fig. 2, pa.) binds under the half-ossified ethmo-palatine ; it is falciform, with the narrow handle inwards, far from that of its fellow of the other side. Hiding the semi-osseous sub-nasal plate, except in front and in the middle, are the two vomers (Plate 2, v.), they are sub-reniform, with a lateral spike ; behind, they are massive and dentigerous: this lenticular part is like what is seen in several Frogs (both “ Platydactyla” and “ Oxydactyla”) from the Notogec. The internal nostrils (fig. 2, 7.n.) are large and round ; the external nostrils (fig. 1, e.7.) are moderate in size and are guarded by two valvular cartilages, the inner and outer upper labials. The are of cartilage running from the palatine to the pterygoid becomes thin, but I cannot find that it is segmented; the latter bone (pg.) is pointed in front, then becomes very broad, and this wide part forks into the sheath of the pedicle (pd.) and a posterior process. The posterior process becomes sub-vertical, binding the inside of the suspensorium to the condyle of the quadrate (¢.c.). The inner fork arches inwards and enclothes the stunted pedicle ( pd.) up to the condyle, which glides on a pad of cartilage on the fore-face of the auditory capsule (fig. 2.). DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 85 In the rounded angle between these two forks we see the large, oval, oblique, Eustachian opening (ew.); this is enclosed behind, by the geniculate stylo-hyal (st.h.). The suspensorium forms a very obtuse angle with the axis of the skull, and its condyle reaches to a transverse line that cuts the neck of the occipital condyles. The hinge (q.c.) is a bilobate condyle, the inner lobe being the larger ; the substance of the cartilage above the condyle is ossified as a quadrate bone, by the engrafting on it of the “ quadrato-jugal ” (q,.). A tract of cartilage can be seen outside the pterygoid (fig. 3, pg.), and this may be traced into the pedicle, and to the borrowed quadrate centre. The backward position of the suspensorium is the cause of the very tilted position of the squamosal (fig. 3, sq.), the lower part of which, partly hidden by the ear-drum, runs backwards as well as downwards ; the supratemporal part (fig. 3, sq.) is very long and sigmoid, for it sends forwards a sharp out-bent postorbital process. The premaxillaries (px.) are large and well formed, having a well-defined dentary margin, a triangular palatine process, and a high nasal process, capped, inside, by the inner upper labial. . The dentary edge of the face is finished by the maxillary, the teeth ending opposite the Eustachian tubes, and the bone opposite the middle of the auditory capsules. There are small “septo-maxillaries (s.mz.),” and the maxillaries (m«.), notched in front, run well up to the premaxillaries and the nasals ; the jugal part is high. The quadrato-jugals (q,j.) are short, high, and are grafted upon the quadrate region of the suspensorium, as aforementioned. The cartilaginous mandible is placed obliquely in its “articular” trough (a7.), that bone lying mainly below, and on the inside ; the dentary (d.) has formed a “ mento- Meckelian” bone (fig. 3, m.mk.), by ossification of the end of the rod ; once the free inferior labial ! The stapes (Plate 10, fig. 4, st.) is thick and reniform; for the antero-superioy edge is sinuous to admit of the large dorsal part of the “columella,” between it and the capsule. The inter-stapedial end of the columella is thin and clawed below; the upper part is large and blunt ; by these it holds the stapes, as it were. The “ medio-stapedial ” bone (m.st.) is dilated where it runs into these spurs, and then runs as a straight rod up to the extra-stapedial (¢.st.). This latter part is at first no thicker than the end of the bone from which it arises; it turns downwards, and soon enlarges into a tranversely oval disc. There is no “supra-stapedial,” even as a membranous band, and the interstapedial also not being segmented off, this columella is much below the normal condition. The “annulus tympanicus ” (@.ty.) is large, and its horns nearly meet above. The stylo-hyal end of the cerato-hyal band (st.4.) has coalesced with the auditory capsule, a moderate space from the front of the fenestra ovalis ; it then turns directly 86 MR. W. K. PARKER ON THE STRUCTURE AND outwards, finishing the Eustachian rim (ev.), and from thence passes downwards and forwards to the basal plate, which differs but little from the norma.” Second genus, Gomphobuates, 18. Gomphobates (Leiuperus). ? sp.—Adult (7) ; 10 lines long, River Plate. This small Frog was, apparently, adult, and yet so arrested in certain respects, as well as so very small, that it answered to the young of the Common species taken about the end of July. The chondrocranium of this kind is so very instructive that the figures are only made to show this, and certain of the bony plates attached to it. The general outline of the skull (Plate 13, figs. 1, 2) is semi-oval and the breadth is a little more than the length. The notochord (nc.) is still large, and lies upon the united moieties of the basal plate. The occipital condyles (oc.c.) are normal; and the emargination between them moderate; but the roof has a much larger crescentic notch than usual over the foramen magnum (f.m.). The auditory capsules are like those of a young Common Frog in general condition, but they have a peculiar character seen in metamorphosing larvee of Pseudis (Plate 11), viz.: a floor to the tympanic cavity projecting beyond the fenestra ovalis and stapes (fig. 1, st.). The tegmen tympani is of very small extent, and only exists in front. The 9th and 10th nerves (LX., X.) are surrounded by a very limited ex-occipital (e.0.), and the prootic (p7.o.) only forms an oval patch behind the foramen ovale (V.); this is entirely on the lower aspect of the skull; the ex-occipitals just reach the upper surface between the ear sac and the occipital arch. The interorbital region is rather broad, a little bulging, and considerably narrower in front than behind ; it is well walled in with cartilage, the optic fenestra (II.) being of the average size ; and the top wall grows over on to the roof for some distance, leaving one oval fontanelle (fo.) about half as large as the whole roof. The anterior part of the tegmen cranii is well developed, and the hind part is nearly as long as the fontanelle. The ethmoidal wings end abruptly above, and articulate by a flat facet with the ethmo-palatine (e.pa.), but are quite continuous with it below; there is no rudiment even of the girdle-bone. The figures given of the upper and lower aspects of the ethmo-nasal region, might * The skull of the adult Pseudis differs from the “norma” in many things, viz. :— . Intense ossification of endocranium. . Sub-pedunculate occipital condyles. No secondary fontanelles. . A prenasal rostrum. . Solid dentigerous lobe to vomers. No proximal segment or distal process to columella. . Hind edge of parasphenoid emarginate. . Extreme solidity of pedicles. OOTIA oP wD . Ossification of quadrate region. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 87 be taken as typical diagrams of those parts in the “ Anura” generally ; for although coalesced, they show the outlines of all the parts that go to make up the whole (figs. 1, 2). The tegmen cranii in the ethmoidal region is continuous with the top of the large high intertrabecular crest, the outline of which is clearly seen, both above and below, as a gradually narrowing tract of cartilage, convex above and below, and ending in a free, blunt spike. Here we have the perpendicular ethmoid passing into the septum nasi and ending in the prenasal rostrum (p.7.). In front of the ethmo-palatine bars the trabecule have a large crescentic notch on each side ; here the inner nostrils (.n.) are situated, yet at this part the latter are not much narrower than in the interorbital region. In front, they spread into wings three-fourths the extent of the auditory masses, behind ; and each wing is divided by a large rounded notch into a large hind, and a lesser fore, lobe. The hind lobe, or angle of the subnasal lamina (s.7./.) forms a rounded hook in front, which turns inwards; on the upper surface (fig. 1) each plate has a transverse crest, which runs inwards and articulates with the out-turned, anterior horn of the nasal roof-cartilage (na.). The anterior lobes of the primary trabecular cornua are nearly as long, and one-third the width of the hind lobes; they are a little arcuate, finger-shaped, and diverge at more than a right angle; these are the pro-rhinals (p.rh.). We have, thus, a five-fingered end to the chondrocranium, and these five lobes are all developments of the pre-pitwitary on-growths of the three basi-cranial cartilages that in the post-pituitary region embrace and enclose the notochord.* Over the nasal sacs, the cartilaginous roof (a.) is seen to be composed of a pair of ear-shaped shells, lying back to back, not touching each other, but confluent with the edges of the septum nasi. Behind, they have also united with the ethmo-palatines (e.pa.), and in front with the transverse ridge on the outer trabecular lobe (s.n.l.). The inner margin of each is semi-circular, the outer is sinuous, widest in the middle, and each end is developed into a cornu; the front horn is blunt and the hinder horn sharp. On the front horn there is seen the second upper labial (w./°.) (the jirst has been left unfigured with the pre- maxillary) ; between this and the front horn is the outer nostril (e.n.). * Yet the true skeletal avis ended behind the pituitary space and body, and the organic fore end of the neural axis ended just in front of that part, through the overbending of the vesicles of the brain. The continuous, or distinct, conjugating bar by which the fore-growth of the mandibular pier becomes attached to the ethmoid, bears an endoskeletal relation to the outer bones in the malar region. The angle of the “‘ subnasal lamina’”’ bears the same relation to the mavillary region, and the pro-rhinals to the inter-mawillary region, and these pre-oral endoskeletal and exoskeletal structures, combined, make up one antagonistic upper jaw, to work against the post-oral lower jaw. Moreover, there are, besides these three pre-oral rudiments (or imitative arch-piers), the “ post-palatine ”’ “ cartilage of the Axolotl, the “transpalatine” of Singing Birds, and the “epiterygoid” of Urodeles, Chelonians, and Lizards, to be accounted for. These three latter, I suspect, are different modes of the same element. 88 MR. W. K. PARKER ON THE STRUCTURE AND The palato-suspensorial arches and their relation to the nasal region are like those of the fourth of my stages of Pseudis (see Plate 12). The “ ethmo-palatine” (e.pa.) is a thick transverse bar, whose flattened upper end articulates with the ethmoidal wall. The “ pre-palatine ” (p.pa.) is split into two laminz, the outer one is rounded, and the inner and upper lobe is sharp; the outer runs into a second spike, turning inwards on the transverse bar (fig. 2). The “post-palatine” (pt.pa.) now forms a crescentic crest on the top of the far-extended pterygoid cartilage (pg.), where it turns inwards, in front. The former was free behind ; and its new and fixed position, as compared with its original condition, is well seen by comparing my third and fourth stages of Pseudis (Plates 11, 12). The rest of the arch keeps its even size as a rather wide band, until it forks; the short inner fork is the pedicle (pd.), the longer outer fork is the quadrate and its hinge (q., q-c.), which reaches to a line crossing the middle of the stapes (st.). The ” palatine and pterygoid bones (pa., pg.) are normal: the former looks like the bone of an old individual; the latter is thin and arrested. The Eustachian openings (ewv.) are large and turned obliquely outwards and for- wards ; the stapes (figs. 5, 6, st.) is an elegant oval, and has wedged inside its front margin a much smaller oval cartilage quite distinct—-the interstapedial hind part of the columella. The main part of the columella (co.) is not developed into regions ; it is long, sinuous, and finger-shaped, with a crest near the stapes: but for the separate interstapedial, this columella would have corresponded with that of Pseudis at its fourth stage (Plate 12, fig. 6). The narrow “annulus” (a.ty.) is arrested ; it is a knee- shaped strap of cartilage bent forwards. The mandible and hyo-branchial plate (figs. 3, 4) correspond with those of a very young Frog; the mento-Meckelian is not finished, and the cerato-hyal (c.hy.) runs short of the ear-capsule; there is no hypo-hyal lobe, and only the hinder part of lateral lobes. Nearly all the investing bones have the same arrested character; the paraspbenoid (fig. 2, pa.s.) is, however, very large; the vomers (7.) are small, angulated crescents, without teeth. The fronto-parietals fail to cover more than half of the great fontanelle; this cor- responds to what has been found in other species of this genus.* The deflection from the norma seen in this small skull is very largely due to arrest, on account of the small size of the species, and perhaps also to a somewhat immature state. These deficiencies, as compared with the skull of an adult of the common kind are :— * Professor E. D. Cope—“ Eleventh Contribution to Herpotology of Tropical America,” Amer. Phil. Soc., June 20th, 1879, p. 264—says that Leiuperus (a synonym of Gomphobates) has an open fontanelle. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 89 The small size of the ex-occipitals and prootics. ° No girdle-bone. The fontanelle single and largely uncovered. The evident outlines of the elements composing the nasal region. The presence of a prenasal rostrum. The articulation of the ethmo-palatines with the ethmoid above. The rudimentary condition of the annulus and columella. 8. The arrested state of the mandible and simple condition of the hyo-branchial structures. 9. The absence of teeth on the vomers. Sale St ee CO) Third genus. Cystignathus. 19. (A) Cystignathus ocellatus (?).—First larva. Brazils.* First larva; 2§ inches long; tail 14 inch ; hind legs not, visible. This is the most immature of the Batrachian larvee of which I have as yet given illustrations in this memoir; I have already referred the reader to much earlier stages, figured and described in my published papers. These skulls are relatively very solid, and what strikes the eye at once is their oblong form and the straightness and width of the cornua trabeculz. The chondro- cranium (Plate 17, figs. 1, 2) is quite complete, and it has passed from the simple primary condition by having had three osseous tracts applied to its surface (f'p., pa.s.). Except for the projecting cornua and labials, the skull up to the quadrate condyles (q.c.) is very evenly oblong, and only one-tenth longer than broad. The occipital con- dyles (ve.c.) project but little beyond the auditory swellings; the halves of the basal plate are separated by a considerable notochord (nc.), and the cartilage which unites the basal plates with the periotic capsules is young and crowded with corpuscles. The tegmen cranii, both fore and aft, is also made of young cartilage, and its boundaries are traceable; it leaves one large fontanelle (fo.), which is pyriform in outline, with the stalk behind. Up to the middle of the internal nostrils (7.m.) the intertrabecular plate has conjugated the trabeculze (¢7.); from thence the massive cornua (c.tr.) diverge gently. Behind, the canals of the ear (fig. 1, a.s.¢., 1.s.¢., p.s.c.) * These were taken with the species here named, and presumably belonging to it, from Rodsio (a tributary of Rio des Macacos) above the Falls, Brazils, May, 1865; they are the gift of Prof. A. Agassiz. I shall treat of these as belonging to the type named, with the above explanation, as also the skull of another, much smaller larva, with no title, but only the locality from which it was taken, namely, Lake Jannarg, Manaoo, Brazils. Whichever genus the larger Tadpoles belong to, to that also the little larva belongs; these two species have larval skulls very distinct from what is seen in Rana, Calyptocephalus, Cyclorhamphus, the Hylidw, or the species of Bufo; they are most like those of the earlier stages of Pseudis. Merely as a study of a variety of the larval Batrachian skull they are very valuable; and I trust that some friend will verify their title for me, and if there be any error, correct it. MDCCCLXX\XI, N 90 MR. W. K. PARKER ON THE STRUCTURE AND shine through the thin walls of the sac, which is a small spheroid, growing into a free edge on the outside, above. That “eave” of cartilage is the tegmen tympani (t.ty.), behind, and the confluent spiracular cartilage (sp.c.), in front; this ear-shaped flap just touches the round elbow of the suspensorium—its otic region (ot.p.). The infero- lateral face of the capsule is splitting to form the “fenestra ovalis” (fo.), but the stapes is not yet formed. Contrary to what was seen in the last two kinds, the inter- orbital or fore half of the cranium proper enlarges from behind forwards; it also bulges a little at the sides. There are scarcely any “ethmoidal wings” (al.e.) for the post-palatine rudiment (pt.pa.) lies close to the skull, and the pterygo-palatine band (p.pg.) is very short ; it turns forwards and outwards, and is surmounted by the thick, oblong post-palatine. In front of the internal nostrils (¢.7.) the trabecule (c.tr.) do not grow towards the pree-palatine rudiment (pr.pa.) as in the last two kinds; they only, there, suddenly become wider. The sides of the ethmoid (a/.e.) can be seen to be older than the tegmen (fig. 1, tr.) between and over them, and this latter part covers the rudi- mentary ‘“ mesethmoid.” The gently diverging cornua (c.f7.) increase a little in size up to their fore end; but they do more than this, they are convex above, turn gently downwards, and at their outer edge curl over; this fold thickens up to the end, where it is very solid ; it there articulates with both the upper labials of that side (w./‘.w.l’.); this is very similar to what is seen in Psewdis (Plates 2 and 11), where the upper labials are not divided. The falcate hinder part of the suspensorium (sp.) unites by its hooked end (pd.) with the basis cranii, and this “pedicle” passes in between the orbito-nasal and vidian nerves (V’., VII’). The sub-oval condyle for the cerato-hyal (hy.f-) is inferior, and projects from the root of the large orbitar process (07.p.) which nearly touches the skull-wall; it is only separated by a thinnish tract of fibrous tissue. Its ribbed edge is decurrent along the suspensorium, fore and aft, reaching nearly to the condyle (q., q.c.) and nearly to the ear. The pre-palatine rudiment (pr.pa.) is an upturned “ ear” of cartilage, giving attachment to the ligament that bounds the inner nostril in front. The quadrate condyle (q.c.) is saddle-shaped, and looks somewhat inwards, as well as forwards and downwards; it has a round knob of the suspensorium outside it. The upper labials (v./u.l’.) are formed (as cartilages) separately ; the inner pair are sub- crescentic, the outer pedate,—the “toe” being the postero-external angle. The mandibles and lower labials (fig. 3, mk., 1.7.) are of the same form as in Pseudis ; they are short, massive, and obliquely articulated together. The other inferior arches (fig. 4) have their own characters. The cerato-hyal (c.hy.) is remarkable for the hollowing out of its upper part, which is a curved shell of cartilage; the condyle (Hy.c.) is oval, and there is a spike in front of it. The basi-hyal (b.hy.) is composed of simple cartilage, the basi-branchial (6.br.) of hyaline, and it has a budding second joint. The large hypo-branchials (/.br.) are wide in front, and gradually narrower behind, where they are notched. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 91 The extra-branchials (ev.br.*) are quite normal, and each has its own cerato- branchial rudiment (c.b7.'*) coalesced below with it. The fronto-parietals (fig. 1, fp.) are small blades of bone, pointed in front, broader behind, and resting on the roof-edge of the skull. The parasphenoid (fig. 2, pa.s.) is very noticeable ; it is U-shaped, and half cleft into two pieces ; this character attains in this, and in its congeneric larva, to an extent not seen by me in any other type of the “Ichthyopsida;” a front notch is seen in every stage of Pseudis, but it is small. Most of the cranial nerves (fig. 2, II., V1, V*%, VIL, VII’, IX., X.) are shown issuing from their foramina, as in the last kind of larva. 19 (continued).—(B) Cystignathus ocellatus (?).—Second larva; 34 inches long; tail, 24 inches ; hind legs, 1 line long. Brazils. In development, this Tadpole corresponds very exactly with my huge youngest larva of Pseudis (Plates 1 and 2); the hinds legs are relatively of the same size and are at the same stage. The chondrocranium (Plate 17, figs. 5, 6) is one-half larger than the last, and certain changes have taken place in it of importance. The notochord (nc.) is shrinking, and the basal plate closing in upon it; the occipital condyles (oc.c.) are better formed, and the cranial cartilage has fused more completely with the auditory capsules. A dis- coidal exoccipital encloses the 9th and 10th nerve (e.0., IX., X.), and the prootic (p7.o.) is beginning to creep up the front of the auditory capsule behind the foramen ovale. The canals of the ear (a.s.c., h.s.c., p.s.c.) are very clearly seen above; below, the capsules are getting flattened sides, which lessen backwards ; in the side, under the large leafy tegmen tympani (t.ty.) the fenestra ovalis has become almost vertical, and a nucleus of cartilage is now to be seen in the soft plug of tissue that fills it; this is the stapes (fig. 6, st.). The tegmen tympani is elegantly angulated behind; in front it is continuous with the spiracular cartilage (sp.c.), a large sub-falcate flap, which has coalesced also now with the budding “ otic process” (ot.p.). The interorbital region of the skull is more perfect, but of the same shape; the pear-shaped fontanelle ( fo.) is widest, now, further back, and the front tegmen is more solid, and has united with the trabecular edges and hidden median cartilage to finish the ethmoidal region, The olfactory nerves (I.) are seen emerging, but the nasal roofs are still membranous, and were removed in the preparation. Yet the intertrabecula is filling in the space between the trabecule up to the membranous margin of the internal nostrils (?.n.). That membrane, however, has in it no spur growing from the trabecular cornu (c.t7.), which is narrowed further forwards than in the last stage ; for the rest, these, and the upper labials (w.l*1.), are only larger. The pedicle of the palato-suspensorial arch ( pd.) is less transverse ; the orbitar process (or.p.) is further out ; the post-palatine rudiment (pt.pa.) is now fairly differentiated from the ethmoidal wing («/.e.), now N 2 92 MR. W. K. PARKER ON THE STRUCTURE AND manifestly seen to be an external crest of the trabecula, a continuation of the cranial wall, but in front of the cranial cavity. For the rest, the quadrate region and its condyle (q.), the mandibles (fig. 7, mk.) and the lower labials (/./.) have merely increased in size since the last stage. The same may be said of the fronto-parietals (f.p.); they are a little straighter, however. But the parasphenoid (fig. 6, pa.s.) has been transformed into the normal shape, or nearly ; the notch is now relatively much smaller ; its shaft is long and wide also ; the basi-temporal wings are now developed, they are pointed, raggedly lobate, and meet in a triangular “handle.” 19(continued).—(C) Cystignathus ocellatus var—Adult male; 54 inches long. Dominica. This large sub-typical Frog has a skull (Plate 16, figs. 1-5), whose length is to its breadth as 7 to 84; it is a very regular half-oval in outline, and the condyles for the lower jaw reach, very nearly, as far back as those for the atlas. Being of great breadth, the auditory regions are extended like outspread arms, the parotic processes being as wide as the capsules proper, outside the horizontal canals. The extent of the ossification of the endocranium is normally Ranine, and there is clear tract of cartilage above and below the foramen magnum (fm.), and the ex-occipitals and prooties (¢.0., p7r.o.) are distinct. The occipital condyles (oc.c.) are large, reniform, postero-inferior, and separated by a space less than their own size; the occipital arch itself is of great width, and well marked off from the auditory capsule, especially above. This is due to the great size of the epiotic eminence (ep.), caused by the arch of the posterior canal; this eminence is unossified. So also is the outer half, below, and the outer third, above, the prootie not reaching, even there, as far as to the squamosal (sq.). Thus the whole of the extensive tegmen tympani (¢.ty.) is unossified. Besides the high epiotic region another ichthyic character appears (seen to a less degree in Rana pipiens, Plate 8), viz.: an extensive “ pterotic ” crest (pt.o.); this is a backward and outward sigmoid process of the tegmen tympani, which doubles the extent of the margin and overhangs the large annulus tympanicus (fig. 3, a.ty.), behind. The prootics (pr.o.) reach as far as to the middle of the space between the 5th and 2nd nerves (V., II.); the latter lie in a fenestra which reaches half-way from the prootic to the girdle-bone (eth.). The interorbital region widens steadily from behind forward, it is rather hollow above and below, for the side view (fig. 3) shows a concave outline both to the roof-bones and the parasphenoid (/f-p., pa.s.). This enlargement forward is due rather to the growth of an “eave” over the fore part of the orbit than to any great increase in the size of the cranial trough within ; but for that superorbital expansion the cranial trough would be seen as somewhat spindle- shaped, bellying in the middle and narrower at either end. The outer edge of the superorbital expansion is unossified, and in its fore edge the orbito-nasal nerve (fig. 3) DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 93 enters. The girdle-bone (eth.) is three-fifths the size of the orbital region and stops exactly at the fore edge of the true ethmoidal region: it does the same right and left. In the smaller species (Plate 16, fig. 6) a pair of secondary fontanelles exist, and I have no doubt of their existence in the large kind also. The great fontanelle is not very large and is covered in well by the roof-bones (/‘p.). That which strikes the eye at once in this skull is the tumid shape of the whole nasal region, which is large in every direction, along, across, and from top to bottom. It is well roofed by cartilage (aln.), but the floor (fig. 2) is still more extensive; in front, the snout is broad and gently convex, not produced, but rounded in the middle and rising well over the openings (e.7.). Below (fig. 2), the growth of cartilage is unusually extensive, the angles of the subnasal laminze (on each side of s.n.) are large and faleate; the pro-rhinals (p.7h.) are long, wide, and somewhat decurved bands ; the wall of the nasal capsule (fig. 3, n.w.) is well developed. But the greatest modi- fication is seen in the large size of the ethmo-palatine arms (e.pa.), which however have no definite pre-palatine spike (pr.pa.). The bony bars correlated with these arms are also very large; they are falcate, with the point growing back under the post-palatine cartilage (fig. 2, pa.). Here we have also what is seen in several Toads, viz. : an additional semi-distinct bony bar, sharp and denticulated, stuck to the lower surface of the main bone at its middle. The pterygoids (pg.) are slenderer than the palatines ; they have used up the feeble cartilaginous tract to which they were applied; they fork in front of the Eustachian opening (fig. 2, ev.) at a right angle, and thus that large space is itself triangular. The part covering the pedicle ( pd.), and faced with unaltered cartilage, is very long and pedate, for this point is not far outside the foramen ovale (V.). The hind part of the pterygoid faces the inside of the strong suspensorium, whose condyle (qg.c.) is a large and well made trochlea. There is some deposit of bone in the quadrate region from the quadrato-jugal (q., ¢,j.). The annulus (figs. 1 and 3, «.ty.) is large and perfect ; the stapes (figs. 1, 2, 5, st.) is oval, umbonate, and cut away in front; the columella has a condyloid cartilaginous proximal end without a segment ; and the strong bony bar (m.st.) has two projecting “trochanters.” The extra-stapedial (e.s¢.) is tongue-shaped, and its ascending process (s.st.) is confluent above. The mandible (fig. 3) is strong; the condyle (a7.c.) is large and reniform ; the coro- noid crest (a.) is not high; the dentary (d.) is only half as long as the ramus; the mento-Meckelian (m.mk.) is well developed. The stylo-hyal (st.h.) is confluent above, the rest of the band (fig. 4, c.hy.) is narrow up to the hypo-hyal lobe (i.iy.) ; the notch is very large, leaving a basal plate (b.h.br-.) of small extent at the middle; the fore side lobes are very large, elegant, and reni- form; the hind lobes are narrow arcuate bands; the thyro-hyals (¢.4y) are rather straight and of moderate size; between them there is a lozenge-shaped ossification of the basal plate (b.br.). 94 MR. W. K. PARKER ON THE STRUCTURE AND The roof bones (f.p.) are strong, well developed behind, where they rise over the arches of the canals (figs. 1 and 3), have a small temporal wing, and are short and rounded in front, where they leave the girdle-bone uncovered, scarcely covering the ereat fontanelle. Also between the temples there is a small gap which answers to the “ parietal fonta- nelle” of a Lizard; the orbital flange of each bone (fig. 3, fp.) is thickish, and very irregular in outline. Answering to the large size of the cartilaginous ethmo-palatines the nasals (7.) are huge bones, the left much the larger of the two, and sending an angular process over the fronto-parietal. Their fore part is pointed and divergent, their facial or descending region thick ribbed and projecting ; their top is strongly convex. The premaxillaries and maxillaries (pz., mx.) are quite Ranine ; they are strong and well developed. The septo-maxillaries (s.i7.) are small and inconspicuous. The quadrato-jugals (q,j.) are grafted on to the quadrate ; they are strong, arcuate, tusk-like bones. Also the top of the squamosal (sq.) is like a tusk, being curved, and turning inwards in front asa sharp fang; the supratemporal part is of small extent, and the descending plate long and broad. The parasphenoid (fig. 2, pa.s.) is perfectly normal; it is large and swelling, has a long point in front, its wings bent backwards, and its handle triangular. The vomers (fig. 2, v.) are large, many-cornered hones; they bind crescentically round the inner nostrils (7.7.), which are large, circular, and very wide apart. The front process of the bone is small, not reaching nearly to the end of the subnasal angle; the terminal plate almost touches its fellow, is a fan-shaped tract, and ends in a thick crescentic ridge, turned obliquely outwards and backwards, and armed with an arched row of sharp teeth. The nasal valve-cartilages (w./.w./?.) are well developed. This skull differs from that of a true Rana in :— 1. The epiotic growth. . The pterotic growth. ‘ . Its widening roof forwards. . The huge ethmo-palatine. . The additional palatine bony crest. . Absence of an inter-stapedial. . The small extent of, and bony tract in, the basal plate. . The large size of the nasals, and their want of symmetry. onNnrTn oOo Fk WH Db 20. Cystignathus -? sp.—Larva; 1 inch long; tail, 2 inch; hind legs, 1 line. Lake ystig I g 3 gs Jannarg, Manaoo, Brazils. This larva of a Neotropical Frog belongs, evidently, to the same genus as the last ; it is less than a third the length of the large Tadpole, whose skull has just been described, and in development it is intermediate between the last two, DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 95 The difference between this skull (figs. 8,9) and that of the lesser of the two large Tadpoles is slight, indeed, but there are noticeable modifications. The occipito- auditory regions are more developed, and the parts more fused; the fenestra ovalis (f0.) is larger and wider open, and more /ateral in position; but there is no stapes. The interorbital part of the skull is of the same shape, and so is the great fontanelle (fo.). The ethmoidal region is in the same unfinished condition, but the cornua trabeculee (c.t7.) are a little shorter, and the thick edge below is less solid, but it creeps round to the inner edge. Also at their root the cornua have a small snag for attach- ment of the membrane in front of the internal nostril (¢.7.). The ribbed edge of the orbitar process (o7.p.) is extended further back, as a strong decurrent enlargement, nearly to the apex of the pedicle (pd.). The pre- and post-palatine rudiments (the snag inside g., and pt.pa.) are much alike in both species; so are the mandibles (mk.), the lower labials (/./.), and the cerato-hyal (c.hy.). But the upper labials (w./.u.l’.) differ; the right and left inner pair are united in front, and all four are rounder and broader ; here we get nearer the Lamprey. The fronto-parietals (fig. 8, fp.) are large, and moreover they have a sigmoid, and not a straight, form, But the parasphenoid (fig. 9, pa.s.) is the true generic diagnostic; it is more split than in the younger Tadpole of the other species; the notch and the two sharp blades being one-fourth longer than the united part, which has more distinct lateral processes. If we reflect upon the short time in which this bone changed from what is figured in fig. 2, to what is shown in fig. 6, it will occur to anyone that there is strong reason to believe that, for a short time, each “trabecula” had its own investing bone. In the Chick, about the middle of the second week, the hinder parasphenoid (* basi- temporal”) is composed of a right and left osseous centre (Phil. Trans., 1869; Plate 82, fis. 2,.6.t.). The great constancy of this bone as an azygous basi-cranial splint in the Ganoids, Teleostei, Dipnoi, Urodela, and in most of the Anura, makes one surprised to see it with its apex bitten off in Rana pipiens and FR. halecina; it would not be more remarkable to find it modified by median division, also, and here this seems to be the case. 21. Cystignathus typhonius.—Adult female ; 14 inch long. Porto Rico. This small species has a much more ossified skull than the last large kind, just as the skull of the small Cyclorhamphus marmoratus (Plate 20, figs. 1-3) is much stronger than that of the large C. culews, as Mr. GARMAN has shown. On the whole this skull is very similar to the last, but its occipital condyles (Plate 16, figs. 6 and 7, oc.c.) are wider apart. The epiotic prominences (p.s.c.) are 96 MR. W. K. PARKER ON THE STRUCTURE AND less, and the pterotic processes (pt.o.) are much shorter. The parotic wings are much less, for this skull is only one-thirtieth wider than long. The canals are well marked (a.8.c., h.s.c., p.s.c.), also the whole occipito-auditory region is ossified, and this bone runs up to the optic fenestra (II). Between and a little below these spaces the floor of the skull is cartilaginous; but from thence up to the middle of the septum nasi, along the axis, the endocranium is all ossified, and the bone runs into the extremely large ethmo-palatine (e.pa.), leaving only the pre-palatine hook (pr.pa.) soft. Above, the bone is continued nearly to the front of the broad transverse snout, so that there are only a few places where cartilage remains. Behind, only the outside of the auditory floor is left below, and the pterotic angle (pt.o.) above, whilst behind these parts only the condyles retain their bark of cartilage. The three fontanelles (fig. 6) are quite normal, but they are rather small, as there is very much ossified overgrowth of cartilage (¢egmen cranit). The orbital region is almost oblong, rather narrower in front than behind, and bulging very little. The roof (a/.n.) of the nasal region is well developed ; the floor is of less extent than in C. ocellatus, but the pro-rhinals (p.r7h.) are large and hooked. The huge bony ethmo-palatines (e.pa.) help the bones to enclose the large circular nostrils (¢.n.), which are wide apart; the palatine bones (pa.) are well developed, falciform, and with an additional inferior sharp crest. The pterygoids are feeble and sinuously inbent, with a much shorter part for the pedicle (pd.) than in the last; they have consumed nearly all the cartilage. The Eustachian passages (ew.), lying in their angle, is sub-pyriform in outline. The quadrate is well ossified by the quadrato-jugal (q,j.), and the condyles (¢.c.) ave large well-made trochleee. The annulus (a.ty.) is large and perfect ; the stapes (figs. 6, 7, and 10, st.) is oblique, emarginate, and has a boss ; the feeble columella has a large undivided emarginate unossified proximal part ; the bony bar (m.st.) is feeble and curved. The extra-stapedial (e.st.) is a small irregular hook, and its ascending process is fibrous. The mandible (fig. 8) is rather slender; it has a long condyle (a7.c.), a sharpish coronoid crest (a7.), a short dentary (d.), and a well-formed mento-Meckelian (m.mk-.). The stylo-hyal (fig. 7, st.d.) is confluent above ; the hyo-branchial apparatus (fig. 9) is similar to the last, very elegant in form, has its front lateral lobe dentate in front, its basi-branchial ossicle (b.br.) divided in the middle, and there is a band of bony deposit along the edges of the basal plate ; the thyro-hyals (t.hy.) are long, straight, and slender. The nasal valve-cartilages (1./'.w.2?.) are well developed. The investing bones are also well developed; the fronto-parietals (/:p.) cover the temporal and part of the epiotic region, just roofing the lesser fontanelles, behind, where they form a straight margin; also in front they form a straight margin, and only just cover in the large fontanelle (fig. 6). The nasals (7.) are large, but less irregular, convex, and ribbed than in the last. The marginal bones (pz., mux., qj.) are all normal ; DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 97 they are slight, but dense; there is a small septo-maxillary (fig. 6, s.m«.). The post- orbital process of the squamosal (sq.) is straight, the supra-temporal tract small, the descending part normal. The parasphenoid (fig. 7, pa.s.) is not so attenuated in front, nor do the wings turn backwards as in the last kind; the tract between the wings is strongly ribbed in a cruciform manner, and the whole hinder part is largely anchylosed to the ossified endocranium. The vomers (fig. 7, v.) are similar to those of the last, but they are broader, and with smaller notches in their margin. To the discrepancies between the last and the type, we must add (1) the highly osseous state of the skull, (2) the ossified ethmo-palatines and quadrates, (3) the anchylosed parasphenoid, (4) the fibrous supra-stapedial, and (5) the additional centres of the basal plate. Fourth genus. Plewrodema, 22. Pleurodema Bibronii.—Adult female ; 14 inch long. Chili. The skull of this species is very evenly semi-oval, and its greatest breadth is to its length as 8 to 7; it, therefore, is one of the short skulls (Plate 18, figs. 1, 2). At first sight its main difference from the “norma” seems to consist in its open fontanelle (fo.) and the absence of the two secondary fontanelles, which open, like sky-lights, in the hind skull of more typical forms. More attentive observation, however, brings out several aberrant characters. The occipital condyles, the basi- and superoccipital tracts of cartilage, and the foramen magnum (figs. 1 and 2, oc.c., fm.) are all normal. The occipito-auditory region is more extended, laterally, and more ossified than in the type. Moreover, there is here a character which I have already described in another aberrant Neotropical Frog, namely, Pseudis ; this is a tract of cartilage running over the crest of the posterior canal (p.s.c., ¢.0.), and which, in Pseudis, is due to segmenta- tion of the originally continuous bony matter answering to both prootic and ex-occipital —that is an unmistakably archaic character. The basioccipital cartilage (fig. 2) is only half the width of the upper tract (fig. 1); this latter is a large wedge-shaped tract (¢.cr.) which runs to, and forms the hind margin of, the great single fontanelle (/o.). This is bounded by the ex-occipitals behind and by the prootics in front (p7r.o., ¢.0.), and the latter run forwards just as far as the cartilage above, up to the optic fenestra (IL.); they therefore take in the whole “ alisphenoidal” wall as well as the “ prootic” region. The ‘‘tegmen tympani” (¢.ty.) is extended far outwards, and this parotic projection is largely cartilaginous; beyond the horizontal canals (/.s.c.) the auditory region loses one-third of its breadth, and this is bevelled off from the hind margin, mainly. In front the cartilage is nearly one-third the breadth of the whole auditory region; behind, it is nearly one-half, that is above. But below (fig. 2) the bony matter reaches to the edge of the sub-convex vestibular region (vb.), with its fenestra and cartilaginous plug or stapes (sé.). MDCCCLXXXI. 9 98 MR. W. K. PARKER ON THE STRUCTURE AND The interorbital region is rather pinched than swollen in the middle, and gently narrows, forwards ; from the hind margin of the optic fenestra (II.) to the front of the true ethmoidal region, half behind, is cartilage, and half, in front, is bony; the latter tract is the “ girdle-bone” (eth.). This bone has a convex fore, and a concave hind, margin; it nearly reaches the fontanelle above and spreads a little into the ethmoidal wings, right and left. There is no trace of a superorbital “ eave.” The nasal region is normal as to its wide floor (s.n.l.), gradually lessening, backwards; it has large pro-rhinals (p.rh.); and it wants a pre-nasal rostrum ; but the roof (fig. 1) is very narrow, and the thick wedge-shaped end of the septum nasi (s.7.) is ossified separately: this bone (p.n.) is a low triangle, with its apex backwards. The large sub-oval orbital spaces, and the strongly bent bow of the palato-suspen- sorial are normal, but wider than usual; the pterygoid bone (pg.), with the pedicle (pd.), are also normal; but the palatine bone (pa.) is composed of two sub-equal pieces. These are large in proportion to their cartilaginous model, and the outer piece is large and falcate. The quadrate hinge (q.c.) is normal ; above it the quadrato- Jugal (q,j.) has ossified a good tract of the cartilage. In the rounded angle formed by the forks of the pterygoid the Eustachian passage (ew.) is of medium size, is oblique, turning outwards and forwards. The “annulus” (a,ty,) is broad, rather small, and imperfect above; the stapes (fig. 4, s¢.) is large, oval, sub-pedunculate, and obliquely emarginate on its antero- superior edge. , The columella (fig. 4) is pistol-shaped, there is no interstapedial cartilage separate from the medio-stapedial; this bar is gently arcuate (m.st.), and ends in a broadly spatulate extra-stapedial (e.s¢.) with a strong supra-stapedial (s.s¢.) confluent above. The mandible (fig, 3, d., a7., m.mk.) is perfectly normal, and so is the hyo-branchial apparatus ; the stylo-hyal (fig. 2, st.h.) is articulated, above. The investing bones (figs. 1, 2) are on the whole similar to those of a Common Frog of the same size, but they are less solid. The parasphenoid (pa.s.) has a similar form, and so have the vomers (v.); these, however, are turned outwards much more. The pre-maxillaries, septo-maxillaries, maxillaries, quadrato-jugals, and squamosals have the typical form, but are all rather feeble. But the fronto-parietals (fip.) are very arrested ; and although they reach from the ethmoidal wings to the junction of the anterior and posterior canals (a.s.c., p.s.c.) they are each only two-thirds the width of the naked space between them. The contrasts between this and the typical skull arise mainly from the following characters :— 1. There are no secondary fontanelles. 2. The fronto-parietals are abortively developed, leaving the fontanelle largely open. 3. The occipito-auditory ossifications, right and left, ave generalised, and not properly divided into prooties and ex-occipitals. 4, There is an “anterior septal bone” in the nasal region, DEVELOPMENT OF THE SKULL IN THE BATRACHIA. oo cr . The palatine ectostoses are in two pieces on each side. 6. The quadrate is partly ossified. 7. There is no distinct inter-stapedial. . The supra-stapedial is confluent, above. co Fifth genus. Lymnodynustes. 23. Lymnodynastes tasmaniensis—Adult female ; 13 inch long. Tasmania. This is a long skull (Plate 18, figs. 5-8), with short auditory, long nasal, and average orbital, regions. The length is the fortieth of an inch greater than the breadth, and the outline of the face is half a long oval. The sides are rather straight in the jugal region, for the broad and long nasal region makes the fore edge a regular semicircle. The whole ethmo-nasal territory is, as compared to the skull cavity, as 2 to 3 in length; the average fore and aft extent of that territory, as compared with that of the cranium proper, is as 1 to 2. The occipital condyles (oc.c.) are of moderate size, and of the usual distance from each other; they project but little, are more posterior than inferior, and the emargination of both floor and roof is crescentic. The auditory capsules are of moderate size and become narrow at the tegmen (é.ty.); the cranial interspace between them is very wide, especially in front. The ethmoidal region (eth.) is also wide, and the broad flat skull is almost like an hourglass, being much the smallest in the middle, and thus the suborbital fenestree are almost oval in shape, the arcuate palato-pterygoids bounding them externally. The whole pre-cranial region is very flat and outspread. The tegmen cranii (¢.cr.) is more complete in this skull than in most of those known to me; it resembles in this respect the skull of the Skate, the roof growing along from the sides, so as to leave, when the roof-bones are removed, a space little more than a third as wide as the narrow interorbital region, and little more than a third the length of the cranial cavity. Also those secondary spaces, so characteristic of the Anura, are absent, and the wide inter-auditory space, above, is completely chondrified. The tract, moreover, in front of the fontanelle (eth.) is large both ways, so that a very unusual amount of the girdle- bone is exposed. The ex-occipitals and prooties (e¢.0., pr.o.) are confluent for some extent at their inner edge, above; below, there is a good space of cartilage dividing them. Above (fig. 5), the whole tegmen tympani (¢.ty.) is soft, and a wedge-shaped tract of cartilage is seen between the bones. Behind this tract the ex-occipital is seen to reach outwards beyond the horizontal canal (/.s.c.) and inwards to such a distance from the middle that the superoccipital region (fig. 5) is left of unusual width. Below, the basioccipital synchondrosis (fig. 6) is only half that width, and is quite normal. The prootics run forwards, above (fig. 5, pr.o.), over the optic passage, and inwards 0.2 100 MR. W. K. PARKER ON THE STRUCTURE AND they each grow into a large faleate tract which nearly reaches the end of the fontanelle, and then come within a moderate distance of each other, near the middle. The roof of cartilage has a concave edge behind, as well as in front; and from the curved tracts of bone the cartilage runs forwards on each side to the end of the fontanelle, and exists, between the prootic and girdle-bone, as an obliquely four-sided space, which occupies about a fourth of the orbital edge of the skull. Below (fig. 6), the prootics and ex-occipitals being of less extent, the cartilage is present as a large wedge on each side running into the floor of the skull; nearly the whole vestibular floor is unossified, and in front of the ear, the prootics, not completely surrounding the foramen ovale (V.), leave a tract of cartilage which surrounds the whole -of the optic fenestra (II.). The outer edge of the girdle-bone (eth.) margins half the orbital rim; it is only typically developed as to extent, a clear tract of cartilage separating it from the orbital fenestra ; while in front it only just reaches the long septum nasi (s.n.). This bone is large, because of the width and size of the matrix in which it is formed, and not because of overpassing its own proper territory, as in several kinds of Anura. On each side the ethmoid bone does not quite cover the wings of the ethmoidal cartilage (fig. 5, pa.), from which the ethmo-palatine bars are extended; these transverse bars and all the huge nasal region, are left unossified. The roof of the nose (fig. 5, al.sp.) is very wide and enlarges from before, backwards ; the floor of the nose (fig. 6, s.7./.) is equally wide, and lessens from before, backwards. Here we see what has taken place since the larval stage—viz.: that the huge cornua trabeculee are now united together by a solid intertrabecular tract, ending, in front, in a rounded knob; whilst near this knob, on each side, a small ligular fork has been given off which passes between the laminee of the premaxillary ; this small secondary “cornu” is the pro-rhinal (p.rh.). Above (fig. 5), the nasal roof throws the external nostrils (e.n.) far apart ; each lamina has a rounded emargination behind the nostril, and then expands again until it passes into the substance of the ethmo-palatine band (fig. 5, pa.) That band is twice the normal width, and spreads externally into a large fan-shaped plate (fig. 6). The hind horn of this expansion (pt.pa.), as it passes into the pterygoid, becomes a mere thread, and the rest of the arch, although normal, is very feeble, and its two forks, the pedicle and quadrate region (q.c.), are short. In this species it is not difficult to detach the palatine and pterygoid bones (pa., pg-), for they graft themselves only slightly on the cartilage within. The palatines (fig. 6, pa.) are flat, lathy bones, rounded only at their inner end; they are sub- sigmoid, with sinuous edges, and an abrupt outer end; their inner ends pass largely under the face of the girdle-bone. The pterygoids (figs. 5 and 6, pg.) are but little larger than the palatines ; they DEVELOPMENT OF THE SKULL IN THE BATRACHIA. | 101 bend inwards in front of their fork, and the inner process covers the short cartilaginous “pedicle” very imperfectly. The quadrate region (q.c.) is short, but the condyle is large and reniform ; it only reaches to the middle of the columella ; this region is not ossified. The Eustachian tubes (ew.) are large and circular; the annulus (fig. 5) is small —two-thirds the typical size—and imperfect above. The stapes (figs. 6 and 8, st.) is large, oval, only straight-sided supero-anteriorly, not emarginate ; for the columella is very small (fig. 8); it is pistol-shaped, has a bilobate unossified tract fitting inside the stapes, but no segment there. The medio-stapedial bone (m.st.) is dilated at the upper end, where it fails to ossify the bilobate tract ; it then becomes a very slender rod, dilating gradually to its end, beyond which the unossified distal part is a small, narrow, cochleate extra-stapedial (e.st.), with no ascending process. The stylo-hyal end of the hyoid (fig. 8, st.h.) is narrow, and articulates with the auditory floor. The mandibles (fig. 7) are long, and quite normal; the mento-Meckelian part (m.mk.) is unusually large, showing that the lower labial kept its size, after fusion with the mandible, more than is the case generally. The hyo-branchial plate (fig. 7) is normal, the basal part is wide, but its processes are slender ; the hypo-hyal lobe (h.hy.) is sharp and slightly perforate ; the lateral lobes are smaller than usual, and the thyro-hyals (¢.hy.) are very slender, diverging rods. The investing bones (figs. 5 and 6) show a feeble skull; the fronto-parietals (fig. 5, fp.) are thin curved shells of bone, hooked outwards, in front, where they bind on the outer angle of the girdle-bone, sinuous along their inner edge, where they form a waist to the fontanelle (fo.), and dilated behind into a large round lobe ; these lobes, right and left, cover much of the unossified supraoccipital tract (f.cr.) ; externally, each bone just rises on to the swelling of the anterior ampulla (a.s.c.). The nasals (fig. 5, 7.) are large, but leave a width of cartilage uncovered between them, equal to their own diameter. These bony shells are notched behind the outer nostril, in relation to it; their postero-external part, or handle, is short and sigmoid. The parasphenoid (fig. 6, pa.s.) is typical, but very broad in the main shaft, and notched in front ; the vomers (v.), as in Plewrodema (fig. 2), are placed almost trans- versely, and the inner nostrils (.7.) open wide apart in the large rounded notch of each vomer. The main wing of the bone, in front of that notch, is, as usual, split into two sharp lobes. The dentigerous stem of each bone is a long thick rib, serrate with recurved teeth, subcrescentic in outline, and nearly reaching its fellow. Hence in opening the mouth of a Frog of this species the vomerine teeth are seen to run nearly straight across behind the inner nostrils, and to be scarcely separated at the middle ; their real direction is inwards, and a little backwards, their ends lying under the fore margin of the girdle-bone. The bones that fence the semi-oval face are normal, but delicate ; the premaxillaries (px.) are wide, and the septo-maxillaries (s.m«.) small; they lie outside the second upper labial; both these cartilages (w./!.w./*.) are normal. The squamosals (sq.) are slight, and only just clamp the edge of the short tegmen tympani (t.ty.); the > 102 MR. W. K. PARKER ON THE STRUCTURE AND maxillaries (m.) are thin, high shells in front, and have a slender jugal process ; the quadrato-jugal (q.j.) only fastens on, does not ossify into, the quadrate. As compared with the normal skull, this of Zymnodynastes has many curious points of difference ; it would fit in among the skulls of Australian ‘‘ Anura,” generally, better than among those of the Palearctic region, even those of its own “ Family” ;— 1. It is a frail skull, the outer bones very feeble. 2. It is greatly depressed. 3. Its nasal region is one-third longer, and much broader than in the type. 4. The prootics are confluent with the ex-occipitals above. 5. The tegmen cranii is extremely developed, behind, in front, and also over the edges of the skull, so that there is only one small median fontanelle. 6. The fronto-parietals fail to cover the space. 7. The fore part of the palato-suspensorial cartilage is very broad and large, and the hind part unusutally small. 8. The dentigenous stem of each vomer is very long and sub-transverse in position. 9. The annulus tympanicus is only two-thirds the normal size. 10. The columella is arrested and very small, without inter- or supra-stapedials. Sixth genus. Camariolius. 24. (A) Camariolius (Pterophrynus) tasmaniensis (?)—Adult female; ? inch long. Australia. * 1. Skull of the adult.—This is one of the smallest of the Batrachia; mine had large ripe ovaries, and was only ? inch (9 lines) long; Dr. G&NTHER (see note) gives 13 lines as the length of two, and 1 inch for the length of a third, species. This is a long skull (Plate 19, figs. 1, 2), for its breadth is but little greater than its length ; the contrast is much greater in a large number of “ Anura.” Knowing this to be an adult, a satisfactory comparison can be made of its skull with that of the typical species. . This skull is the frailest I have yet seen in the group; and this not only as to the small quantity of bone that enters into its composition, but because of the economy as to cartilage also. I shall describe, in their turn, other very minute skulls from the same region. The occipital condyles (Plate 19, figs. 1, 2, oc.c.) are rounded in form, and largely inferior in position ; the interspace is moderate, and gently emarginate. The auditory capsules have the large ovoidal fulness of form seen in the young of the larger kinds, but they are intensely ossified ; their retention of the ovoidal form, with but little teymen, is one cause of the relative narrowness of the skull. * For a deseription of this genus see GiinrHER, Proc. Zool. Soc., 1864, pp. 46-49. That author has examined this specimen for me, and considers it to be either C. tasmaniensis, or the one most akin to that species. ee DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 103 The wide temporal regions become concave in passing into the orbital, and then the skull narrows gently and bulges a little in front, near the ethmo-palantine “axils.” The ethmoidal region (eth.) is covered by a very short “anterior tegmen;” the margin of this part has a large evenly rounded edge, and so also, in the other direction, has the “posterior tegmen,” which is limited to the superoccipital region, and is barely twice the breadth of the narrow ledge in front. These two opposite margins of the great single fontanelle (fo.) are very far apart ; the hinder is nearly as far back as the middle of the auditory capsules ; the fore edge is nearly as far forwards as the arcuate ethmo-palatine (fig. 1, pa.). The fontanelle, open over three-fourths of the cranial ‘“ barge,” narrows in the temporal region, and only recovers three-fourths of its size behind; the roof-bones (f-p.) are at a considerable distance from it in every part. The cartilage which bounds it laterally is narrow in front and widens in the temporal region to thrice its first breadth; this is the lateral remnant of the tegmen eran, and it is, for the most part, quite independent of the walls (see fig. 2), which from a little behind the ethmoidal “axils” to the back of the optic foramina (IIL.) are entirely membranous. Thus the optic fenestra (the “foramen” is a hole through it) is excessively large; a good series of gradations in this respect are to be seen in the Australian Anura, and will be described in due time. The floor of the skull is very narrow—only half the width of the roof—and is merely composed of the trabecule and an ‘“ intertrabecular” band of the same width. The nasal regions are quite normal, except that the subnasal angle is simpler than in the type. There is no “rostrum,” and the pro-rhinals (p.rh.) turn inwards, and are not much dilated distally. The roof (fig. 1), widest behind, and the floor, widest in front, are normal ; the septum (s.n.) is thick and clearly seen. The ossification of this little semi-membranous skull is of great interest, for its very minuteness has enabled it to escape from the strict morphological bonds that keep the larger kinds in order. Nearly all the ossification of the endocranium is behind the orbits ; there are more, and there are fewer, osseous centres than in the typical kinds. The additional centres are the median bones that have been crushed out, so to speak, by the special law of Batrachian morphology ; these have crept in again. Fewer centres are seen in the lateral parts, for the occipito-otic centres (prootic and ex-occipital, pr.o., e.o.) are here, as in Pseudis and other generalised types, ossified fore and aft, without the transverse dividing band of cartilage, seen in those in which the bone begins only at the proper nerve outlets. And they are intensely ossified too; only a trifling edge of cartilage is left in the tegmen and in the floor (figs. 1 and 2), and in front the bony matter takes up half the “ alisphenoidal” tract, giving a good bony margin to the “ foramen ovale” (V.). The upper and lower median synchondroses are of moderate width ; the lower is oblong, and is less than a third the width of the inter-condylar 104 MR. W. K. PARKER ON THE STRUCTURE AND space, and contains the rudiment of a basioccipital bone (fig. 2) as a frail bony “ cephalo- style.” The upper tract is of the same width at tl foramen magnum, but widens out, wedge-like, to the fontanelle ; it contains the rudiment of a supraoccipital bone as a slight endosteal deposit. The antorbital (avillary) tracts of the ethmoid are ossified (eth.), not, however, as a “girdle-bone,” but as in “ Urodeles” as a pair of ‘ sphenethmoids ;” these occupy about three-fifths of the ethmoidal cincture—perfect here for a short distance. Below (fig. 2), the two bones are separated by a distance less than their own width; above (fig. 1), they come a little closer together, ending there as sigmoid “ horns,” turned forwards till they nearly touch the conchoidal nasal roof. These bones, laterally, just run round the end of the long oval interorbital fenestra ; in their interspace below there is between these two ecto-ethmoids an evident, endosteal, ‘“‘mesethmoidal” tract like the azygous endosteal patches in the hind skull. Here, once more, we come athwart the familiar “ perpendicular ethmoid,” and the basi- and supraoccipital bones. The palato-suspensorial arches have thin bony plates (pa., pg.), and these, as well as the bones that surround the face, are normal, if compared with a young Common Frog of the first summer. The vomers (fig. 2, v.) are small and toothless; the parasphenoid (pa.s.) is narrow, especially in the fore part, yet it is, essentially, typical. But the fronto-parietals (fig. 1, fp.) are by far the slightest and most delicate I have yet seen: they reach to the conchoidal nasal roof in front, and partly cover the anterior canal (a.s.c.) where these needle-like bones become roughly pedate. These bones only cover the outer edge of the arrested lateral rudiments of the “tegmen cranii.” The nasals and the squamosals (7., sg.) are under-sized but quite normal; so also are the parts of the mandible (fig. 3), and the hyo-branchial plate with its processes (fig. 4). The annulus (fig. 1, a.ty.) is open above and rather small. The stylo-hyal (s¢.h.) is articulated, above ; the Eustachian passage (ew,) is of medium size and reniform, with the concavity looking outwards and backwards. The stapes (fig. 5, st.) is between a lozenge and an oval in shape; it is large, and has a short stalk, The cartilaginous enlargement on the oblique end of the columella is not segmented off as a distinct inter-stapedial. The medio-stapedial (m.st.) lessens rapidly from this cartilaginous process, and ends as bone, a little behind an oval, decurved, thick but small, foot of cartilage—the extra-stapedial (e.s¢.) ; there is no ascending process from this lobe. The labial cartilages (v.l'.u./?.) are normal. Many of the things that dis- tinguish this from the typical skull depend upon arrest, and are such as can be seen in a young Common Frog, five or six months old. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 105 24 (continued).—(B) Camariolius tasmaniensis (?).—Larva, $ inch (9 lines) long ; legs, so ch. Same locality. This Tadpole was the largest of several of this species examined by me, and its free hind legs showed that it has attained to its full larval condition ; it was as long with its tail as its parent without. In the possession of paired nasal sacs, each with a distinct roof, and of a well defined fenestra ovalis, this larva had gone beyond the Lamprey; but there was no bony deposit even below the skull (parasphenoid), and the hinder half of the long hypo- physial space was still membranous; the rest of the space was composed of a very thin layer of young, half consistent cartilage. This skull, therefore, was in the best possible state for comparison with that of Petromyzon and its allies, Leaving out the prenasal structures this skull (Plate 15, figs. 6, 7) is nearly square ; it is unusually short, and although the head was no larger than that of a Blow-fly, the chondrocranium had become a very solid structure ; and cartilage was forming even in Tadpoles one-third the length of this—that is, in newly-hatched specimens corres- ponding to my earliest stage in Bufo vulgaris (Phil. Trans. 1876, Plate 55, fig. 1). The first-formed cartilage can now be well seen, as it is much more massive, and richer with proliferating cells, than the newer tracts. The cartilage, which at first only enclosed the apex of the notochord as the ends of the trabeculz, has now spread along the whole hind floor and a very definite tract crosses in front of the notochord. That rod is expanding to its full (spinal) size just where it emerges from between the parachordal bands (iv.); they pass outwards and backwards, and end in a free rounded point behind the ganglion of the 9th and 10th nerves (IX., X.), which lies in a notch in the outer border of each band. These bands are distinct until they reach the apex of the notochord, and then, as just mentioned, unite in front of it; there they form the hinder boundary of the pituitary space. Directly in front of the notch for the ganglion each parachordal dilates suddenly into a broad crescentic wing, which forms a concave floor for the antero-internal half of the auditory capsule (fig. 7) right and left. There is a shallow crescentic notch in the middJe of the hinder margin of each wing; this forms the inner boundary of the fenestra ovalis (fo.). In front, there is a deep round notch between the wing and the trabecule ; im and over this space lie the more or less united ganglia of the 5th and 7th nerves (V., VIL). The sub-orbital fenestra, and the band of cartilage which encloses it are, together, scarcely of larger extent than the auditory capsules behind. These have lost their simple oval form, for the three large canals and their globular ampulle (a.s.c., h.s.c., p.s.c.) above, and the sacculus (vb.) below, have wrought the sac into their own form. The alate basal floor is thin, and composed of rather young cartilage, but the auditory sacs are very solid, for they become cartilaginous directly after hatching, coevally with the trabecule and suspensoria. MDCCCLXXXI, Pp 106 MR. W. K. PARKER ON THE STRUCTURE AND They are now perfectly distinct from that floor, and a little pressure serves to shell them out under the cover-glass; this causes no tearing of the tissue. The large fenestra ovalis (fo.) is as wide as one of the canals, and as long as the diameter of an ampulla; it lies along the middle, obliquely parallel with the horizontal canal ; its antero-internal edge is double, being composed of both floor and capsule; the vesti- bule swells out on its inner side in a crescentic manner. From the inner edge of each capsule, above (fig. 6), there is a little tract of new roof-cartilage, which will finish the occipital arch. The basis cranii in the orbital region widens a little forward ; the trabeculee (t7.) occupy nearly a third of its breadth; the trabeculee have met in the ethmoidal region, and are becoming fused together ; the line of fusion is nearly equal to the intertrabecular space, behind (v.t7.), which is now half filled in with new cartilage. In front, the trabecule are free, diverge at more than a right angle, and curve out- wards ; these are the horns of the trabeculz (c.tr.). Newer cartilage has appeared in several places ; laterally, the trabeculze have grown into orbital walls, and these walls are growing over the roof a little, especially in the ethmoidal region (fig. 6). Out- side, in front of these rudiments of the “tegmen cranii,” the “ ethmoidal wings” have appeared, embracing the nasal sacs (7.7). Between these sacs the trabecule are very thick, where they are narrowest, and then suddenly expand into the “cornua.” These latter, or their inner edge, are developing a thin, crenate expansion, ready to finish the nasal floor. The trabeculze are most compressed where they form, below, the semi-circular inner boundary of the internal nostrils (¢.n.). Over the sac, right and left, there is a small cartilage, like the valve of a small Entomostracan, attached to the ethmoidal wing : this is the nasal roof- cartilage (first “ paraneural,” mr.) ; here it is more distinct than usual have ever seen it in this group. The huge suspensorial bands reach by their distal condyles (qg.c.) in front of the nasal sacs, whilst, behind, they curl themselves up against the intruded auditory capsules (au.), and bend suddenly upwards as they pass inwards, to form the narrowed (but really broad) pedicle ( pd.).* At this part the cartilage rises so high that it would seem to form the counterpart of the “ascending process ” of the Urodeles, but it drops again before it passes into the trabecula, and although riding over the vidian nerve (VII'.) it passes under the 5th (figs. 6, 7, V.). This thick high part will be further outwards, afterwards, and will form the otic process (ot.p.). In front, the quadrate passes inwards as a broad lamina from the saddle-shaped condyle, and grows into a hook—the pre-palatine rudiment (fig. 7) bounding the nasal passage, below, as the ethmoidal wing does, above. The pterygo-palatine conjugational band is short, wide, and out-turned; on it, as yet, I see no definite rudiment of the post-palatine process. Over it, from the large projecting sulcate condyle for the hyoid (hy,f.), the orbitar process (or.p.) grows ; more than [| * The pedicle lies on a higher plane than the pterygo-palatine band; it is, like the ribbed edge, and orbitar process, a temporary tract of the suspensorium. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 107 it is a very long, narrow leaf of cartilage, and lies in a recess of the ethmoid, nearly touching the wing; its fore edge is ribbed. The mandible (mk.) is a thick, short, chubby cartilage—like a little short wlna; the lower labial (/./.) meeting its fellow by a broad face, is thick and strong; and the temporary upper labials (w./.) are thin crescentic leaves of cartilage. The hyoid bar (fig. 7, c.iy.) is normal; above, there is the sinuous condyle, and the rudimentary styloid process; and ventrally it is dilated largely—most behind—and then suddenly less, it becomes the basi-hyal ; this is composed of simple cartilage. This chondrocranium is essentially like that of the youngest Pseudis, another member of the Cystignathidze (see Plate 2) ; that is the /wrgest, and this the smallest larval Frog I know. The larval skulls are, however, very uniform throughout the “Opisthoglossa ;” yet they have non-essential differences that are of great interest. The skull of the adult Camariolius is like that of a young typical Frog, several months after transformation. This is seen in the general lightness of the investing bones—the moderate extension, backwards, of the quadrate hinges, and the divided condition of the ethmoid. But there are many things that cannot be put down to mere arrest ; these are :— 1. The continuity of the membranous space, above ; only one very large fontanelle. 2. The extension of the orbital (or optic) fenestra over three-fourths of the wall, in that region. 3. The intensely ossified and generalised condition of the occipito-auditory region, right and left. 4, The very slight and arrested, tripartite state of the elements of the girdle-bone; a “meso-ethmoidal ” rudiment appearing. 5. An upper and a lower median rudimentary osseous centre in the occipital arch. No septo-maxillaries. . Extremely rudimentary fronto-parietals. . Quadrate region partly ossified. . Columella rudimentary ; no “ inter-” or “ supra-stapedial.” SD OND . Vomers rudimentary and toothless. Seventh genus.— Cyclorhamphus. 25. Cyclorhamphus marmoratus.—Adult female ; 13 inch long. Vinco Caya, Peruvian Andes ; height, 16,000 feet. This is a stout, evenly semi-oval, short skull (Plate 20, figs. 1-6) ; the length is to the breadth as 62 to 8. The quadrate condyles only reach opposite to the middle of the stapedial plates, the epiotic eminences are almost flush with the occipital condyles, the roof is imperfect, and the ossification is intense, and attended with much anchylosis behind. The ear is imperfect and the nasal roofs short, the vomers small, and the inner P 2 108 MR. W. K. PARKER ON THE STRUCTURE AND nostrils much nearer together—against rule—than the outer. This is one of the lowest and most generalised of the Ranidee. The occipital condyles (Plate 20, figs. 1, 2, oc.c.) are large, oval, and postero-inferior ; they are separated by a concave notch of their own width. The whole hind skull has a remarkable form, for the auditory capsules are very large and thoroughly ossified, yet the canals project very much both before and behind, pushing the epiotic eminence almost as far back as the occipital condyles, and the anterior (prootic) swelling, caused by the anterior canals, into the orbital region. The parotie projections suddenly become one-half the size of the main capsule and are only one-third at the tegmen (fig. 1). The extreme edge of this part (inside sq.) is unossified, and there is a small semi-osseous tract in the superoccipital region ; all the rest of the endocranium is ossified up to the middle of the nasal septum (fig. 2, eth.) below, and still further forwards above (fig. 1, s.n.). Moreover, the parasphenoid behind, and the fronto-parietals for fully two-thirds of their extent, are anchylosed to the bone within. The temporal fossee are deep rounded hollows ; the orbital region of the skull, from thence, swells gently, and then remains of the average width up to the antorbital bars; then there is a slight superorbital projection, but it also is ossified. The tegmen of this hard endocranium is extensive before, but more so behind ; for that tract reaches so far as to lie over the front edge of the optic fenestra (II.). The edges of the open part are wide, so that the single fontanelle (/0.) is only one-fourth the length of the cranial cavity, and is narrow, and pinched in the middle. Here the roof is almost covered in, as in some old Skates’ skulls. The girdle-bone, as in some Salamanders, is not marked off behind; in front it runs far forward into the true nasal region, and right and left, ossifies the ethmoidal wings, but stops at the ethmo-palatines (fig. 1, e.pa.) exactly where the segment is in the genus Bufo. The nasal region (m7.) is very remarkable, being, like that of Bombinator (Plate 25, fig. 1), merely a transverse double pouch ending in a broad sub-arcuate snout, which is scarcely covered by the nasals, which lie mainly on the proper ethmoidal tract. Hence the outer nostrils (e.n.) are very wide apart; but the floor being very narrow where the trabeculz originally came first into contact in the internasal region the inner nostrils (fig. 2, 7.7.), although large and round, are very near together tor a Batrachian ; the relative distances are exactly reversed in this case. The roof (n.7.) is narrow from the ethmoid onwards, but the nasal wall (fig. 1) forms a pouch behind the outer nostril. The inner labial (w./!.) is small, and lies against the front of the broad snout; the outer piece (w./*.) is in front of the nostril. Below (fig. 2), the trabeculze in becoming the floor have scarcely changed their form at all; but they have budded out into a pair of large falciform, secondary cornua at their inner angle ; these are the pro-rhinals (p.rh.). The palato-suspensorials are normal in the palatine region, and the bone (pa.) is of the usual f-shape ; it forms but little union with the cartilage. DEVELOPMENT OF THE SKULL IN THE BATRACHTA. 109 The pterygoid (pg.), on the other hand, has used up most of the cartilage to which it clings, and growing beyond the pith of the pedicle (pd.) makes this part pointed. On the right side, that independent ossification of the outer and inner elements which is so common in the “ Anura,” shows itself in a remarkable manner, for the very apex of the cartilage becomes detached, as a “meniscus,” and then thoroughly ossified, as a small, free, diamond-shaped “ metapterygoid” (figs. 5 and 5a, pd., pd.m.). In Rana temporaria, and others of that genus, the metapterygoid, which appears during metamorphosis, is formed in the outer fibrous layer: here it is endoskeletal. The quadrate region is neither long, nor far retreated (fig. 2, q.c.), it is all ossified by the quadrato-jugal, except the apex of the otic process, which is continuous with the slight remaining pith of the pterygoid, and the edge of the tegmen tympani (fig. 5, sp. q., pg.c., pr.o.). The condyle (¢.c.) which ends opposite the middle of the stapes (st.) is a large, well-formed trochlea, strongly clamped on its inner side by the pterygoid. That process forms a very obtuse angle with the pedicle, and here a very small crescentic slit can be seen just through the skin of the mouth, its convex side lying against the bone: this is the blind Eustachian opening (ew.). Outside the suspensorium a very small annulus (figs. 3 and 6, a.ty., m.ty.) is seen; it forms three- fourths of a circle, its diameter, and the breadth of the band, is about one-third that of a typical Frog. The mandible (fig. 3) is very strong, especially in its hinder part; it is, however, quite normal. The stapes (fig. 6, st.) is of the average size; it is a very regular oval, but has a rounded process behind, and an oval “ boss” outside. The columella is extremely delicate, and is not segmented. The medio-stapedial bone (m.st.) occupies the dilated stapedial end, and runs along the thread of cartilage so as to be one-third the length of the whole rod, which ends as an arcuate extra- stapedial (e.st.) ; this is spatulate, and without a fork. The stylo-hyal (fig. 2, s.th.) is confluent with a trace of cartilage on the outside of the floor of the vestibule ; it enlarges in its descent (fig. 4, c.hy.), and becomes partly ossified. There is a small, sharp hypo-hyal lobe (/.hy.) in front, and one on the side of the short retral part ; between these bars the “notch” is a very large crescent. For the basal plate is very wide and short, and is of the same kind as we see in Calyptocephalus (Plate 21, fig. 4)—a near relative. There is an adze-shaped fore, and a styliform hind, lobe on each side; the thyro-hyals (t.hy.) are large, dilated, and separated by a space which is two-thirds of a regular oval. The proximal part of these bars take up half the basal plate, and they almost meet at the mid-line. The rest, all but the front edge and part of the fore-lobes, is semi-osseous. Like some other sub-typical Frogs—Pleurodema, Lymnodynastes—the fronto- parietals (fp.) do not finish the roof but are scanted in their front half, and end as rounded bars, each one-third the breadth of the roof. Behind, they are strong, smooth, sinuous, falling in at the temples, and swelling over the anterior canals, and 110 MR. W. K. PARKER ON THE STRUCTURE AND are thoroughly anchylosed to each other, and to the subjacent bone. The pre- maxillaries and maxillaries (px., ma.) are strong, smooth, high, and typical; and there is a large septo-maxillary (s.mx.) on each side. The falciform quadrato-jugal (figs. 1, 2, 3, and 5, q¥.) is perfectly continuous with the bony quadrate (q.); the squamosals (sq.) are like a hammer with a wide handle and a very short head ; this latter part lies on the tegmen tympani but little (fig. 1, sq.). The nasals (figs. 1 and 3, n.) scarcely hide the fore part of the girdle-bone, and leave the pouched short nasal roofs naked ; they are narrow, convex, curved shells of bone, with a descending narrow process, outside. The vomers (v.) are small and have a post-narial spur, but none in front of the passage ; the fore part is narrow and bifid; the dentigerous lobe is oval and less than normal. The parasphenoid (fig. 2, pa.s.) is large, normal, rather blunt and ragged in its processes, and is somewhat confluent with the superjacent bone, behind. Next to the skull of Bambinator and Pelobates this Frog shows most what is low and generalised in its skull, it differs from the ‘‘ norma” in : 1, The large relative size of the auditory capsules. 2. The shortness of the nasal capsules. 3. The unfinished roof. 4. The intense and almost universal ossification of the endocranium, and the anchylosis of the outer roof and floor bones. 5. The differentiation of an endoskeletal metapterygoid on the right side. . The smallness of the upper part of the squamosal and of the applied ‘ annulus.” . The arrested state of the columella. . The almost entire closing of the first cleft. . The short, wide form of the basal plate, and its semi-osseous condition. oS © ON SO . The narrowness of the interuasal region, and small size of the vomers. 26. Cyclorhamphus culeus*.--Larva ; length, 3} inches; tail, 2 inches; hind legs, 7 lines. Puno, Lake Titicaca, Peru. I am very fortunate in being able to give the structure of the larval skull in a geographical neighbour of the great Calyptocephaius ; this is one among many of the things I owe to my friends AGAssiz and GARMAN. Moreover, among the various kinds of larval skulls worked out by me, this comes vhe nearest to that of that helmeted Frog; and as we know in the case of Bombinator and Pelobates how weak one skull, and how strong another, may be, in congeneric types, there is no difficulty on that head as to the relationship of these two Neotropical forms. Cyclorhamphus marmoratus (Dum. and Brs.) evidently comes between these two forms in respect of its adult skull (AGAssiz and GARMAN, p. 277). * For an excellent figure of the adult of this fine species, see AGAssiz and Garman, “ Bulletin of the Mnseum,” No. 11, plate 1 (Cambridge, Mass., November 26, 1875). ———oO— DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 111 This larval skull will be best described by comparing it with the next (Plate 22, figs. 6—9 with figs. 2-5) ; all these objects are shown as magnified five diameters ; the Tadpole of Calyptocephalus was fully one-third larger than that of Cyclorhamphus, and in the latter the hind legs were four times (relatively) as long as in the former, and therefore it must have been somewhat more advanced in development, generally. Indeed, this will be seen by comparison of the figures; yet this is too slight to effect the general form and relations of the parts. The interauditory region, in this, is altogether narrower; the interorbital much more uniform, not narrowing, forwards, half as much ; and the palato-suspensorial bars converge more. On the whole, however, these two skulls might easily be taken for those of two species of the same genus; part for part, and process for process, there is a very close resemblance between the two. The notochord (ne.) is still found between the halves of the basal plate; the occipital condyles (0c.c.) have the same form ; the ex-occipitals (¢.0.) are more deve- loped, answering to the longer legs, and are seen beyond the twin nerve-passages (I1X., X.). The hinder “tegmen” only runs half as far forwards, and is not fenestrate, whilst the fore ‘‘ tegmen” is finished and the single fontanelle ( fo.) is a large evenly oval space, instead of being a smaller space, shaped like an oval leaf, with the stalk in front. The auditory capsules are not thrust so far out; they are naked, and show the canals (a.s.c., /.s.c., p.s.c.) through their diaphanous walls ; the spurs growing from the “tegmen tympani” are very similar (¢.ty.*, sp.c.) ; the stapes (st.) is like that of the next. The almost oblong interorbital region (fig. 7) is shorter, and scarcely bulges at all; its walls are perfect and run an edge over the roof, on which the bony “ wall plates” lie. The front tegmen runs forward as a wedge-like mass, for it has coalesced with the more developed intertrabecular crest, and the “lamina perpendicularis ” (p.e.) is now formed. The lateral ethmoidal wings (a/.e.), and the upgrowths of the tra- beculze, outside the emerging olfactory nerves (I.), have conspired with the median crest and the roof to finish the plaster model for the future “ girdle-bone.” The first rudiment of the septum nasi exists, now, merely as the foremost part of the inter- trabecula ; beyond this frail commissure of the paired bands the cornua trabeculz (c.tr.) diverge, arching both outwards and downwards ; they are longer and narrower, and more diverged than in the next, making the lozenge-shaped interspace left by them and the upper labials (w./*.v./’.) much wider. The palato-suspensorial arch (pd., p.pg., q.) shows its likeness and its unlikeness to that of the other species. The pedicle (pd.) is longer and the otic process is not distinct from the rounded “ elbow” of the cartilage, which is bent backwards more. This, with the gradual convergence of the main bars, makes the orbital space wider behind ; it is, also, shorter. The orbitar processes (07.p.) do not overlap so much as in the next, but keep outside the post-palatine rudiment (pt.pa.); this crest fails in its * In fig. 6 this is wrongly lettered f.cr. V2 MR. W. K. PARKER ON THE STRUCTURE AND hinder lobe, which is merely a blunt projection. The trabecular and pre-palatine spikes (c.tr., pr.pa.) help to surround the inner nostril (7.n.) in the same manner, and the quadrate region (q.) is alike in both, but most turned inwards in this. The passages, only, of the 9th and 10th nerves are shown, but the common root of the upper and lower maxillary branches of the fifth (V**.), the orbito-nasal (V1), the facial (VII.), and the long, undivided palatine or vidian are shown; the latter (VII’.) is seen passing forwards outside of, and a little below, the orbito-nasal; it pierces the ethmoidal wing, then runs close to the mid-line, and supplies the upper lip. The optic (II.) and olfactory (I.) are seen in situ.* The spiracular cartilage (sp.c.) is shorter; the upper labials (u.l/.w./’.) are quite cut through into four pieces; the lower labials (fig. 8, /./.) and the mandibles (mk.) are similar to, but still more solid than, those of the larger kind. The hyoid cornu (fig. 9, e.hy.) is extended out into an unciform stylo-hval (st./.) and the condyle (hy.c.) is narrow and oblong. The extra-branchial pouches (ex.br.) are similar, but the small cerato-branchials (c.b7.) are smaller still in this, and the second is confluent with the hypo-branchial plate (h.br.), as well as with its corresponding external bar. All things, however, taken together, the want of conformity between these two in the chondrocranium is much less than between many other kinds that are congeners ; the eye sees at once that they must belong to types very near akin to each other. This is shown, also, in the parasphenoid (fig. 7, pa.s.), but it has a long point at its fore end, and its basi-temporal wings are pointed, but the two bones are very much alike: the great difference is in the want of any trace of granulation in this. The fronto-parietals (fig. 6, fp.) are also quite devoid of granulation; and their size is in extreme contrast with those of the larval Calyptocephalus; they belong to a somewhat riper stage, and yet, instead of being large expanded plates, are mere styles of bone, lying like wall-plates on the edge of the skull, dilated a little on the inside where they will be more or less segmented into two bones, and on the outside, where they bind on the auditory sae. Third Family. “ DiscoGLossrpz.” First genus. Discoglossus.: 27, Diseoglossus-pictus,— Adult male ; 2$ inches long. South Europe. This is a true Frog, it comes next the Ranine, and above the Cystignathine species, but is modified by having a discoid tongue, dilated sacral apophyses, and opistheccelian vertebrae, The other Frogs that possess the first of these marked characteristics are * The sudden curve, upwards, of the pedicle (fig. 6, pd.) gives it the appearance of lying over the . 5th nerve at its exit; this is not the case, however, the actual end passes into the skull much lower down. ett DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 113 much less typical, and some of them are very aberrant and generalised. Hence it comes to pass that this is rather a motley group, and the question arises as to whether it should not be broken up and redistributed ; if this were done the boundaries of the groups that should receive their own relations back again, would have to be made more elastic. This type would have to go to the Ranide; Calyptocephalus to the Cystignathide ; Pelodytes would ask for admittance, either among the Bombinatoride, or the Alytide ; whilst Xenophrys, and the kinds agreeing with it, would probably have to be made into a new family ; that species is exceedingly generalised, and shows affinity with families far removed from the Ranidee, far more clearly than with those typical Frogs. My mention of these things is as a protest against the family group as it now stands: the Zoologists must re-arrange it as they see fit. As far as the skull is concerned, this type might have been left in the genus Rana, or still better, put with the species of Pyxicephalus ; with that of P. rufescens it agrees very closely. This is a typical semi-elliptical skull, whose breadth is to the length as 15 to 14; it is more frail, or less ossified, than in the average Frog’s skull; and its roof is imperfectly covered. The extent of the three regions of the skull is normal, the orbital being the largest, the nasal next, and the auditory the shortest. The occipital condyles (Plate 20, figs. 7, 8, oc.c.) are large, semi-oval, almost directly posterior, and separated by a narrow notch, The epiotic projections (p.s.c.) are gently convex, and come short, behind, of the condyles. The quadrate hinges are opposite the large twin nerve-passage (¢.¢., TX. X.), The small ex-occipitals (e.0.) are wide apart both above and below, and are at a con- siderable distance from the equally arrested prootics (pr.0.); these latter bones enclose the foramina ovalia (V.), run out to the facet for the pedicle (pd.), and up over the anterior canal (a.s.c.). Beyond these bones the tegminal region (fig. 7) is only half the width of the canal region, and runs only a moderate distance beyond the horizontal canal (h.s.c.). The skull is well roofed, for the main fontanelle ( fo.) is a rather small, long, heart-shaped space, and the small secondary fontanelles are wide apart on the large hinder part of the tegmen cranii. The temporal region is wide, and from thence the orbital part of the skull narrows up to the rudimentary superorbital projections (s.0b.) ; these increase the width a little in the ethmoidal region, but, as in Py«icephalus rufescens (Plate 14, figs. 1, 2), the cranial boat is very constricted in front. As in some small Ranide from Australia (e.g., Camariolius), the girdle-bone (eth.) is imperfect ; it occupies only about a fourth of the interorbital region at its widest ; its halves are united by a narrow isthmus below, and scarcely meet above. Most of the ethmoidal region, with its wings, and all the wide, well-developed, normal nasal region, are cartilaginous. Both roof and floor (figs. 7, 8) are wide, the septum (s.n.) is high and thick, the snout gently convex, with a slight prenasal bud; and the pro-rhinals (p.vh.) and subnasal outer angles are large and well MDCCCLXXXI. ; Q 114 MR. W. K. PARKER ON THE STRUCTURE AND developed. So also is the annular growth of the nasal wall (nw.), and the appendages (w./'..7?.) that finish the nostrils (e.7.). The inner nostrils (fig. 8, 7.7.) are very large, short oval, and almost transverse ; they turn a little forwards, within. The palato-suspensorials are well developed, the fore part wide, with a slight pre- palatine projection; the foot-like pedicle (pd.) projects well inwards, and the quadrate condyle projects well outwards, reaches back behind the stapes (st.), and is large and reniform, with a post-condylar lobe; there is no ossification beyond the setting on of the quadrato-jugal (q/j.). The palatines and pterygoids (pa., pg.) are quite normal, but remain wngrafted. The Eustachian passages (ev.) are only half the size of the inner nostrils, and are reniform. The mandible (fig. 9) is perfectly normal; it is rather high. The annulus (a.ty.) is large and perfect. The stapes (fig. 11, st.) is oval, emarginate in front, and has a boss. The interstapedial (7.s¢.) is nearly as large, is gently notched below, and its fore third is ossified. The medio-stapedial (m.st.) is a strong phalangiform bar, and ends in a perfectly normal spatulate extra-stapedial, with a cartilaginous supra- stapedial, confluent above (e.st., s.st.). The stylo-hyal is also confluent above, and the hyo-branchial apparatus (fig. 10) is perfectly Ranine. The investing bones are also quite Ranine, but the fronto-parietals ( fp.) are very narrow in front, and fail to cover the fontanelle perfectly; they are like those of Py«icephalus rufescens, but do not unite, behind. The nasals (7.) are normal, but wide apart; all the marginal bones are normal, but the squamosal (sg.) has an exceedingly long postorbital process. There are no septo-maxillaries; the parasphenoid (pa.s.) is exactly like that of a Rana, but the vomers (v.) are not; they are sub-quadrate, with short snags. Each bone touches the septum nasi, the spike in front of the inner nasal opening is suppressed, the front part ends in a spike, and the post-narial spike is short. The dentigerous elevations are large, oval, and oblique. This Frog differs from its narrow-backed relations, with forked tongues, in having :— 1. The whole skull feebly ossified. . The interorbital region very narrow in front. The main fontanelle left partly uncovered by the roof-bones. . The four bones of the hind skull small, and the girdle-bone imperfect above. . No septo-maxillaries. . The squamosal very long in its post-orbital region. NMIQa Fw Dw , The stylo-hyals and supra-stapedials confluent, above. Second genus. Pelodytes. 28. Pelodytes punctatus.-—Adult male; 13 inch long. Europe. The skull of this species is short, the breadth being to the length as 11 to 10; yet jt looks shorter than it is, on account of the great breadth of the nasal region. In DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 115 this respect it is intermediate between that of Plewrodema and Lyimnodynastes, types to which it can claim no near relationship.* But there are two European Frogs to which this type has some claims of relationship ; these, Bombinator and Alytes, however, are very exceptional forms, having “ opisthocce- lian” vertebree, whilst Pelodytes has them normal or “ proccelian ;” they have ribs, but this type has not (see Mivart, P. Z. 8., 1869, pp. 290, 291, and 294). Nevertheless, in spite of all that, I would rather put this near them—the natives of the same geogra- phical territory—than place it with any from a far country, having a great belief in the faculty of the Anura for modifying their internal structure, during secular periods. The occipital condyles (Plate 23, figs. 1, 2, oc.c.) are large, oval, and posterior ; the oceipito-auditory region is wide proximally, and narrow distally. The interorbital region is rather wide, lessens gently forwards, and bulges a little; the nasal region is very broad, a little longer also than usual, and then the snout is very transverse. In the endocranium the fontanelle ( fo.) is single and large, the roof-bones more than half hide it, and leave it of an hourglass shape. Half the interspace between the occipital condyles is unossified, and the super- occipital cartilage (above fim.) is of the same breadth; this is normal. But there is no distinction between the prootic and ex-occipital ( pr.o., e.o.) right and left; this I take to be a primary confusion of parts, and not due to coalescence ; if they have been separate at all, it has been secondary and temporary. The bony matter half encloses the foramen ovale (V.); half the interorbital region is unossified ; in this part the moderately large optic fenestra (II.) is seen. The other half is taken up by the girdle-bone (eth.), which simply ossifies its own ethmoidal territory in front, and only slightly passes into the wings; its axille are shallow, and its upper part loses half its extent at the middle, in bordering the great fontanelle. The nasal roof and floor (figs. 1 and 2, s.n., s.v.l.—put by mistake on fig. 1) are well developed, the former widest behind, and the latter in front. The “ pro-rhinals” (p.rh.) are small, and the angles of the floor large; there is a small “ prenasal rostrum ” running forwards from the thick septum nasi (s.7.). The palato-suspensorial arches are quite normal, but the fore part is very broad, and the hinder part rather feeble ; as in Dactylethra and Bombinator there is no palatine bone, and the pterygoid (pg.) is feeble like that of a young Common Frog ; it scarcely affects the cartilage to which it is applied, and a considerable pad of cartilage forms the facet of the movable “pedicle” (pd.). The re-entering angle of the pterygoid and suspensorial cartilage is a right angle, and in it the Eustachian passage (ev.) is seen—it is large, oval, and transverse. * Dr. GiinrHer, who gave me my specimen, predicted that I should find it a very generalised type ; this is quite true, for I am completely puzzled as to where it ought to be placed. The “ Family” in which systematists put it, namely, the “ Discoglosside,” is merely a “ Cave of Adullam,” to which all sorts of irregular, lawless, and aberrant types are relegated. Pelodytes and Calyptocephalus have proccelian vertebrae, whilst in Xenophrys and Discoglossus they are opisthoceelian. Q 2 116 MR. W. K. PARKER ON THE STRUCTURE AND The annulus (a.ty.) is of the typical size, broad, and its horns are not united. The stapes (figs. 2 and 4, st.) is large and oval, with its fore and upper edge but little bevelled. The medio-stapedial (m.st.) is a strong and somewhat sigmoid rod, with but little cartilage at its proximal end, where it wedges in—without any inter- stapedial segment—between the stapes and the vestibule. Beyond the bone the cartilage soon expands suddenly into a sub-peltate extra- stapedial, with a ligulate supra-stapedial confluent above (e.st., s.st.). The mandible (fig. 3) is normal, but very long and slender. The stylo-hyal end of the hyoid (fig. 2, st.h.) is sub-acute, and loosely attached to the vestibular floor ; it does not bind strongly round the hind margin of the Eustachian opening. Two-fifths of the band (fig. 8, chy.) at its lower end is reduced to a mere fibrous tract, and the hypo-hyals (h.hy.) are quite loosed from their proper stem and are confluent with the basal plate (fig. 3, b./.br.), into which they run by a ligulate stalk. This distal part is a very large ear-shaped emarginate leaf of cartilage, which, by its oblique, external angle overlaps the front lobe of the basal plate, a some- what smaller, more regular leafy growth, which runs almost transversely into the base by a shorter stalk. Behind this, on each side, there is the normal finger-shaped hinder lobe, a much smaller outgrowth of cartilage. At its root the fore end of the phalangiform thyro-hyal bone (¢.hy.) is set into the cartilage; this is terminated by an unossified lobe ; and the right thyro-hyal is considerably larger than the left. To the lower face of the proximal part of the right bone there is a curious splint applied : it is (V-shaped, with all its points sharp; the short stem behind, and the forks, which are crooked, run, the right forwards and outwards, and the left obliquely outwards in the other direction, under the root of the left thyro-hyal, This, which exists:as a splint, I take to be an abortive attempt to produce an ectosteal basi-branchial bone; it will soon be described again in Alytes, and I shall have to refer to it when describing a more normal (“endosteal”) basi-branchial in Bombinator, Diplopelma, Callula, Engystoma, Rappia, and Pelodryas.* The fronto-parietals ( fp.) are similar to what we find in many sub-typical Anura, they are wedge-shaped shells of bone, with sharp fore and dilated hind ends. By these bones the great oblong fontanelle is reduced to an hourglass-shaped space, only half as large. The rest of the investing bones are normal; the nasals come within a moderate distance of each other, aud are not very broad. With the upper labials— which are normal—there is, I believe, a small seed-shaped septo-maxillary. The pre- maxillaries ( px.) are widely transverse; the maxillaries (m.) thin and _shell-like in front, and sharply styloid behind; and the styloid quadrato-jugal (¢j.) is not evidently grafted on to the quadrate. The squamosals (sqg.) are slight but normal; they fail to cover the short unossified “teomen tympani” (¢.ty.). The parasphenoid (fig. 2, pa.s.) is normal, and large in * This is one among many instances in which the metamorphosis to which these ichthyic types have been subjected has only partially obliterated the form and structure of the Fish. i. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. ia the fore part ; its basi-temporal wings are narrow. The vomers (v.) are also normal, their dentigerous lobe is small and transverse; they strongly curl round the inner nostrils (7..), and have a moderate front lobe. I shall have to refer to this skull once and again; it seems to me to belong to an archaic type of the “ Anura,” and to be a form very difficult to place, zoologically. The main departures from, or failures in attaining to, what is typical, are as follows :— 1. There is only one fontanelle. . This is not covered by the roof-bones. . The ossification of the occipital arch and ear-capsules is generalised and continuous. . The nasal region is very dilated. m CO bo Or . There are no palatine bones. 6. There is no inter-stapedial, the supra-stapedial is confluent above, and the extra- stapedial is sub-peltate. 7. The hyoid arch is very feeble, partly absorbed, and has its dislocated distal ends extremely dilated. 8. The front lobes of the basal plate are similarly dilated. 9. The thyro-hyals are non-symmetrical. 10. There is an irregular forked superficial basi-branchial on the under surface of the basal plate. Third genus. Xenophrys. 29. Xenophrys monticola.—Adult male ; 3 inches long; length of hind leg, 43 inches. Darjeeling. This fine specimen, the gift of Dr. GUNTHER, is twice as large as the one described by him in his ‘ Reptiles of British India’ (plate 26, fig. u, H’, p. 414). That specimen only measured 19 lines from snout to vent, with a hind limb 31 lines long ; this gives nearly double these lengths. Moreover, this appears to be much better developed; it has a rudiment of the “interdigital membrane,” and the fingers and toes have a discoid end one-third wider than the contiguous part of the digits. I am very doubtful as to the position of this large Frog, and of its equally fine relative, Megalophrys montana (ibid., p. 412). The figures of the skull of Xenophrys (Plate 23, figs. 5-10) have been purposely put on the same plate with those of Pelodytes (figs. 1-4) for contrast. The Family “ Discoglosside” (GUNTHER, “ Batr. Sal.,” p. 34; and Mivart, Proc. Zool. Soc., 1869, p- 294) must be one in which the members “agree to differ” to a very great extent, especially now Professor Mivart has added the “ Asterophrydide ” of Dr. GUNTHER, and Professor Copr’s Neotropical genus Grypiscus, which has mandibular teeth, and heads a “ Sub-family” of its own. This type has opisthoccelian vertebrae. I shall look for its true relations among the “ Oriental” Batrachia, and these, when 118 MR. W. K. PARKER ON THE STRUCTURE AND found, will most probably differ from it in one or other unimportant character, and yet correspond with it in non-essentials. This is a short skull (Plate 23, figs. 5 and 6); the breadth is as to the length as 5 to 4, but the length is increased externally by the great retreat backwards of the quadrate condyles (q.c.), which make it, if measured by them, one-tenth more. It is also an extremely flat skull, and has a decurved snout; it is very much unlike the skull of any of the “ Ranidee” known to me, and the crania that answer to it best are not yet described. I must therefore give an account of it as it is, and then refer to it afterwards, when its image appears again in other types. The occipital condyles (oc.c.) have large, oval, posterior faces, but they project very little and are separated by a gentle emargination. With many deficiencies that make it lie some depth below the “norma,” this skull is ultra-Ranine in its general form, for it is extremely depressed, the otic regions are wide wings, the jugal arch strongly bowed, and the hinge of the jaw is carried far behind the occipital articulation. The skull is of full width in the temporal region; then it does not become narrower, but much broader, towards the antorbital region, The margin above has a concave, and not a convex, outline; the walls do not bulge, but are scooped (fig. 6). The dilated ethmoidal region is made still wider by superorbital expansions (so.D.), and in front the whole nasal territory is but little narrower than the ethmoidal. Thus the fore half of the cranium is in remarkable contrast with the hind half, and its copious cartilaginous matrix is very similar to that which is seen in Skates and even in the Chimera; to make it still more archaic, the prenasal (p.v.) is a thick decurved beak, such as is seen in an early stage of the embryos of many Vertebrata. The ossification of the occipito-auditory region is generalised, bemg continuous on the same side (p7.o., e.0.); the synchondroses above and below (f:m.) are rather wide and sub-equal. The fontanelle is quite covered ; it is single and lJanceolate, with the narrow end forwards; the tegmen, behind, is short; in front, it rather appears, than is, extensive, for the ethmoid (eth.) is of great extent. The occipito-auditory bones (pr.o., e.0.) nearly reach the fontanelle, under the roof-bones, above ; below, they touch, over the parasphenoid, but leave the vestibular floor unossified. Above (fig. 5), the bony matter leaves merely the tegmen tympani soft ; it is very narrow fore and aft. In front, below, the bony matter almost encloses the foramen ovale (V.), and from thence half the interorbital region is cartilaginous in this tract; behind the middle of it is the large oval optic fenestra (II.). The girdle-bone (eth.) occupies more than a third of the endocranium ; it borders nearly a third of the fontanelle, and runs somewhat into the proper nasal territory. Below (fig. 6), it affects the wings a little ; above (fig. 5), its edge is rounder ; it runs quite clear of the superorbital “eave.” A cross-shaped expansion, below, marks off the shallow axils of this depressed bony mass; these fossze are large, but shallow. The cross is made more apparent by the extension forwards of the bony matter into the DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 119 hind part of the septum nasi (s.n.), and the hiding of the outer part of the bone by the vomers (see figs. 6 and 9). Contrary to rule, the upper and lower outlines of the nasal region are very similar. Above, a large and shallow, and, below, a narrow and deep, notch on each side, shows where the ethmoidal territory ends and the nasal begins, and these emarginations show that the nasal region is short ; the thickness of the convex roof hides the form of the dividing septum nasi. Below (figs. 6 and 9), the girdle-bone ends in a sinuously- rounded margin, a short distance behind the notches in the sides. The bone is sinuous on its surface, first scooped in the middle, then swollen on each side, a groove sub- « transversely dividing these swellings from the “wings.” In front of the notches, the subnasal laminee (fore part of trabecular cornua, s.7./.) are ossified, each bone taking up one-third of the floor. A bone half the size of these has been formed in the cartilage of the median tract ; it reaches from the nerve-passages (7.7.) to the decurved rostrum (p.r.). These three bones, like the ethmoid or “ girdle,” being ossifications of the ’ three basi-cranial bars—like those found in so many types in the region of the brain cavity—may be entitled to the name of “serial homologues” of such segment-forming bony centres.* The lateral subnasal ossifications are rugged and prickly at their margins ; their ce middle is somewhat elevated, transversely ; the ‘ pro-rhinal process” of the cartila- ginous matrix (p.7h.) is small. The upper surface, formed mainly (7.e., except at the middle line) by the nasal roof- cartilages (fig. 7, al.n.) is broad, gently convex, and passes insensibly into the prenasal rostum (figs. 5-7, p.n.). The suborbital space (fp. to pg.) is large and sub-oval ; its outer fence is the palato- suspensorial arch. The ethmo-palatine bar, like the rest of the cartilaginous pith of this arch, is slender, and runs outwards and forwards, like a continuation of the wing of the ethmoid; it is bi-aculeate in front (fig. 9, e.pa.). The post-palatine portion (behind e.pa.) can be seen to be but little affected by the palatine bone (fig. 6), which is a normal sub-faleate blade, separated by almost its own length from its fellow, and not reaching the pre-palatine spur in front. But the pterygoid (pg.) either ossifies or conceals much of the hinder part, yet the inner fork or pedicle (tig. 5, pd.), and the tract leading to it in the edge of the folded pterygoid (fig. 7, pg., sp.) shows that the axis is nowhere quite lost. But on the under surface (fig. 6, pd.) the pedicle is seen to be tied down by the pterygoid bone so that all motion is lost, and this inner tongue of the bone forms a strong squamous suture with the prootic and parasphenoid ; this is a Bufonine character, as we shall see. But the outer fork, or quadrate region, becomes much larger, and the part above the condyle is largely ossified by the quadrato-jugal (figs. 6, 7, q., q-¢-, QJ-)+ * In Birds, ossifications of this kind in the precranial region of the base are very common; but in their * compressed prognathons head the fore part of the intertrabecula is early absorbed; the paired trabecule end behind the intertrabecula and the ‘ subnasal laminw”’ are only exceptionally developed. 120 MR. W. K. PARKER ON THE STRUCTURE AND The condyle for the mandible (q.c.) is large, oblique, and reniform, and has the inner trochlea twice the size of the outer; the hinder and outer fork of the pterygoid bone runs vertically, splint-like, inside this far-retreated part of the suspensorium. The large Eustachian opening (fig. 6, ev.) is, by this, made to turn backwards, as well as outwards ; its hind margin is a thick fibrous ligament. The mandible (fig. 7) is quite normal and of great length, answering to the extent of the gape; the mento-Meckelian part (m.mk.) is large, and so is the cylindroidal condyle (a7.c.) ; the cartilaginous bar (mk.) is but little affected by the trough-shaped “articulare” (ar.); the dentary (d.) is a little more than half the length of the mandible. The “annulus” is very remarkable ; it is large (Plate 23, fig. 7, a.ty.), thick at the edges, oblong in shape, and whilst one horn is attached to the fore part of the portico formed by the squamosal (sg.) over the tympanic cavity, the other horn passes behind and under the hind part; its position is oblique, being carried backwards, below, by the suspensorium and its splint (q., sq.). The stapes (fig. 10, st.) is thick and of a short-oval shape, with the fore margin emarginate. The columella fits by its thick, bilobate, scooped apex, within and around the fore part of the stapes. The inter-stapedial (7.st.) is represented by the larger, unossified lobe ; but it is doubtfully segmented. The medio-stapedial (m.st.) has also some cartilage on the lesser lobe, and the thick bony end carrying the cartilage is almost discoidal. The narrow shaft is bent almost at a right angle on the dilated proximal part, and runs more than half way to the distal end. That end, gradually thickening to its middle, is the extra-stapedial (e.st.) ; at first it is merely a continuation in the same line as the shaft, but its distal two-thirds is bent down at a little more than a right angle, is thick below, and cochleate above. There is no supra-stapedial. This generalised, but large, columella is seen (fig. 7, e.st.) to emerge from the cleft, and then to pass, downwards and outwards, inside the membrana tympani. So far are these parts carried outwards, backwards, and downwards, that in the upper view (fig. 5, a.ty.) they are scarcely seen when the eye is focussed to the upper face of the skull; hence the apparent minuteness of the annulus in the figure showing that aspect. I could find no cartilage in the hyoid arch from the Eustachian opening downwards until I reached the hypo-hyal region. There (fig. 8, /.hy.), there is a faleate hypo-hyal lobe which passes backwards and inwards to a basal plate of normal size (b.h.br.), and with the usual small posterior wings, and bony, divergent thyro-hyals (¢.iy.); but there are no anterior wings to the basal plate. Here the absorption of the hyoid band is equal to what is seen in Pipa. The investing bones, like those of the endocranium, are very unlike what we see in Rana, and in the “ Ranidee’ The fronto-parietals (fp.) Just overlap the fontanelle and the side-walls of the cranial cavity ; they dre almost square over the hind part of that cavity, and then expand , generally. forwards to their end. Their hind margin has a pair of shallow notches, and their fore DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 121 margin is cut away, but more irregularly, into four shallow emarginations. The ribbed orbital edges are higher than the middle, so that the top of the skull is gently con- cave ; behind the superorbital edge there is a temporal notch. In front, these bones scarcely cover in the fontanelle ; behind, they almost reach the foramen magnum. The nasals (figs. 5 and 7, 7.) are curious stalked shells of bone; the stalk is thick and is pointed below, the dilated part fits on to the convex cartilage, does not come near its fellow, and has a strongly crenate margin; much of the middle and all the fore part of the cartilaginous roof is exposed. In this very “ Phryniscine” skull (see Plate 41) the przmaxillaries are thrown beneath the prenasal rostrum (Plate 23, figs. 5-7, pw.) ; they extend outwards a good way, but are narrow, have a small palatine process, and an average nasal process. As with the dentary below, so with the premaxillary, above, the bone is grafted on to a labial cartilage, which cartilage it largely ossifies; in this case it is the “inner upper labial” (w./'.). All the cartilages of the outer nostril are large and solid—we shall see the like again in the “ Pelodryadidee ”—and whilst above the opening (e.7.) the nasal roof (al.n.) is very thick and crescentic, the front of the nostril is guarded by the large “outer upper labial” (fig. 7, w./?.), which is pedunculate and tridentate. The maxillaries (mx.) are long, narrow bones; the jugal process reaches far back ; and the quadrato-jugals (q.j.), which are half as high as the maxillaries, are rather long: they are extensively fused with the quadrate (q.). The real height of the oblique backwardly-turned squamosal (sq.) is equal to half the length of the skull. Its supratemporal part (fig. 5) is extensively developed over the auditory capsule, as well as over the marginal “tegmen ;” the bone is cleft, as it were, to bind fore and aft upon the parotic mass. The postorbital process (figs. 5 and 7) is faleate; the edge of the temporal part forms a crescentic porch to the ear, and the descending stem is a narrowish lath of bone binding on the retreating suspensorium. The parasphenoid (fig. 6, pa.s.) is large, well-developed, and Ranine; its short “handle” is split, its basi-temporal “guard” is strongly sutured to the pterygoid (where it les over the pedicle), and the pointed “ blade” reaches as far as to the wings of the ethmoid ; the trough of the bone is deep (fig. 7, pa.s.). The vomers (fig. 6, v.) are remarkable ; they are thick, and bent like “ knee-timbers,” are wide apart, have a solid boss for the teeth, an out-turned front lobe which reaches the pre-maxillary angle, and a round notch for the internal nostril (¢.7.), which passage is guarded by a long spike of bone on the outside. As compared with the skull of the typical species, this is very abnormal; its most important modifications are :— 1. The extreme flatness of the skull. 2. The great extension outwards of the parotic wings. 3. The extremely backward position of the quadrate articular condyles. 4. The fusion above of the occipito-auditory bones. MDUCCLXXXI. R 122 MR. W. K. PARKER ON THE STRUCTURE AND 5. The absence of the secondary fontanelles. 6. The increasing width, forwards, of the cranium, involving the great breadth of the girdle-bone. 7. Superorbital projection of the ethmoid cartilage. 8. The width, convexity, and solid rostral end of the nasal region, 9. The three anterior subnasal ossifications, 10. Slightness of palato-suspensorial arch, with its fixed pedicle. 11. Bony quadrate. 12. Oblong, oblique, open annulus tympanicus. 13. Generalised columella, without supra-stapedial process or infra-stapedial segment. 14. Absence of hyoid band and of antero-lateral lobes of basal plate. 15. Inferior position of pre-maxillaries. 16. Reversed shape of fronto-parietals. 17. Extensive double growth of supratemporal region of squamosal. 18. Solidity and arcuate form of vomers. 19. Massiveness of upper labials, and union of the inner pair with pre-maxillaries. 20. Absence of septo-maxillaries. These twenty marked characters of difference from the type show how far removed this Frog is from the Ranidze; its generalised nature is so great that I shall have to refer to it for comparison when I come to the Polypedatide, the Hylide, the Phryniscidee, the Engystomide, and the Hylapleside. Fourth genus. Calyptocephalus. 30. (A) Calyptocephylus Gayi.—Adult female; 54 incheslong. Chili. (A) The adult.—This large, massive, and almost ganoid skull (Plates 21, 22) is, on the whole, very Ranine, but it has retained a copious overgrowth of granular sub- cutaneous bone, which makes it like a strong rough helmet. Its general form is half a short oval ; it is similar in outline to rather more than the narrow end of a short Hen’s egg; for its outline narrows in again behind. Compared with its length (from snout to occipital condyles) the breadth is as 44 to 3, and the quadrate hinges project 2 lines beyond the occipital hinges, making the skull one- eleventh longer at the outside. The actual measurements are-— 1. Greatest breadth, 2 inches 5 lines. 2. Length from snout to occipital condyles, 1 inch 9 lines. 3. Extent of quadrate hinges behind them, 2 lines. This skull carries the peculiar Batrachian type to its greatest degree, viz.: the great width as compared with the length. So great is the development of the investing bones that the endocranium is almost DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 123 entirely roofed over by them; and the orbital spaces, so large generally, have only one-third the extent above that they possess below. The endocranium is not intensely ossified, and there is no increase of the number of bony centres either in it or in the external plates such as is seen in Rana pipiens ; but it agrees with that large Frog in the moderate ossification of the inner skull, and in the extension backwards of the condyles for the lower jaw. But on the whole, leaving out of view the dense overgrowth of the subcutaneous bones, this skull differs less from that of Rana tigrina and R. hexadactyla (Plates 6 and 7) than it does from the skull of R. pipiens. — It agrees with that of the Oriental Bull-frogs in the form of the endocranium ; in the relative size of the cranial trough (much wider and larger than in R. pipiens), and in the severely typical paucity of bony centres, whether external or internal. On the whole, there is little in this skull of divergence from the “norma,” except the exuberant growth of its roofing plates ; and in this genus, and also in Ceratophrys, it is in this overgrowth, and not in indefinitely numerous bony patches, that we have the mark of ancientness: the great size of the species, however, is in itself suspicious in a group whose members are for the most part small, and even very small. The occipital condyles (Plate 21, figs. 1, 2; and Plate 22, fig. 1, oc.c.) are large, oval, postero-inferior in aspect, and separated by the basioccipital cartilage, which is gently emarginate, and one-third of the width of one condyle. The superoccipital arch (Plate 22, fig. 1) is excavated above, and less than a line in front of the basal outline (Plate 21, fig. 1, fim.); it is not finished with bone, for there is a considerable wedge-shaped tract of superoccipital cartilage. Both the auditory and ethmoidal regions stand out at right angles to the axis, and the former is twice the extent of the latter. The interorbital region between these is of the average length, and twice as broad as in Rana pipiens (Plate 8); it narrows gently up to the ethmoidal region, and is almost straight-sided, not bulging, as in many types. The depth (fig. 3) is proportional to the width, and the sectional form is boat- like, gently widening upwards. The foramen ovale is partly rimmed in front by the prootie (fig. 2, V., pr.o.); the optic fenestra (II.) is of the medium size and is midway in a tract of cartilage that occupies half the interorbital region. The girdle-bone (eth.) takes up the rest; it spreads some way into the ethmoidal wings, but does not reach quite to the front of its own proper territory, nor ossify the limited superorbital lamina (s.ob.). Behind, the ex-occipitals and prooties (fig. 2, e.0., pr.o.) are separated by a triangular tract of cartilage, and do not extend outwards as far as the parasphenoidal wings ; thus the lesser vestibular and the larger tegminal tracts (fig. 2) are left soft. The tegmen tympani (t.ty.) is largely extended outwards, beneath the over-arching squamosal (sq.), as a flat, sub-quadrate, pterotic tract ; the vestibular floor is gently convex, and is confluent on its outside with the dilated stylo-hyal end of the hyoid bar (st.h.). yas] Lo 124 MR. W. K. PARKER ON THE STRUCTURE AND In the nasal region there is no rostrum to the septum (s.n.) in front; the pro- rhinals (fig. 2, p.7h.) are broad, short, pedate, and they fill the space between the nasal and palatine processes of the pre-maxillaries (pz.). The outer angles of the subnasal laminze (s.n./.) enter a similar space in the fore end of each maxillary (mz.) ; the lamin themselves bulge downwards, and the septum ends below on a higher plane (fig. 2, s.n.). Both these inferior (trabecular) plates and the nasal roof (al.n., al.sp.) are moderately developed, and the latter can be seen to project beyond the fore end of the nasals (figs. 1 and 3, al.n., n.). The small round outer nostrils (figs. 1 and 3, e.n.) are unusually near together, are formed in a scooping of the nasal roof—or rather that roof runs round them, and they are protected by the large, solid inner and outer “upper labials” (w./.w.P?.). The inner nares (fig. 2, 7.n.) are twice as wide apart as the outer, twice as large, oval, and directed cross-wise ; they are quite included in the space between the middle process of the vomer and the ethmo-palatine cartilage and bone (pa.). That cartilage (pr.pa.) is of the usual form, and that bone is not more than usually developed ; it is straight in direction, but sigmoid in form, with a dilated (external) blade. The post-palatine region of cartilage is unossified for a considerable extent, and as it passes into the pterygoid region is at first only covered on its inner side by a delicate stylo of bone, for the pterygoid ( pg.) in its fore part, whilst it still clings to the cheek, is aborted considerably by the huge maxillary (fig. 2, mz.). Becoming free, before it forks, this bone suddenly enlarges ; then in front of’ the Eustachian opening ( pg., ew.) it sends, straight inwards, a large tongue of bone having an ectosteal relation to the pedicle, which reaches almost to the foramen ovale (V.). By this process, as in Xenophrys and Bufo, this partially absorbed dorsal end of the pier of the mandible is hidden and fixed; we miss here the gliding joint seen in Rana, and many other types. The outer and hinder fork of the pterygoid passes backwards, and then outwards, in a geniculate form, to clamp the inner face of the quadrate region of the suspen- sorium (q.). This region, ossified below by the quadrato-jugal (q,j.), ends in the large “ trochlea ” for the lower jaw (q.c.); it is a very elegant pulley with a large inner, and a smaller outer longitudinal convexity, and a deepish concavity between. This part, with the condyle, the practical suspensorium, but only made out of the distal end of the larval cartilage (see Plate 22, figs. 2, 3, q.), is directed almost equally outwards, downwards, and backwards (Plate 21, fig. 1-3, q.). The large, oval, Eustachian passages (ew.) are turned a little backwards, and not ” directly outwards ; the “annulus” (fig. 3, ty.) is of medium size; is open above, and is a rather broad band. The mandible (fig. 3) is quite normal ; it is a strong bar, with a large cylindroidal condyle, with a moderate coronoid crest to the articulare (a7.), an unusually long and high dentary (d.), and a thick, short mento-Meckelian bone. The stapes and columella (figs. 2-6), partake of the general stoutness of the skull; the DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 125 former (st.) is a solid, sub-pedunculate mass, sub-oval in form, but scooped to receive the columella. This latter part has no separate inter-stapedial segment ; but there is a large thick unossified lobe to the end of the medio-stapedial (m.st.) which represents it. Between this and the main stem there is a deep notch, faced fore and aft with cartilage, then the bony bar gradually lessens, is arched, and ends in a terete rod of cartilage which dilates into the thick, oval, bi-convex extra-stapedial (e.st.). From the inner face of this lobe there arises, at a sharp angle, the terete, stout supra- b] stapedial (s.st.), which is confluent with the “tegmen ” above. The stylo-hyal end of the cerato-hyal (figs. 2 and 4, st.h., chy.) seems like a flabelli- form continuation of the vestibule; it is wide at first, then narrows as it passes directly outward, margining the Eustachian opening; is then wider again as it creeps along the inside of the quadrate, and after this becomes a little less again, as it passes, sigmoid in form, to the hypo-hyal region (h./y.). There is no enlargement, but it forms a round loop, and then gently dilates as it passes into the basal plate (fig. 4, b..br.). The antero-posterior extent of the basal plate is small, as in the “ Hylidee ;” its lobes are large, the foremost flabelliform, the hindmost—here quite as large—is stalked and emarginate. There is no bone, save in the “thyro-hyals” (¢.hy.); they are of moderate length, divaricate well, are stout, and soft-footed. This skull, with its strong roof and sides, is constructed, externally, of seven pairs of bones, and an odd one; in the young the main roof-bones were divided across, which gives seventeen “parostoses” originally investing the proper endocranium ; the plates applied to the cartilage, and grafted upon it, are not counted, but the two “ dentaries ” in the lower jaw bring the sum up to nineteen. The other bony tracts in skull and face amount to seven pairs and an odd one—that, however, the girdle-bone, was double once, so that there are sixteen bony centres that are properly “ endoskeletal ;” in all, there were thirty-five ; there are thirty-two bones in this skull. But the ‘“quadrato-jugal” is grafted on the quadrate and becomes partly endo- skeletal, whilst, on the other hand, the bony centres that were formed in the chondro- cranium (ex-occipital, prootic, &c.), are largely recruited, in their growth, from the perichondrial layers of membrane, thus they link on to the palatine and pterygoids that begin in membrane, and get their “‘endosteal” additions afterwards. But the ? truly parosteal “sub-species” of bony centre concerns us now; I shall describe this group, and then show what was the form of the original “chondrocranium,” on and mm which this strong architecture is perfected. Behind, the roof-bones come close to the edge of the foramen magnum (/7.), and are built across from side to side, without interruption, over the condyles of the quadrate region; the hind margin of this wide tract is emarginate, crescentically. The nasal, frontal, and saggital sutures form one line of division from snout to occiput, and each region of this long suture is. about equal. The two pairs of main 126 MR. W. K. PARKER ON THE STRUCTURE AND roof-bones, viz.: the parietals and frontals, are now only one pair (/:p.), the coronal sutures having early disappeared. This double fronto-parietal tract is twice as wide over the occiput as over the orbits ; its hind margin, like its front margin, is almost straight, but the edge on each side uniting it with the squamosal (sq.) converges forwards, and is a roughly dentate suture—this answers to the squamous suture in Man. Each squamosal (sq.) is a bone with a descending and retreating handle, and a roughly pentagonal blade; these are the temporal and pre-opercular regions of the bone. The pointed fore end of the temporal plate —its postorbital projection—is united by suture with the upper edge of the maxillary (fig. 3, mw.); this is a sinuous suture, and behind it the edge of the squamosal rises and falls twice: first outside the pterygoid (pg.), and then over the “annulus tympanicus,” forming the ledge under and to which it is attached (fig. 3). The preorbital rim is formed by the nasals (n.), these are very large, about half the size of the fronto-parietals. In front of their concave preorbital edge they form a large descending pedate process (fig. 3), which rests upon, and is united by suture with, the fore end of the maxillary, on each side. The median spur of each nasal is blunt, and the fore edge has a large round notch, exposing the nasal roof (al.n.), which is itself notched for the nostril (e.n.). The main roof-bones dip into the orbit (fig. 3, fp.), forming an orbital plate. The arcuate premaxillaries ( px.) form a divided but strong key-stone to the great arch of the face; their palatine and nasal processes are well developed aud normal. The maxillaries (mz.) are notched in front, where they articulate—without the inter- vention of a septo-maxillary—with the premaxillary. The lower dentary edge of the bone is gently arched, and the jugal process reaches as far back as the tegmen tympani (¢.ty.). Half the depth of the side of the skull is formed by the maxillary, whose edge is cut away below the eye, and is again deeply notched behind the postorbital suture; it then becomes the rapidly lessening jugal process. The quadrato-jugals (q,j.) are one-third the length of the maxillaries, but are only the size of the jugal processes ; they strongly bind on the lower part of the squamosal (fig. 3), and are grafted largely on the quadrate, and there pass under the squamosal ; this is best seen in the palatal view (fig. 2). That view also shows how large the palatal plate of the maxillary is, especially where it is locked to the pterygo-palatine arch (pq.). The parasphenoid (figs. 2 and 3, pa.s.) is a large bone, and reaches to within a short distance of the foramen magnum, behind, and to the beginning of the unossified septum nasi in front. It projects by a very obtuse angle behind, its “ basi-temporal wings” are broader ee ll SS ee ae DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 127 than the main bar, and end by a ragged oblique edge, which runs to a point in front, reaching nearly to the Eustachian opening (ev). The middle of this hind part is raised into a lozenge-shaped convexity, but the main bar, which is three-fifths the width of the trough of the cranium, is itself a deep trough. Between the foramina ovalia (V.) it is compressed a little, is widest between the optic nerves (II), and then scarcely lessens to its rounded ragged fore end, where it reaches the thick bosses of the vomer (v.) that carry the teeth. Those bones (fig. 2, v.) are very thick in their dentigerous portion, where they form a lobe that runs outwards and a little forwards ; the thinner part forms a strong spike in front of the internal nostrils (7.7.), and a broad diverging lobe in front of the spike, which reaches the maxillary under the angle of the “subnasal lamina” (s.7./.). This cranial building is thus finished below and above ; its roof, floor, side-walls, partitions, chambers, and outworks are all essentially such as we see, on a small scale, in the Common Frog. But the great size of this skull, and the extension under the skin of thick dense bone, ornate with a honeycomb pattern of hollows and papulate ridges passing into pearly grains, makes it in appearance very unlike its simple “norma.” These modifications may be classed under one head as modification No. 1. 2. The upper fontanelle appears to be single. 3. The long inner process of the pterygoid aborts the jovt of the pedicle. 4. There are distinct superorbital ledges. 5. The quadrate is largely ossified by the quadrato-jugal. 6. There is no inter-stapedial segment. 7. The extra-stapedial is suborbicular, and its supra-stapedial process is confluent above. 8. The stylo-hyal is confluent above. 9. The basal plate is very short, the hypo-hyals have no lobe, and the hind lobe is very large and emarginate. These are the few, and for the most part gentle, differences between this skull and that of the norma. 30 (continued).—(B) C. Gayi.—Larva; 4? inches long; tail, 2? inches ; hind legs, % inch. Chill. In the large Tadpole of this species the chondrocranium (Plate 22, figs. 2-5) is seen to differ in many points from that of the larve of the American Bull-frogs (Plates 3 and 4), and of Pseudis. It may be compared with the same (my first) stage in Rana pipiens (Plate 3, figs. 1-3), Rana clamata (Plate 2, figs. 5-7), Pseudis paradowa (Plate 2, figs. 1-4), and with a somewhat more advanced stage of Cyclorhamphus culeus (Plate 22, tigs. 6-9), and with the larval skulls of Cystignathus, Hyla, Bufo, &c., in other Plates. Its true relationship will be at once seen by a comparison of the figures on this 128 MR W. K. PARKER ON THE STRUCTURE AND Plate (22) with the others. Cyclorhamphus is its nearest ally, by far, and in this comparison we have dug deeper for evidence of kinship than in any comparison that could be made of the skull in the adult. Small hind legs have budded out in this larva, and as a correlate of these growths three investing and two intrinsic bones have appeared in the skull; these lift it out of the sphere of simple “ chondrocrania.” The notochord (nc.) is still of considerable size, but the copious development of hyaline cartilage has already obliterated several morphological landmarks. The occipital condyles (0c.c.) are now well formed, and the arch itself is very wide, rather low, and has a huge doorway (j.m.). The large oval ear-sacs are completely chondrified, and are confluent with the basis and tegmen cranii; they abort its walls largely, growing into the sides of the skull; the old lines of junction can, however, be traced. The cranium proper is very wide, especially behind, and the tegmen is developed, already, nearly as much as in Sharks—more than in most Skates ; it has two lesser fontanelles in it (fo’.). The interorbital part is wide, gently bulging, and gradually narrowing to the ethmoidal region, which is now in the act of closing in upon the fore part of the brain capsule. Above (fig. 2) the principal fontanelle (/fo.) is only a quarter the size of the general tegmen of cartilage, and in front it is unenclosed—the lateral halves of the primary ethmoidal region of the skull-wall have not met over the brain cavity. Also the unfinished state of the skull is seen still further in this, namely, that the “perpendicular ethmoid” (p.¢.) is still a mere oval upgrowth of cartilage—a tuberous ascending development of the intertrabecular tract which is still visibly distinct from the trabeculz, all the way from the pituitary floor to where the trabeculze become again free in front as the “ cornua” (c.t7.). In front of that little swelling of cartilage (fig. 2, p.e.) the foundation of the septum nasi is seen in the narrow foremost part of the cartilage which conjugates the two trabeculee. We get here additional light upon the vegetative growths of cartilage that finish, by closing in, the cranial capsule, in front. Where the walls have most converged in front they there suddenly grow out as the “ethmoidal wings ” (al.e.), running by con- tinuous growth of cartilage into the pterygo-palatine bands (p.pg.). The olfactory nerves (1) escape here from inside the thickening wall which gives off ” the wings; in the middle the mesethmoidal “ tuber” (p.e.) becomes a flatter structure, and grows into a vertical partition wall between the nerve-outlets, the foremost of the three sense-capsule fenestre. The right and left roofs—growths of the ethmoidal wall—unite with each other and with the middle wall ; and besides this, in front of the cranial cavity, each trabecula at its outside develops a process of cartilage, similar to, but less than, the median intertrabecular tuber. These three, growing round the olfactory nerves, finish the round passages (fenestree), and, with the lateral ethmoidal walls (growing into the roof), close in the skull-building. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 129 Directly in front of these ethmoidal crests the trabeculze are at their narrowest, are closest together, their scooped edge forming the inner margin to the internal nostrils (i.n.) ; the complete hind rim is formed by the erescentie fore edge of the pterygo- palatine band (p.pg.). In front, the quadrate part of the suspensorium (g.) and the trabeculee (c.t7.) of the same side each send out a bud (pr.pa.), and these approaching growths are united by a fibrous band which finishes the nasal opening. The cornua trabeculee maintain their average width up to the frontal wall, but are narrower and are longer than usual, diverge very gently, and end in a moderately pedate process, which is turned outwards. The nasal roofs are not yet chondrified, the eyeballs have been removed with the other soft structures, but the auditory capsules are shown, for they are built into the walls of the cranium as side chambers. Below (fig. 2), there is no columella at present, but there is an open space, and into this space—the fenestra ovalis—the oval “ stapes ” (fig. 2, vb., st.) fits. The roof of the future tympanic chamber (t.ty.) is very largely developed ; outside the large horizontal canal (/.s.c.) it sends out a small hinder, and a large front, spur ; the latter is the confluent “spiracular cartilage” (sp.c.)—the “annulus” that is to be; it just touches and overlaps the otic process (a free elbow now) of the suspensorium (ot.p.). The anterior canal (a.s.c.) is covered with bone (f:p.), the posterior (p.s.c.) is naked, and bulges gently above and behind. A small circular plate of bone, the ex-occipital (e.0.), has grown inside the double nerve-passage (IX., X.) and encloses and subdivides it externally. The roof is covered to a great extent by the fronto-parietals (fp.) ; these are many times larger than those of any other larva at this stage dissected by me (see Plates 2, 3, 4; also Plate 22, figs. 6, 7). The two bones are very granular and rugose, quite unlike what is seen in the other Tadpoles, they, together, have an emarginate bow-shaped hind edge spread over the large ear-capsules up to the horizontal canal ( h.s.c.), which they partly cover, and then occupy two-thirds of the interorbital region of the roof. Each bone is deeply notched on the inside, half-way ; this is the beginning of the segmentation into two bones; the space thus left bare is roofed by the front part of the hinder tegmen. Behind the notch the bones come closer together, ready to form a sagittal suture, in front they diverge to their pointed end and have a rounded inner edge. But the parasphenoid (fig. 3, pa.s.) is also rugose, pitted, and shows signs of a “ganoid ” nature ; this is lost in the adult, as if the larva retained more of the nature of some ancient armed type than the adult. The bone is similar in form to that of Cyclorhamphus (fig. 7), and not much larger, relatively ; it is very large in both kinds; it is broader altogether, and notched at the tip; and the basi-temporal wings are, like the tip, less acuminate than in the other Tadpole (fig. 7). This condition of the roof and floor-bones in a larval Batrachian just putting forth its hind legs is of extreme interest ; and the size, massiveness, and sculpturing of these MDCCCLXXXI. s 130 MR. W. K. PARKER ON THE STRUCTURE AND bones is surprising when compared with what is seen in the large Tadpoles of cognate “ Ranide.” The nerves figured as issuing from the skull in this view are the Ist (I.), the optic (II.), the trigeminal (V1, V**.), the facial (VII.), with its vidian branch (VIT".) which runs forwards to the upper labials; the 8th is out of sight; the 4th and 6th were not drawn, and the 9th and 10th (IX., X.) are indicated by their twin-passage. ” The “palato-quadrate” arch in this Tadpole is full of interest (figs. 2 and 3, p-py., pd., q-); it is very massive, and its hinge is opposite the beginning of the terminal third of the cornua trabeculee (c.t7.). Altogether these bars are twice as wide as, and much thicker than the trabeculz, and every thing about them is large and solid. They are almost parallel with the skull, converging very little ; the pedicle (pd.) is short, for the elbowed and forwardly turned bar is of great width and takes up the room; its “otic process” (of.p.) is a rounded bud overlapped by the spiracular cartilage (sp.c.). The main bar becomes half its former width near the pterygo-palatine conjugation (p.pg.) ; there it sends up over the temporal muscle that large, outwardly rimmed scooped and sessile leaf, the “ orbitar process” (or.p.), which at its base has the infero-external crescentic hollow for the condyle of the hyoid (hy.f.). As yet the pterygoid band is not free from the ethmo-palatine, but the palatine region of this great arch has all its three sub-regions well marked out. The first of these is ¢ that which runs into the trabecula, namely, the ethmo-palatine (p.pg.), it is half as broad as the trabecula and runs outwards and forwards behind the inner nostril (7.7.). Run- ning inwards in front of that passage is the pre-palatine spur (pr.pa.); and from the wide space between the outer and the conjugating bar, in the front part of the long sub- reniform orbital space, a falcate lobe of cartilage grows backwards and inwards, taking ( pt.pa.), whose axis will afterwards be coincident with that of the pterygoid band with which it ” up about one-fifth of the large membranous tract: this is the “ post-palatine also will be confluent. The distal quadrate region (q.) is more than twice as wide and thick as the trabe- cular cornu; it is sinuously flat above and sub-carinate below ; its condyle is saddle- shaped, like that of a human “ humerus,” and looks a little inwards. The free mandible (Plate 22, fig. 4, mk.) is a stout phalangiform cartilage, hooked externally like the “ulna,” whose condyloid notch it imitates ; its distal part is solid, broad, and excavated at the end for the lower labial (/./.). This latter segment now runs across to its fellow with but little downward bend ; the two meet by a straight fibrous joint; each piece is oblique and bulging, where it fits into the end of the mandible. The upper labials (Plate 22, figs. 2, 3, u./.) are nearly divided from behind, forwards, into two flat, oblique plates, the outer of which has a long, blunt, angular process ; here the subdivision of the labial tract, right and left, takes place very early. The hyoid cornu (Plate 22, fig. 5, chy.) is very massive, with a sermi-globular DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 131 condyle and a flabelliform body ; its basal piece (b.dy.) is composed of simple car- tilage. The basi-branchial (b.b7".) is a solid cartilaginous disk, with the rudiment of a second (b.b7.) behind it. The hypo-branchial (A.br.) is a triangular plate, which passes outside the postero- internal angle of the hyoid, and has articulating with it the four normal “ extra- branchials” (ex.br.). They give rise near the lower end to processes that run into the septa on the inner and upper surface of the pouches; these (¢.br.) are (all but the last) small rays, with a pedate free end: they are “ cerato-branchials.” Fourth Family. “ Atyrrp®.” Genus Alytes. This Family, as enlarged by Professor Mivarr (P. Z. 8., 1869, p. 291), suggests this remark, namely, that that assemblage of characters by which the Alytina, Sca- phiopodina, and Uperoliina are characterised, after all only serves to faggot them together; and the possession of neck-glands, dilated sacrum, maxillary teeth, and sharp toes by Alytes in Europe, by Scaphiopus in North America, and by Hyperolius (Uperoleia) in Australia, proves nothing as to the genesis of these three genera. Before all things, in classification, natural geographical grouping has to be looked to, and then afterwards the modification into Families, Genera, and Species. All the Australian Anura I have dissected, as yet, look to me rather like branches of one Australian stock than scattered scions of Families and genera from distant geographical regions; and whilst appraising at their full value every character that can be discovered in any type whatever, I shall keep as far as possible the distribution of the types before my eyes. The Dog-faced Opossum is far nearer akin to the Kangaroo, in spite of his canine features and form, than he is to the true Dog of other regions. 31. Alytes obstetricans.—Adult female; 1 inch 10 lines long. Europe.* The skull of this type is extremely like that of the delicate Australian Tree-frogs, e.g., Hyla Ewingii and H. phyllochroa (see Plate 31); whilst in the next instance, an Australian Frog, the skull might be taken for that of one of the “ Ranidie,” and a not very aberrant form of the European or the Oriental species of Rana. Yet a careful examination of this skull will show that the skull of Alytes has some important points of coincidence with that of Pelodytes and Bombinator; and the Obstetric Frog is more likely to be one branch of the same stock as these other European kinds than a migrated relative of an Australian species, or the Australian species a migrated relative of this. * My specimens of this Frog are the gift of Professor TroescueL, of Bonn, 8 2 132 MR. W. K. PARKER ON THE STRUCTURE AND This skull (Plate 24, figs. 1, 2) is short, flat, and a very delicate semi-osseous structure altogether; the length is to the breadth as 5 to 6, and the quadrate condyles (q.c.) reach as far back as the fore edge of the stapes. The occipital condyles (oc.c.) are large, separated by a very gentle emargination of the basal plate, and directly posterior in aspect. The auditory regions are greatly winged outwards, becoming narrow externally; the common occipito-auditory bones are single on each side, reach as far out as to the wings of the parasphenoid below, and touch the narrow squamosal above; thence the outside of the vestibule (vb.) and the tegmen tympani (t.ty.) are left soft. So also is a rather wide tract above and below at the middle; whilst in front the prootic region of the bone reaches the foramen ovale (fig. 2, V.), which it half encloses. The optic fenestra (IL.) is large, and is well surrounded by cartilage, and this cartilage is continued beyond the middle of the orbital territory ; two-fifths of this territory only is taken up by the girdle-bone (eth.) This bone just reaches the “wings” and the fore end of its own proper territory, and above (fig. 1, eth.) it is imperfect ; the right and left half, which are united, but leave emarginations below (fig. 2), are not fused above: their horns just meet in the fore margin of the anterior fontanelle (/o.). The “tegmen cranii” is very short in the ethmoidal region ; it forms a flat and widish wall- top, runs a band across the roof in the postorbital region, and becomes superoccipital in front of the middle of the auditory regions, right and left. The main space thus spanned is divided into two fontanelles (fo., fo’.); the front space is one-third larger than the hinder, and is emarginate behind; the hinder space is transversely oval. This is a rare modification; behind, there are, as a rule, two secondary spaces, and in several types none. The tegmen grows out into a small superorbital “eave” in front, and then there is a wide stretch of cartilage—ethmoid and ethmo-palatine. Outside the small super- orbital projection of the “wing” there is a very large superorbital cartilage (s.ob’.) ; it covers nearly a third of the orbital space ; is roughly oval, its large end foremost, and it is placed obliquely, with its hollow face downwards. This is also a rare character, it occurs in Phyllomedusa bicolor; but in that type it is much smaller. The common septum of the ethmoidal and nasal regions (fig. 1, s.7.) is large, and is continued in front as a prenasal rostrum (p.n.). As in Bombinator the ale of the nasal roof (fig. 1) are very little developed along the sides of the septum, the cartilaginous pouches being nearly all in front, lying on, and coalesced with, the “subnasal laminze.” These plates (fig. 2, s.n.l.) are normally wide, and have long narrow pro-rhinals (p.7/.). The muzzle being very obtuse, the nostrils are wide apart, they are protected by the usual inner and outer labials (w./'.w./?.), which are large and well developed. The palato-suspensorial arches are wide apart, and the orbital spaces are oval, the pre-palatine spike (pr.pa.) is large, the stem (e.pa.) wide, and the post-palatine and pterygoid tracts are but little affected by the palatine and pterygoid bones (pa., pg.), which are quite typical. This cartilage is seen in both forks—the pedicle (pd.) and DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 133 the quadrate (g.)—the latter is partly ossified near the trochlea or condyle (q.c.) by the quadrato-jugal (¢,j.). The Eustachian openings (ev.) are large, reniform, and transverse ; the annulus («.ty.) is large and its horns are confluent, above. The stapes (fig. 5, st.) is large, truncate in front, and has a large knob for muscular attachment. The inter-stapedial (¢.st.) is well developed as a “sesamoid” cartilage ; the medio- stapedial (m.st.) is roughly pistol-shaped ; the extra-stapedial (e.st.) is a large spatula, giving off a ligulate supra-stapedial (s.st.), which is confluent, above. The mandible (fig. 3) is normal, but the dentary (d.) is much longer than usual, and the articular portion has an endosteal “ articulare ” (a7!.) as in Bombinator. The basi hyo-branchial plate (fig. 4) has the notch very deep; the cerato-hyals (c.hy.) expand to twice their upper size, in the middle, and have a small hypo-hyal lobe ; they are free above. The lateral lobes, fore and hinder, are well developed, but the thyro-hyals (¢.hy.) are near together in front, and are short; they have a pedate unossified free end, and each has, on its outside, a small oblong nucleus of cartilage sticking to the bony shaft.* Here again we see, as in Pelodytes, the \/-shaped perichrondrial bone, with its angle backwards, lying on the under face of the basal plate near the aa it is an attempt to form a basi-branchial bone. The investing bones correspond very closely with those seen in the skulls of Pelodytes and Renbemnn and they are also very similar to those of the more delicate Australian Hyla. The fronto-parietals (fp.) are thin laths of bone, lying on the skull-wall, and curved so as to bind round the two fontanelles and the front of the ear-capsule, just at its inner edge. But they fail to cover the superoccipital region, and only partly hide the girdle-bone. The nasals (n.) have the usual form, and in their fullness of size and shape cover in the imperfect nasal roof. The premaxillaries ( pw.) are extended widely ; the maxillaries, quadrato-jugals, and squamosals (mz., g.j., sq.) are normal but feeble ; the quadrate is partly ossified by the second of these; the third only binds on the edge of the tegmen tympani. The parasphenoid (fig. 2, pa.s.) is large in its fore part, but its wings are narrow externally, and the hind part is but little produced. The vomers (v.) are very large, but in this wide-muzzled Frog they only reach the septum nasi inwardly, and outwardly do not touch the maxillaries. The dentigerous part is oval and transversely placed ; the interspaces between these tracts is almost equal to the tracts themselves. There are no septo-maxillaries. It is impossible to compare this skull with that of Bombinator and Pelodytes without seeing that these three have many things in common, and that in so far as they agree * Mr. Howes has shown me, in his exquisite dissections of the Sturgeon at the South Kensington Museum, that the visceral arches of that fish have many small nuclei of cartilage inside the periosteum, 134 MR. W. K. PARKER ON THE STRUCTURE AND with each other, they differ from the more typical Anura of the European sub-division of the Palearctic region. They differ greatly from the normal Bufo, as well as from the normal Rana, and they are isomorphic, more or less, with types in the “ Notogeea,” to which they can claim only the remotest Batrachian relationship. The main characters by which Alytes differ from the common typical Frog are :— 1. The flattened, very feeble skull, as a whole. 2. The open state of the fontanelles, through the deficient growth of the roof-bones. 3. The fusion of the normal pair of secondary fontanelles into one. 4. The presence of a clear superorbital edge to the skull, and the articulation with it of a large, distinct superorbital cartilage. 5. The nasal roof imperfect, and the “rostrum” well developed. 6. No septo-maxillaries. . Occipito-otic bones continuous. . Quadrate region partly ossified. . Supra-stapedial confluent. 10. An endosteal “ articulare.” 11. Cerato-hyal wide, and basal “ notch ” very large. 12. A \/-shaped ectosteal basi-branchial, and periosteal cartilages on thyro-hyals. NoTilo oFiba. | Fifth Family. Hyperoiip”. Genus Hyperolius. 32. Hyperolius (Uperoleia) marmoratus.—Adult female ; 1$ inch long. Paramatta, Australia. This type represents the “ Uperoliide ” of Ginruer (“ Batr. Sal.,” p. 39), but Pro- fessor Mivarr (P. Z. 8., 1869, p. 291) melts this lesser group mto the “ Alytide ;” the correspondence in certain characters, however, gives them no real title of near relationship. This more extended group of “ Alytidee” must be taken as a convenient temporary zoological bundle ; a very small spark will devour the “tow” that binds these alien types together. This skull (Plate 24, figs. 6, 7) is evenly semi-elliptical in outline ; its breadth is a little greater than its length, and the quadrate condyle (q.c.) is only opposite the foramen ovale (V.). This skull is at once seen to be both generalised and arrested; yet it belongs to a minority among its Australian congeners, in having its fontanelle covered. Altogether it is more like the skull of a young Common, than of an old Obstetric, Frog. It is, indeed, more lke the typical skull than that of the “Cystignathide” of Australia—Lymnodynastes, Camariolius (Plates 18 and 19)—whose skulls approach those of the 7’ree-frogs of the same region. The occipital-condyles are almost hemi- spherical ; they are posterior, and are separated by a straight-edged interspace larger DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 135 than themselves. Below, there is a narrow, above, a wider and widening tract of cartilage, dividing the bony masses of the ear and occiput, which are moderately ex- tended, wholly ossified, and furnished with a large, wide tegmen tympani (¢.ty.), which overhangs the vestibule (fig. 7) to an unusual extent. These right and left bony regions reach to the fore edge of the foramina ovalia, below, which they almost encircle. The skull is wide behind, and narrows evenly, with scarcely any bulging up to the axils of the ethmoidal region. The covered fontanelle (fig. 6) is very large; it reaches to within a short distance of the foramen magnum, behind, and leaves very little tegmen cranii in front. The edges formed by the walls of the skull are straight, and those walls have scarcely any “coping.” The converging and slightly bulging inter- orbital walls are unossified from the foramina ovalia to the ethmoidal axils; in the hinder third of these cartilaginous tracts, the optic fenestra (II.) is seen to be of moderate size. The girdle-bone (eth.) scarcely covers its own territory in front, and behind, only ossifies the axillary or pinched part, reaching some way back below ; above (fig. 6) it is imperfect, its halves scarcely meeting at the mid-line. The nasal region is entirely unossified, there is no prenasal rostrum, and the pro- rhinals (fig. 7, p.rh.) are well developed ; the subnasal lamina (s.7./.) is very broad, and the roof (fig. 6) is of normal width. The external nostrils (e.7.) are only half as wide apart as the internal (7.7.). The palato-suspensorial arch is quite normal, both in its cartilage and in its bones (pa., pg.) ; but both the pedicle (pd.) and the quadrate (q.c.) are short ; the latter is scarcely affected by the quadrato-jugal (q,j.). The Eusta- chian openings (fig. 7) are rather small, and directed obliquely outwards and forwards. The annulus (fig. 8, a.ty.) is a very small crescent, one-third the average diameter. The stapes (fig. 10, st.) is large, ear-shaped, and ossified, except at the edges; it has a strong apophysis. The columella has no inter-stapedial segment, the medio-stapedial (m.st.) is pistol- shaped, the large, proximal pedate end not ossified, and the terminal part is a thick, short spatulate extra-stapedial (e.st.), with no ascending process. The mandible (fig. 8, m.mk., d., av.) and the hyo-branchial apparatus (fig. 9, c.hy., b.h.br., t.hy.) are perfectly normal ; the stylo-hyal end (fig. 7, st.h.) is confluent, above. The labials (w./'.w./°.) and the investing bones are, on the whole, quite normal—that is if compared with those of a half-grown Common Frog. The fronto-parietals (fp.) are very wide behind, deficient inside, in front, and on the left side (reversed in fig. 6) behind, are unusually developed over the arch of the anterior and posterior canals. The parasphenoid (fig. 7, pa.s.) is large and normal, but the vomers (v.) are mere films of bone, bordering the inside of the inner nostrils; they are toothless. There are no septo-maxillaries ; the maxillaries (ma.) send up a distinct process to articulate with the “manubrium” of the nasal (v.), which has the normal conchoidal shape, and comes near its fellow, above. The pre-maxillary ( px.) has united, by its nasal process, with the inner upper labial, so that this cartilage, like the lower labial, is partly ossified. 136 MR. W. K. PARKER ON THE STRUCTURE AND The wide contrast between this skull and that of Alytes is seen at a glance by a comparison of the figures ; it differs from the “ norma” in the following particulars :— 1. The skull, as a whole, corresponds with that of a young specimen of the typical kind. 2. The occipito-auditory region is quite ossified, showing no distinction of front and hinder centres, 3. The skull, proper, is very broad behind, and largely unossified in the interorbital region, the girdle-bone being imperfect above. 4. The annulus and Eustachian tubes are small. . The stapes is very large, projecting, and bony. . The columella is arrested ; there is no inter-stapedial segment or ascending process. . The stylo-hyals are confluent above. . The premaxillaries are grafted on to the 1st upper labials. . The vomers are very small and toothless. Note—That the highly ossified bind part of the skull is in great contrast with the feebly ossified fore part ; the large size of the stalked stapes is in great contrast with the arrested columella and the feeble annulus. oS ON HN Sixth Family. “ BompinaToriD&.” Genus Bombinator. 33. Bombinator igneus.—Adult female ; 13 inch long. Europe. The semi-oval skull is of the average form; the length is as to the breadth as 7 to 8, and the quadrate condyles reach to a point opposite the fore edge of the stapes; in Alytes they were opposite its fore edge, but the breadth of the skull was much greater. This and the next type have many things in common, but their skulls are in strong contrast, on account of the feeble ossification of the one, and the unusual degree of hardness attained by the other; the contrasted figures will show how they differ (Plate 25, figs. 1-4, and 5-11). The main difference in form of outline between the skull of Bombinator and that of the Common Frog, is that in the former the snout is much wider; the interorbital region is much alike in both, it is rather narrow, lessens from before, backwards, and bulges gently in the middle. The auditory region is also wider, and the whole skull more archaic, and also much arrested, The occipital condyles (o0c.c.) are separated by an interval their own breadth, and but little emarginate ; they are swpero-posterior in position as in Pipa, which is a rare modification. The roof over the foramen magnum is also gently emarginate, and both in roof and floor the bones (e.0.) are divided by a moderate wedge of cartilage. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 137 The prootics and ex-occipitals (e.0. to V.) are one undivided mass of bone, right and left, and the tracts are margined by cartilage, outside and in. The outer tract, above, is the widish oblong tegmen tympani (t.ty.), beyond the horizontal canal (h.s.c.); below, the vestibular region (vb.) is largely unossified. In front, the bone reaches to the foramen ovale (V.). Thence the cranial trough 1s unossified for three-fifths of the region of the orbits; in the hinder half of this tract the optic fenestra (0.s., IT.) is very large. The girdle-bone (eth.) does not ossify all the ethmoidal region in front; it reaches further below than above, and there it projects somewhat into the “wings,” and reaches forwards most in the middle. The fore part of the great, single, open fontanelle ( fo.) cuts away a semi-circle from the girdle-bone ; three-fourths of it is uncovered by the arrested roof-bones. The fore part of the ethmoid and all the nasal territory is unossified ; the septum nasi (s.7.) is large and clearly marked, and the subnasal lamine (fig. 2, s.7./.) are very large, even at their narrowest part, mesiad of the internal nostrils (7.7.); their notched outer edges nearly reach these passages. The “intertrabecula” only forms as a knob in the prenasal region; the angles of the subnasal lamin are triangular, aud spread into the maxillaries at their fore end. The secondary cornua, or pro-rhinals (p.7h.), are long, slender, and out-turned. The nasal roof-cartilages form a mere rim to the septum nasi (fig. 1, al.n., s.n.), but they swell out into a curious bag of cartilage in front. This bag lies on the fore part of the subnasal lamina, and is confluent with it; the premaxillary encloses part of this bag, as well as the pro-rhinal band. The thick upper and fore part of the nasal roof stops abruptly, the cartilage dipping down in front ; after forming the bag, it grows up again, behind the large outer nostril (e.n.), which is protected, in front, by the three-toothed outer and upper labial (w./*.) ; this is attached to the inner labial (w./4.), a large oval segment, inside the nasal process of the pre-maxillary. These nasal pouches look like those of Dactylethra, but in that type these bagpipe- shaped pouches are formed out of the large upper labial ; the true nasal roof being a small band over the nasal sac, on each side (see Phil. Trans. 1876, Plate 59, figs. 1, 8, 5,,and 6, w.1.). The ethmo-palatines (e.pa.) are wide, diverging, cultrate bands, emarginate in front, with a wide pre-palatine blade (pr.pa.), and having the post-palatine bar (pt.pa.) continuous with the pterygoid as one arcuate cartilage, but little affected behind by the normal pterygoid bone (pg.). There is no palatine bone, but the pterygoid runs far forwards ; behind, it bifurcates as usual, to invest the large pedate pedicle (pd.), and to clamp the inside of the quadrate stem (fig. 2, sp.) This region is only a little affected by the quadrato-jugal (q,j.) ; its condyle (q.c.) is small, but normal ; it reaches only a little behind the very small Eustachian pouch (ev.). There is neither annulus tympanicus, or columella ; the stapes (figs. 1 and 4, st.) is MDCCCLXXXI, i 138 MR. W. K. PARKER ON THE STRUCTURE AND therefore an uninjured oval, having no bevelling or emargination in front; it forms a fit lid for the vessel of the vestibule (vb.). The stylo-byal end of the hyoid bar (figs. 2 and 4, st.h.) is small, pedate, and articulates loosely with the ear-mass behind the almost closed first cleft. The mandibles (fig. 3) form an elegant arch ; the mento-Meckelian (m.mk.) is not much developed ; the dentary (d.) is small, and the articulare (ar.) is large, has a high coronoid crest, and within its trough behind the articular portion of MECKEL’s cartilage has a large endosteal nucleus (fig. 3, a7’.), as in Chelone viridis. The hyo-branchial structures show the same semi-transformed condition as the other parts of the skull, and some very generalised characters, besides. At first the stylo-ceratohyal band is narrow (fig. 3, c.dy.); it then widens largely, narrows again, has a knee-like dilatation at its fore part, and then goes straight back- wards as a narrow hypo-hyal (h./y.). The notch between the lateral parts of these structures is three times the extent of the solid basal plate (b.h.br.) ; and this basal plate has a bony “basi-branchial” (b.br’.) in its centre. The thyro-hyals (¢.hy.) are normal, but the posterior lateral lobes are twice as large as those hinder forks, and instead of being a slight cartilaginous snag, they are large, and largely ossified. The anterior “lateral lobes” are very large; they, also, have retained much of the larval cartilage, and each flabelliform outgrowth is notched deeply and sinuously in front. The investing bones are very similar to those of newly metamorphosed Frogs and Toads of the more typical sorts. The frontal portion of the two roof-bones (fig. 1, fp.) is merely a narrow lath, lying like a wall-plate on the flat top of the skull-wall ; this ends at the second third of the interorbital region. The parietal portion widens, covers the hinder part of the large single fontanelle, spreads a little over the temporal region, runs sinuously over the hind skull, and ends some distance in front of the foramen magnum. ‘The right bone overlaps the left, and the suture is very irregular. The nasals (n.) are large thin shells of bone, coming near each other by their round backs, and having two shallow emarginations outside. The premaxillaries (px.) are of great transverse extent, but their processes are feeble. There are no septo- maxillaries ; the maxillaries (mwx.) are long thin bones, with a rather high shell-shaped fore end. The quadrato-jugals (q,j.) are but little united to the quadrate; they and the squamosals (sq.) are feeble. The parasphenoid (fig. 2, pa.s.) is a short, wide, thin bone, normal in its processes, but ending far behind the antorbital region in a lathy ragged manner, like that of a young Common Frog. The vomers (fig. 2, v.) are very large, have a massive dentigerous lobe that nearly reaches the septum nasi, are coiled round three-fourths of the inner nostril (¢.7.), and grow out in front to the suture between the premaxillary and maxillary. Here we have a combination of characters that betokens an old type and an arrested DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 139 form ; and these things are not the same, although they often co-exist in the same species. The great difference between the skull of this kind and that of Pelobates partly arises from the Bufonine stoutness of the latter, and the Hyline delicacy of the former ; and this in general form as well as in the internal histological differences. The skull of Pelobates is as much overwrought, in comparison of the “norma,” as that of Bombinator is underwrought ; moreover, the former has a rudimentary columella, whilst Bombinator has none, Pelobates is as remarkable in showing ichthyic bony patches in the cerato-hyals as Bombinator in the basi- and hypo-branchials. Bombinator, in its skull, looks a little towards Dactylethra—that is, in the condition of the nasal capsules ; and it is, for the most part, like a young typical Frog. Its divergence from the last-mentioned species, or the type, is seen in the following particulars :— 1. The fontanelle is wide open, through the arrest of the roof-bones. 2. The ossification of the occipito-auditory regions is continuous, 3. The occipital condyles are supero-posterior, 4. The nasal capsules are almost entirely deficient in their roof-cartilage along the septal region. 5. There are no septo-maxillaries, 6. The investing bones are all feeble, except the vomers. 7. There is no annulus tympanus, 8. The Eustachian passage is a very small diverticulum, 9. There is no columella. 10. There is no palatine bone. 11. The coronoid process of the “articulare” in the mandible is unusually distinct, and there is an “ endosteal” articulare besides the outer plate. 12. The cerato-hyal is greatly dilated distally, and the notch in front of the basal plate is very large. 13. The limited basal plate has a bony basi-branchial centre. 14. The posterior lobes of the basal plate are very large, and ossified as a second pair of hypo-branchials. Seventh Family. “ PrLopatip®.” Genus Pelobates. 34. Pelobates fuscus.—Adult male ; 2 inches 5 lines long. Europe. In these types extremes meet under an apparent external harmony. Peneath similar Taxonomic characters there are hidden the greatest contrasts in these two European representatives of two of the lowest Families of Frogs. T 9 140 MR. W. K. PARKER ON THE STRUCTURE AND In the architecture of their skull they are as nearly much contrasted as any two species that can be found in the whole group of the Anura; we shall see how much of this is due to the feeble condition of one, and to the stout, stony strength of the other. In the skull of Pelobates (Plate 25, figs. 5-11) the length of the skull as compared to its greatest breadth is only as 7 to 84; the occipital condyles project so far beyond those of the quadrate that the median length is one fourteenth greater than the extent of the skull at the sides. Here we miss the excessively wide gape of Xenophrys and the American Bull- frogs ; there, all was for elasticity ; here, all is for compactness. The occipital region is moderately wide; the great width of the skull is mainly due to the suspensorial structures (sq., q.). The endocranium, best seen from below, and the end (figs. 6 and 8) is wide, but not flat as in Xenophrys (Plate 23, figs. 5-7); yet it increases in width from behind, forwards, as in that type; it has also a very broad but short nasal region, which is, as in Xenophrys, a well-fused convex mass, with a thick, short, decurved ‘ prenasal rostrum” overhanging the front; but the premaxillaries under the skull are more normal, Strong as are the outworks (ectocranial elements) of the skull, the equally strong endocranial territories can be made out very clearly, on the whole, without unroofing. The occipital condyles (figs. 6 and 8, oc.c.) are large, sub-reniform, postero-inferior in position, and separated by a space two-thirds their width, which is gently emar- ginate. The whole occipito-auditory region, right and left, is completely ossified up to the foramen ovale (V.); the super- and basioccipital tracts of cartilage are quite obliterated, There is cartilage from the front of the foramen ovale to the front of the optic fenestra (II.), and then the girdle-bone (eth.) occupies all but small supero-lateral tracts in the front ; the sides of the nasal region, only, are unossified. The girdle-bone thus occupies the fore part of the orbito-sphenoidal region, all its own territory, and all the middle part of the nasal. Yet the terminal part was probably formed by a separate centre (or centres), as in Xenophrys, anchylosis taking place afterwards, From whatever aspect the endocranium is examined (figs. 5-8) great strength is seen, the end view (fig. 8) especially shows how strong the occipital arch is, terminated by the large condyles, and flanked by the massive auditory capsules. These show the large arches of the canals (c.s.c., h.s.c., p.s.c.) ; on the outside right and left, the two “teomina ” (fig. 5) are clamped by the large squamosals (sq.) as in a vice. | The outspread girdle-bone (eth.) has shallow “axils,” and sharp angular wings ; these articulate with the palatines (pa.). The ali-nasal cartilages (figs. 5, 7, al.n, n.w.) are crescents of cartilage outside the double bony roof, nasal and prenasal; and the external nostrils (e..) are half embraced by their concave edge; these openings are well finished in front and at the side by the normal inner and outer upper labials (w.0.uP.). Below (fig. 6), we see that the bifureate cornua trabecule—pro-rhinals DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 141 and angles of subnasal lamine (p.7h., s.n.l.)—are not ossified. The pro-rhinals are triangular, pointed and in-turned, as in the genus Bufo; inside them the passages (n.1.) for the terminal branches of the “ orbito-nasal” nerves are large. The thick roof hides the fontanelle, but I suspect it to be single, as in the thin skull of Bombinator (fig. 1). That I should claim for the pterygoid and palatine bones (pg., pa.) the title of endo- skeletal and “ectosteal” will not seem very strange if the condition of these bones in Pelobates is considered. All, or nearly all, the cartilaginous palato-suspensorial arch has been eaten by these strong bones. Here the palatine (fig. 6, pa.) has the form of a hatchet blade, for it has taken up the prenasal spike; it is deeply notched, and half the long-oval, transverse inner nostril (7.n.) is in this notch, and the hole is finished inside by the vomer and wing of the ethmoid (v., eth.). Curving elegantly round in the hollow of the maxillary (mz.) the palatine meets the pterygoid (pg.); seen from below, these two bones appear of equal length: a thing contrary to rule. The pterygoid (fig. 6) has the strength of a “flying buttress,” jammed in between the dentary and palatine lamin of the maxillary (mx.), and then springing obliquely across to clamp the front of the hind skull over the reduced and useless “ pedicle,” it fixes and binds the cheek to the cranium. That inner fork, like a strong foot, stands stoutly on the prootic, and is sutured to the parasphenoid (fig. 6, pr.o., pa.s.). The outer fork passes back outside the minute Eustachian pouch (ew.), and clamps the inside of the quadrate (q.), the distal part of this huge “pier.” By grafting of the quadrato-jugal (q,j.) on the cartilage, this part also is well ossified, all but the large convexo-concaye reniform condyle (¢q.c.). Answering to its pier, the mandibular arch (fig. 10) is unusually stout ; yet it does keep the Meckelian rod—its pith—unossified. The mento-Meckelian (m.mk.) is well developed; the dentary (d.) is continued from it as a strong splint along the front two-thirds of the arch; and the “articulare” (a7.) forms a strong inwardly crested trough for the primary cartilaginous rod. I can find no “annulus tympanicus;” the spiracular cartilage is probably combined with the edge of the “tegmen.” The stapes (figs. 8 and 11, st.) is rather large and has an “umbo ;” its form is three-fourths of a circle, with the hind margin cut away to fit against the sub-convex bony margin of the fenestra ovalis, behind. The columella (co.) is a little phalangiform bone, as long as half across the stapes, thick behind, thin in the middle, and knobbed at its free end; it fits against the inferior margin of the stapes, just below the facial nerve (VII.). The stylo-hyal (s¢./.) has a pedate end which articulates with the auditory mass close behind the minute Eustachian pouch (ew.). The continuation of the hyoid bar gently increases in size until it comes to the hypo- hyal region (fig. 9, c.hy., h.hy.); that part is a large rounded ear of cartilage, which, behind, passes by a broad stem into the basal plate (b.4.b7.). The cerato-hyal has two small osseous centres in it; a rare modification. As in the Hylide, the pre-basal notch is of great depth, and the basal plate of small 142 MR. W. K. PARKER ON THE STRUCTURE AND extent, axially; three-fifths of this is taken up by the huge, very diverging, widely pedate, thyro-hyals (¢.hy.). The posterior lateral lobes are normal ; the anterior are small, irregular lobes. In this skull the roofing and outer walls are in harmony with the inner parts of the building ; the bones are very solid, and their outer surface is almost Ganoid, being richly sculptured with honeycombings and clear granules, like the .surface of a large thick-walled Globigerina bulloides. The roof in this old male is one bone, its parietal region projects as a thick triangular process over the archway of the occiput, and the nasal end (n.) reaches to the verge of the bony prenasal (figs. 5 and 7, n.px., eth.) ; there it ends in two broad, rounded lobes, divided by a sharp notch—all that remains of the nasal suture. This bilobate end of the roof is formed by the nasals, which were very large whilst distinct; they form a right and left wing over and down the antorbital region ; this part is of great breadth, quite unlike the usual spike that articulates with the maxillary. There is a regularly erescentic orbital margin to the frontal region, right and left ; then a short, stout, postorbital process; then a rounded notch in the temporal region (belonging to the parietals); and then an oblique, crenate margin up to the terminal spike. This large, completely anchylosed, convex slab of sculptured bone (fig. 5) overhangs the flat orbital edge of the endocranium, its width, in the orbital region, increasing from before backwards, whilst the inner skull, in that part, increases from behind, forwards (fig. 6). The only part of the slab which articulates with other investing bones is the nasal wing—right and left; much of the extended ear-mass divides the parietal edge from the squamosal (sq.), and the nasal lobes do not reach the nasal processes of the pre- maxillaries (figs. 5 and 7). The parasphenoid (fig. 6, pa.s.) is large and typical; the blade is sub-carinate and trough-like ; the guards, on each side, are also hollow above, and receive the ear-masses into their hollows ; they bend downwards, right and left, in accordance with the dipping of the hind floor of the skull, towards the suspensoria (fig. 8). The handle of the parasphenoid is short and bifid; it does not reach to the edge of the base, behind. The vomers (fig. 6, v.) are curious foot-shaped bones, and are extremely solid, especially the dentigerous “toe”—an oblique, oval mass, looking straight towards its fellow, from which it is separated by a space half the width of the lobe. The “ heel” of this bone lies over the internal nostril (i.n.), from thence it becomes a thick wedge reaching to the junction of the premaxillary with the maxillary. The solid dentigerous lobe of each vomer, and the internal nostrils, form together a transverse series of long ovals, broken by the exposed base of the girdle-bone, and bounded by the bony palatines. The premaxillaries ( px.) are strong, wide, have a small flat palatine, and a large bowed sub-vertical nasal process (fig. 7, 7.pax.), which overlaps the inner and upper labial 2 DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 145 = (u.l.1) hardly as large as usual. Attached to this cartilage is the second or outer labial (u.l.2); it, also, is crescentic and solid; its concave face lies towards the nostril. Below this, and wedged in between the premaxillary and maxillary, there is a granular septo-maxillary (fig. 7, s.mz.). The fore margin of the high, long, and thick maxillary (mz.) is notched where it receives the small bone; its greatest height is where the nasal wing rests obliquely upon, and is sutured to, it. The upper surface is cut away in crescentic manner, and this margin and the orbital edge of the frontal region, together, form most of the oval orbital space ; it is only deficient where the post- orbital process fails to meet the antero-superior angle of the squamosal (sqy.) in the temporal region ; the temporal muscle is thus uncovered at its “ origin.” Behind the suborbital margin the maxillary rises to form a suture with the post- orbital process of the squamosal, and then lessens to its end, which is a blunt point overlapping the quadrato-jugal (¢j.) This latter bone is high, thick, and quite one with the ossified quadrate region (q.). The squamosal (figs. 5, 7, 8, sg.) answers well to the rest of the outer garniture of this strongly-built skull; its facial granular plate is large, oblique, has two large rounded emarginations above, and a lesser round notch below. Its stem (tig. 7) is strong, sinuous, and notched below; its supra-temporal tract runs over the tegmen tympani, binding hard upon the prootic, up to the horizontal canal (h.s.c.), to the form of which it is adapted. The end view (fig. 8, sy.) shows how solid and convex this bone is, and what a stout knee its “ pre-opercular,” or descending process makes, where it binds upon and clamps the suspensorium, and articulates with the pterygoid by its hinder edge. I shall compare this skull now with that of the type, as I have done with that of Bombinator which has just been described. The most striking modifications here seen are as follows :— 1. The form of the skull, which is very short, high, convex, and of a regularly semi- oval contour. 2. The intense ossification of all the parts, outer and inner, and the sub-ganoid condition of the exposed parts of the investing bones. 3. The anchylosis of all the sx roof-bones. 4, The complete union of all the four occipito-auditory centres. 5. The great extent of the girdle-bone, from the optic fenestra to the fore edge of the skull. 6. The inferior position of the premaxillaries. 7. The obliteration of the joint of the pedicle. 8. The lateral junction of the squamosal and maxillary. 9. The complete ossification by the quadrato-jugal of the quadrate. 10. The minute size and imperforate condition outwards of the Eustachian passage as the tympanic cavity. 11. No annulus tympanicus, 144 MR. W. K. PARKER ON THE STRUCTURE AND 12. Columella extremely minute, generalised, and completely ossified. 13. The two cerato-hyal centres, the short basal plate, and the minute size of the antero-lateral lobes. I might, perhaps, add 14, A single fontanelle. But this is doubtful; there is only one in Bombinator. I. A. b—PLATYDACTYLA. First sub-division.—Treefrogs with narrow sacral apophyses and no neck-glands. First Family. Potypepatip&. First genus. Polypedates. 35. Polypedates chloronotus.—Adult male; 2 inches long. India. The difference between this skull and the next (Plate 26, figs. 1, 2; and Plate 27, figs. 1, 2) lies mainly in its general form; its shape is half a long ellipse, and the length is to the breadth as 84 to 73—an almost exact reversal of the proportions seen in the other. The mid and hind skull are greatly affected in their shape by this, the parotic pro- jections being much less and the orbital region much longer ; it is also much narrower. The nasal region has lost little of its breadth, but the ethmo-palatines are shorter and lie within less divergent maxillaries; its three fontanelles are normal. The occipital condyles (Plate 26, figs. 1, 2, oc.c.) are very similar in both; they are oval, posterior, and separated by a concave emargination, which is narrower in this species than in the other. Here the ossification is more intense in the hind skull; the bony masses of the two sides ( p7.o., e.0.) are confluent both above and below, and there is searcely any cartilage left—it is only seen at the outermost part where the pedicles articulate—and the bone reaches half-way between the foramen ovale and the foramen opticum (V., IL.). But the girdle-bone (eth.) is somewhat more restricted, and only reaches to the middle of the orbits; likewise in front it does not run along the septum nasi, but the whole bone stops short a little behind the margin of the proper ethmoidal region. The sinuosity of the bounding lines of the mid-skull is similar in both, but the curves are gentler in this narrower skull. The actual breadth of the nasal roof and floor is but little less in this kind, and the snout is more directly transverse as in Rhacophorus maximus (Plate 26, figs. 5, 6). The pro-rhinals (fig. 2, p.rh.) are large, and the septum nasi (s.7.) is very thick and solid, as is seen both from the upper and lower surfaces. The double palato-suspensorial bow (with its outworks) is rather tightly drawn ; and the quadrate hinges (q.c.) retreat a little less than in the next kind. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 145 Here the whole structures is very Ranine, but unusually narrow from side to side, yet the form of the palatines and pterygoids (pa., pg.) and the amount of unossified cartilage is quite normal; as also is the pedicle (pd.), but the quadrate (q.) is partly ossified by the quadrato-jugal (¢,/.). The Eustachian openings (ew.) are quite as large as the inner nostrils ; they are oval, and are turned outwards and forwards. The annulus (a.ty) is large and complete, and the stapes and columella (figs. 1 and 2, st., est.) are like those of P. maculatus; the stylo-hyal (st.h.) is confluent above. The mandible and hyo-branchial apparatus (figs. 8, 4) are normal, but the articulare (a7.) is fastening upon the ossifying cartilage to make a more solid lower jaw than that of most species; also the hypo-hyal (h.hy.) has a free lobe in front, as in the “norma,” which is not present in P. maculatus. That species approaches Rhacophorus maximus; this comes very near Hylarana temporalis (Plate 29, figs. 1-5). The investing bones only differ from the type in being fitted to a skull with a wider cavity and narrower cheeks. As regards conformity to the Batrachian type, this kind is one of the highest :— The whole cranio-facial structure is longer, and with a broader skull-cavity. The occipito-auditory bones are all confluent, right and left. The columella has an imperfect supra-stapedial. The cartilage both above and below the mandibular joint is partly ossified. . There are no septo-maxillaries. On the whole, this is perhaps the least modified from the typical Ranine skull of any of the “ Platydactyla.” Ss oe eh 36. Polypedates maculatus.—Adult male ; 2 inches 1 line long. India. There are two noticeable varieties of the Oriental Polypedatide, namely, those with short and flat, and those with long and narrow skulls; this (Plate 27, figs. 1-4) belongs to the first variety, which culminates in Rhacophorus maximus (Plate 26, figs. 5-9). The outline is evenly semi-elliptical; the length is to the breadth as 8 to 9, and the condyles of the quadrate come short of those of the occiput by one-thirtieth of the whole length of the skull. The term Hylarana given to one of the genera in this Family perfectly expresses the relationship of this group, whose members are almost exactly intermediate between the species of Rana and the species of Hyla. In this skull the antero-posterior extent of the nasal and orbital regions is equal, but that of auditory, only one-half; this latter region, however, is of great transverse extent. The chondrocranium, as a whole, has had a fair half of its bulk converted into bone, and the skull in all respects has a medium strength; it is much depressed, and the orbital region is at least one-third wider than in most of the species of Jana. MDCCCLX XXI. U 146 MR. W. K. PARKER ON THE STRUCTURE AND The small, oval, directly posterior occipital condyles are separated by a concave space more than the width of both together, and more than half of this basal inter- condylar tract is cartilaginous. Also a wide “cross” of cartilage intervenes between the four main hind-skull ossifica- tions, below ; above, the superoccipital synchondrosis is narrow, and the prootic and ex-occipital (fig. 1, pr.o., e.0.) of the same side are entirely confluent, the bony substance reaches to the squamosal (sq.) under which the “ tegminal edge ” is soft. Even below, the trigeminal nerve (fig. 2, V.) passes through the bone, which is not far from the girdle- bone, over the optic nerve (II.). Some cartilage and three small membranous fontanelles lie across the hind skull, behind the large heart-shaped principal space. The breadth of the skull at the top lessens along the girdle-bone, in which, however, it spreads out again gently in front. This bone (eth.) takes up three-fifths of the orbital region, it passes, in front, nearly to the boundary of its own (ethmoidal) region, and above, runs a little along the septum nasi (s.7.). Over its wings this common ethmoid has a small superorbital “ eave” of cartilage. The nasal region is very large both ways ; it ends abruptly in front, has large cres- centic roofs (fig. 1), a long and moderately deep septum (s.n.), and a wide floor (fig. 2), The pro-rhinals (fig. 2, p.rh.) are slender, but well formed and uncinate. The super- added cartilaginous valves of the external nostrils (w./.u.P’., e.n.) are large, and those passages are wide apart. Very wide apart indeed are the internal nostrils (7.n.) ; they are large oval holes con- verging forwards. The palato-suspensorial arch is large and well developed ; all its parts are wide, for width characterises the whole skull and all its parts. Yet, in essentials, both the cartilage and its grafted bones (pa., pg.), with the condyles of the quadrate and the pedicle (q.c., pd.), are quite normal. The Eustachian openings (eu.) are less than the inner nostrils, but are large; they are oval, with the long axis directed outwards and a little forwards. The annulus (a.ty.) is Jarge, broad, and perfect; the stapes (fig. 4, sé.) is large, oval, and flattish: it is broader in front than behind. The inter-stapedial (¢.st.) is a thickish subquadrate seoment, nearly half the size of the stapes; the medio-stapedial is long, straight, and its imperfectly ossified, wide, proximal end is scooped out ; the extra-stapedial (e.st.) is a reversed spatula supra-stapedial band. The narrow stylo-hyal (fig. 2, st.h.) is confluent above; it increases only gently in size downwards (fig. 8, e.hy.), has no hypo-hyal lobe (h.hy.), and soon reaches the basal plate it is narrower distally than proximally, and has only a fibrous as it turns backwards. That plate (b.4.b7.) is short, wide in front, narrow behind, has small fore-turned anterior, and long, slender, posterior lobes ; the thyro-hyals (t.hy.) are long and slender; they diverge so as to be at a right angle with each other. The mandibles (fig. 3) are quite normal, but are long and slender; the dentary (d.) is half the length of the ramus. The investing bones are normal but partake of the general flatness and breadth of en a DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 147 all things else in this skull. The roof-bones ( fp.) run short in front ; the parasphenoid (pa.s.) has poimted arms under the ear-capsules; the vomers (v.) are wide apart and have only a smallish crop of teeth; there are no septo-maxillaries ; the premaxillaries ( px.) are wide but feeble, so also the maxillaries (mw.) have a wide facial plate, but it is thin; the quadrato-jugals (qj.) are not grafted on the quadrate; the squamosals are but little developed over the tegmen (fig. 1, sq.). Here, in a few things, the ‘‘norma” is not reached ; besides the general breadth of the parts— 1. The auditory occipital masses of bone are confluent above. . The lesser fontanelles are very small and have a still smaller space between them. . There are small superorbital projections. . There are no septo-maxillaries. . The supra-stapedial is membranous. . The hyo-branchials are feeble. Dore WS LD This is a poor list of indictments against this species as coming short of the norma ; the flatness of the general shape, and of the individual parts, are also no great modi- fications in the morphology of this very Ranine skull. Second genus. hacophorus. 37. Rhacophorus maximus.—Adult male; 3§ inches long. North India. This large species bears the same relation to the Tree-frogs of India that Pelodryas does to those of Australia, and Phyllomedusa to the Neotropical kinds. That which characterises the Oriental Polypedatide, generally, is here carried to excess, namely, the great dilatation of the endocranium, especially in the nasal region. The length is seven-eighths as great as the breadth ; the quadrate condyles (Plate 26, fig. 5, 6, g.c.) are opposite the passage for the vagus nerve (fig. 6, X.). The cranial “barge,” from the foramen magnum to the great transvere snout, is twice the breadth of an average skull of a species of Rana of the same size as this kind. The fore skull, leaving out the massive premaxillaries, is as long as the mid skull, and is wider: it is extremely wide in the ethmoidal region. The mid skull lessens gently to the temporal region, where it is pinched in ; in front it is three times as wide as it is deep (figs. 5, 6, 7). The antero-posterior extent of the hind skull is half that of either of the other two regions, yet the breadth of the cranial cavity is not lessened there, but the parotic wings extend far out, as the great breadth would indicate, beyond the horizontal canals (fig. 5, h.s.c.); these wings are rather narrow. The occipital condyles (0c.c.) are rather small for so strong a skull, they are posterior, and are separated by an arched (emar- ginate) line of their own breadth. The occipito-auditory ossifications are all continuous, as in Polypedates chloronotus, U 2 148 MR. W. K. PARKER ON THE STRUCTURE AND and just a little cartilage is left on the outside of the wings; the bony matter reaches half way between the foramen opticum and the foramen ovale (II., V.). The tract of unossified skull-wall round the optic passage is only one-fourth the extent of the mid skull, for the girdle-bone (eth.) is very large; in front, above, it occupies half the nasal roof (fig. 5, al.n.), and below, it runs up to the dentigerous lobes of the vomers (v.). Under the roof-bones (fig. 5, fp.) the three normal fontanelles can be seen, and as those bones only reach, in front, halfway from the foramen magnum to the transverse end of the snout, the girdle-bone is but little roofed in above. Below, the bone passes into the ethmoidal wings, but above, there is a considerable selvedge unossified ; this is the apiculate superorbital (s.ob.). The pre-cranial mass, which is the fore half of the ethmoidal, and all the nasal region, is square, and its breadth is equal to the widest part of the roof in the orbital region. Its fore edge is almost directly transverse, being only gently rounded. The depth of the nasal region is only moderate (fig. 7); the septum (fig. 6, s.n.) is very thick ; the floor and roof (figs. 5 and 6) are equal in width; the pro-rhinals (p.rh.) are small and wide apart outside the nerve outlets, which also are wide apart. The external nostrils (figs. 5 and 7, en.) are very wide apart, and the nasal roof (al.n.) is very thick over them; so, also, the second upper labial (w./?.) is a thick cres- centic valve. The 1st upper labial (w./!.) is rather large, and the nasal process of the premaxillary (px.) is grafted upon it, ossifying much of its substance. The internal nostrils (i.n.) are very wide apart. The palato-suspensorial arches are not large in proportion to the skull proper; and they have only the average ossification, the two bones (pa., pg.) leaving a considerable tract unaffected. The palatines (fig. 6, pa.) are thin, straight, splint-like bones becoming twice as wide at their outer end as at their inner. The curved and bifurcated pterygoids (pg.) are moderately strong; the inner fork on the pedicle is half as long as the outer, which runs inside the quadrate (q.c.). Both the pedicle and the quadrate are formed of a solid mass of cartilage, especially the former (pd.), which forms, here, the most massive condyle I am acquainted with, at this part. This sinuous facet glides on a definite tract of basal cartilage, which is confluent, externally, with the stylo-hyal (st.h.). The quadrate, above the reniform condyle (q.c.), is not ossified to any noticeable extent. The Eustachian openings (ew.) are large, oval, and have their long axis directed forwards. The annulus (fig. 7, a.ty.) is large, broad, and complete; its edge has a strong rim, The mandible (fig. 7) is normal ; it is rather slender, and has a high coronoid process to the articulare (a7.). The stapes (fig. 9, st.) is oval, and not raised externally to any appreciable extent. The inter-stapedial (z.st.) is massive and semi-ossified ; it runs some distance on the shaft. The medio-stapedial (m.st.) is a rather slender, terete rod ; the extra-stapedial (e.st.) is large and spatulate ; the (s.st.) supra-stapedial is cartilaginous and confluent, above, DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 149 The hyo-branchial plate (fig. 8, b.h.b7.), with its outgrowths, falls short of the type in having no hypo-hyal lobes, and in the small size of the front lateral lobes; the thyro-hyals (¢.iy.) are rather slender, and only diverge gently. The investing bones are very similar to those of the type, but modified so as to fit to a more outspread cranium. The roof-bones (fp.) are strong, very wide, and as in most of this Family are very scant in front, especially near the mid-line. The nasals (n.) make no pretence of covering the skull in their region, for they are very narrow crescentic shells of bone. The premaxillaries (px.) are, perhaps, the largest in the Order, bounding, as they do, so wide a nasal fore wall. There are no septo-maxillaries ; the maxillaries (mz.), the quadrato-jugals (q.j.), and the squamosals (sq.) are all strong, well developed, and normal. So also is the para- sphenoid (fig. 6, pa.s.) ; itis very sharp, both behind and before. The vomers (v.) stretch far across the nasal floor, but fail to meet; they have the usual four-snagged form ; the large thick tooth-bearing lobes are arcuate, and their concave margin is behind. The relation of this type of skull to the others in this Family is evident; this is the fullest expression of the Polypedatine type, and thus the greatest divergence from the form of the Ranine skull is obtained. We have to remark upon— 1. The short auditory and occipital region, the wide orbital, and the wide and long nasal region. 2, The fusion of the bones of the hind skull. 3. Unusual solidity and strength of the pedicle. 4, The inter-stapedial ossifying as part of the shaft, apparently without absolute segmentation of the cartilage. 5. The confluence above of the supra-stapedial and stylo-hyal bands. . The feebly expressed cartilages of the hypo-branchial plate. . No septo-maxillaries. . A distinct superorbital eave. . The grafting of the premaxillary on the “ 1st upper labial.” All these modifications are gentle and non-essential. OS ON SD Third genus. Lvalus. 38. Ixalus variabilis.—Adult female ; 1 inch 1 line long. Ceylon. This skull differs very little indeed from that of Polypedates maculatus; it would differ still less from that of a young individual of that species of its own size.* It is a short skull, and the length bears to the breadth the proportion of 7 to 8 ; it is therefore a trifle narrower than that of Polypedates maculatus. The outline is extremely regular, and is half an ellipse, whose long diameter may be put as 13 and * Dr. Giinruer is now inclined to consider this species as the young of Polypedates maculatus ; but the well developed supra-stapedial band, and the absence of an inter-stapedial and of secondary fontanelles, makes this doubtful, 150 MR. W. K. PARKER ON THE STRUCTURE AND its short as 8, for the outline of the face stops a little short of the axial length; the mandibular hinges are slightly less retreated than in the large kind. The occipital condyles (Plate 27, figs. 5, 6, oc.c.) are similar; the cartilaginous tracts at the mid- line are larger, and the projection of the ear-mass less, and less ossified ; the “ tegmen” (fig. 5, t.ty.) is soft. The hind part of the orbital region is not so wide ; the girdle- bone is of less extent, and the superorbital tract of cartilage is not present. The nasal region is similarly broad, on the whole, but the outer nostrils (e.7.) are nearer together, and the inner nostrils (7.7.) are scarcely oval. What difference of length over breadth there is, is shown differently—the holes are transverse. There are no secondary fontanelles (fig. 5); the “annulus” (a.ty.) is smaller, but complete ; the Eustachian openings (ew.) are similar in both kinds. The columella has no separate inter-stapedial (Plate 29, fig. 6, m.st.); and the supra-stapedial (s.st.) is confluent above ; the stapes (st.) has a boss. The investing bones differ in these two kinds, as those of the old and young of the same species differ ; there are no vomerine teeth. The mandibles and hyo-branchial apparatus (fig. 7) are very similar, but the former are more evenly arched, and the latter has a young appearance, for the front lateral lobes are not such elegant “ears,” and they are not definitely marked off from the hinder projection (b.h.br.). The basal plate is longer, and the thyro-hyals (¢.hy.) are slenderer, and are inbent. This small Ranine Tree-frog differs from the type in the following particulars :— “1. In the general broad, short, outspread condition of the skull. There are no secondary fontanelles. There are no septo-maxillaries. . There are no vomerine teeth. The tympanic annulus is less. There is no inter-stapedial segment, and the supra-stapedial is confluent above. a Oa eae There is no lobe to the hypo-hyal. Fourth genus. Hylarana. 39. Hylarana malabarica.—Young ; 3 inch long. India. The specimens dissected by me were only one-fourth the length of the full-grown adult, which attains to nearly 3 inches. Nevertheless I found the skull in these young individuals (Plate 28, figs. 1, 2) to be much more perfectly ossified than in the adult of H. temporalis (Plate 29, figs. 1, 2), next to be described. This is one of the longest skulls in the “ Order;” its length is to the breadth as 74 to 6L—a Frog’s skull which is as long as it is broad, is long as compared to most « Anurous” skulls, The proportionate length of the three regions of the skull is very different from DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 151 what is seen in [. temporalis, where the muzzle is very broad : here the fore and hind regions are equal, and each is three-fourths the length of the mid skull, which is both long and broad. The occipital condyles (Plate 28, figs. 1 and 2, oc.c.) are postero-inferior, and are moderately large. A considerable tract of cartilage remains in both the basi- and superoccipital regions ; there is a triangular endosteal superoccipital bone (figs. 1 and 5, s.o.) ; and the bony tracts (pr-.o., e.o.) of the same side are quite confluent. Below, the bone leaves an even crescent of cartilage at the margin of the floor of the vestibule (au.) ; above, the bone is snagged, and leaves an irregular cartilaginous tegmen (fig. 1). I could find no secondary fontanelles, but a very large long main space with no lateral ingrowth for a considerable distance along the sides, quite unlike the next. Moreover, this deficiency of the ‘ tegmen cranii” shows itself both behind and before : behind, the roof is devoid of cartilage up to the ear-capsules; and in front the roof extends a very short distance over the hemispheres. As this is a young specimen, the ear-masses and the nasal roofs show much of their primary form; they are not drawn into new shapes by the overgrowth of the peripheral bones and arches. Hence comes the narrowness of the skull in the auditory region, for the tegmen tympani is merely a narrow selvedge to the ovoidal ear-sacs; this would be at least somewhat modified in an old mdividual. On the whole, the orbital region, or mid- skull, is very Polypedatine, but is broadest behind, narrows gently, and widens out again towards the ethmoidal region. What is very remarkable in so young a skull is the trespassing of the so-called “prootic” round the front of the foramen ovale (V.), even so as to include the lesser cranial nerves between it and the foramen opticum (II.). Therefore, so early, the orbital region is half of it bony, the girdle-bone (eth.) being as long as the cartilage behind it. It runs into the septum nasi (s.n.) both above and below ; then it takes in all its own cartilage, athwart the ethmoidal “ wings ;” above, it scants a little, and leaves the edge of the ethmoidal roof, at its junction with the nasal roofs (fig. 1), naked. These roofs are very large and broad (fig. 1), larger than the floor (fig. 2) ; the inter- trabecular wedge (or wall) between both grows into a distant “prenasal rostrum” (p.n.) in front, which is a new thing in an Oriental member of this Family. Never- theless, the snout is very broad, and the rostrum, which is drawn as dissected out in the figures, in reality turns downward, as in embryos of all kinds that possess this projecting cartilage. In this kind, as in its congeners, the nostrils (¢.7.) are very wide apart, nearly as wide as the “choanze” or inner openings (fig. 2, 7.n.). The pro- rhinals (p.rh.) and the labials (wv./'.w.7?.) are well developed. The arrest of the “ parotic wings,” in a skull with such long regions, causes the facial “bow” to be strongly bent; otherwise all is normal; the bones (pa., pg.) are already of the normal size and strength, and leave a cartilaginous palato-suspensorial of the average size. The forks of this arch differ greatly, for the pedicle (pd.) is very short, and the Laz MR. W. K. PARKER ON THE STRUCTURE AND quadrate, with its condyle (q.c.) rans downwards and outwards to a considerable distance ; backwards, it only reaches to the fore end of the stapes (st.); above the condyle the quadrato-jugal (¢.j.) has spread into the cartilage, ossifying it. The Eustachian openings (ew.) are large and oval; they are turned forwards; the annulus (a.ty-) is large and perfect; the stylo-hyal (st.h.) is confluent above. The stapes (fig. 4, st.) is large, convexo-concave, oval, and emarginate in front; the inter- stapedial (¢.st.) is a thick mass, ossified largely, and joined by a suture to the medio- stapedial (m.st.), which is a straight rod. The extra-stapedial (e.st.) is a wide spatula, and its ascending fork (s.st.) is cartilaginous and confluent above. The mandible (fig. 3) is perfectly normal, and the mento-Meckelian (im.mk.) is well ossified. The coronoid crest of the articulare (ar.) is moderate, and the condyle is sub-reniform. The hyo-branchial apparatus is normal, but its actual shape is much like that of the next (Rappia, fig. 8); the hypo-hyal lobe is small, and the cerato- hyal broad (h.hy., c.hy.); the semi-oval notch in front is short, the front lateral lobe very large and separated by a small space from the hind lobe ; the basal plate (b./.br.) is both long and broad, as compared with that of other kinds, and the thyro-hyals (t.hy.) are long and slender. Altogether this part is exactly intermediate between what we find in the ordinary Oriental Polypedatidze, and in the next but one, a small Ethiopian type. The investing bones are such as would be found in a Common Frog’s skull of the same size and age. The fronto-parietals (f:p.) are more Ranime then in the last, being pointed above ; they would be more truncated in an old specimen. The nasals (7.) are better developed, and cover more of the nasal roof than in its congeners. The para- sphenoid (pa.s.) is very large, both in the median and transverse parts; and the vomers (v.) are well developed and dentigerous. J find no septo-maxillaries; the marginal splints (pw., mx., q.j., sq-) are feeble but normal. An intermediate place between Hylarana temporalis and the West African Rappia is asked for this species : we shall see this in the two next instances. Compared with the type, we have some curious discrepancies :— . The great size, both length and breadth, of the orbital region. . The intense ossification of the occipito-auditory region. A well-marked endosteal swpraoccipital bone. No septo-maxillaries. . The confluent condition of the apices of the stylo-hyal and supra-stapedial. . The intense ossification of the inter-stapedial. 7. The large size and close contiguity of the lateral lobes of the basi-hyobranchial plate ; none of these are deep-seated differences. anf wond 40. Hylarana temporalis—Adult male ; 24 inches long. Ceylon. Dr. Giwruer (“ Rept. of Brit. Ind.,” p. 427) gives his reasons for believing this insular species, which at that date (1864) he had received from Ceylon only, to be DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 153 distinct from the continental form—H. malabarica. I have just described that species, and when this has been compared with it then all doubt will vanish as to their distinctness ; both kinds attain the length of three inches, or thereabouts ; my speci- mens of H. malabarica are all young, and only a quarter the length of the adult ; yet the ossification of their skull is much more perfect than in the adult of the Ceylonese kind ; there are also other differences, as we shall see. This is a long skull (Plate 29, figs. 1, 2), for the length is equal to the breadth ; it is extremely like that of Polypedates chloronotus (Plate 26, figs. 1, 2), but is not so long ; in that species the length is about a twelfth greater than the breadth; and that kind has its ossifications intenser, and has no evident superorbital “eave.” Although belonging to the narrow-skulled group of this Family, the cranium proper is in reality very broad, half as broad again as in an average species of Rana, and therefore settling this as a character of the Oriental flat-toed Frogs. They are, indeed, a broad-muzzled, flat-skulled race, but when flattened most their skull is very Ranine as compared with that of a true Hyla. The occipital condyles (Plate 29, figs. 1, 2, oc.c.) are large, posterior, and separated by a crescentic emargination half their breadth. Measured along the axis, the occipital region is half, and the nasal two-thirds, the extent of the orbital, which is very large. The “tegmen cranii” covers two-thirds of the skull, and grows in well from the sides, so that the larger cordiform fontanelle is small, and the secondary spaces very minute. The parotic wings stand out well, and the canals (q.s.c., h.s.c., p.s.c.) are prominent. The ossification, right and left, is only equal that of the type, and the four normal centres (pr.0., ¢.0.) are separated by extensive tracts of cartilage. Above, the prootic stretches out in front so as to underlie the squamosal a little. Below, the floor of the vestibule (aw.) is unossified, for the ex-occipitals, which flank the posterior canal above, do not reach to the fenestra ovalis below. In front, the prootics just reach the optic fenestree (II.). Of the remainder of the long orbital region the hinder half is unos- sified ; the rest, or the region of the “girdle,” is partly ossified, and partly calcified, the latter tract marking out the usual extent of the bone, which however is only fully developed in front into an irregular U-shaped bone, thick and rugged, but whose halves do not meet above (fig. 1, eth.). In front, this imperfect girdle scarcely covers the proper ethmoidal territory, and the whole nasal region, and likewise the extensive superorbital crescents (s.vb.), are unossified. Leaving these out, the form of the orbital region of the skull is oblong, modified by a slight bulging, and a gentle increase of breadth, forwards. The nasal territory (al.n.) is large and four-square, with the roof spreading wider than the floor ; the septum (s.n.) is thick, and both above and below we can see clearly how much has been added to the original “trabecular horns” by the solid inter- trabecular wedge (or wall), and how much by the large shells of cartilage that form the roof. Their division also, into the floor (with its angles that bury themselves MDCCCLXX XI. x 154 MR. W. K. PARKER ON THE STRUCTURE AND in the maxillaries), and the new “horns” or pro-rhinals (p.rh.) that have budded from their infero-medial face, is clearly seen. The broad snout is gently emarginate in front, and the nostrils (e.n.) look outwards and forwards : they are wide apart, and are protected by the normal valves (w.l.wl.). The large circular inner nostrils (¢.n.) are still wide apart, not much nearer together than in the more flat-headed “ Polypedatidee.” The palato-suspensorial arches are perfectly normal, but slender ; their bones (pa., pg.) partake of a like slenderness, and the well-retreated condyle of the quadrate (y.c.) is large. So also is the condyle of the pedicle (pd.) ; it is sub-convex; the other is deeply bilobate : above this hinge the suspensorium is ossified to some extent by the quadrato-jugal (q., q,j.). The Eustachian openings (ev.) are large, oval, and look outwards and forwards. The annulus (a.ty.) is very large, broad, and perfect. The mandible (fig. 3) is long, slender, and normal; it has a rather high coronoid process to the articulare (a7.) ; the dentary (d.) is half the length of the ramus, and rises into a crest behind the mento-Meckelian (mk.), as in the Oriental Rane (Plates 6 and 7). The stapes (fig. 5, st.) is large, sub-oval, and has a boss for muscular attachment ; the inter-stapedial (7.st.) is a large, thick, sub-oval cartilage ; the medio-stapedial (m.st.) is a short pistol-shaped bone, very large and expanded above ; the extra-stapedial (e.st.) is large and spatulate ; and the supra-stapediai (s.st.) is almost entirely fibrous. The hyo-branchial apparatus (fig. 4) is normal, but all the “lobes” narrow. are small and The investing bones are intermediate between those of a middle-sized Rana and a broad-headed member of its own Family; both the roof-bones (fip.) and the para- sphenoid (pas.) reach further forwards than in its congeners. The nasals (7.) are angular, jagged shells, and leave a large cruciform tract of cartilage uncovered. The outer bones of the face are all slender ; there are no septo-maxillaries ; and the quadrato-jugals (q,j.) are grafted upon the quadrate cartilage. The vomers (v.) are large, have a considerable dentigerous lobe, and the three other spikes are well developed. Compared with the “norma,” this skull is— 1. Remarkable for the breadth of the cranium proper; for the great length as well as breadth of the mid skull; and for the width of the fore skull, or nasal region. 2. There are no septo-maxillaries. 3. The superorbital “eaves” are large. 4. The girdle-bone is imperfect. 5. The stapes has a boss, and the stylo-hyal is confluent. 6. The supra-stapedial is imperfect. These are all slight modifications, and on the whole this skull is extremely Ranine, und agrees especially with the Oriental Rane in having a crest on the fore end of the dentary, and a large perfect annulus. or on DEVELOPMENT OF THE SKULL IN THE BATRACHIA. ] Fifth genus. Rappia. 41, Rappia ——? sp.—Adult female ; % inch long. Lagos, This specimen was collected, with others, several years ago at Porto Novo, near Lagos, by R. B, WaLkeEr, Esq., and was sent to me by Mr. T. J. Moore, of Liverpool, on March 24, 1874. Dr. GUyruEeR was doubtful of the species, although there is a Rappia lagocnsis (see GintueEr, P. Z.S., June 25, 1868, pp. 478-490; a figure of that species is given in plate 40, fig. 2, and a description at p. 487). Dr. GUNTHER gives the length of P. lagoensis as 28 millims. ; my specimen measured only 21 millims., although a female. In the paper just referred to, 15 species of this genus are given—all African. In the “ Batrachia Salientia,” p. 85, these are described under the generic term Hyperolius, a name now used instead of Uperoleia (ibid., p. 39) for a genus related to Alytes. The only Australian kind described in that work is H. bicolor (p. 89) from Port Essington. The author remarks of this, whose skull I shall describe next, that—“ This species is very probably the type of a separate genus.” This suggestion is a true one; the Australian type has an exceedingly different skull from that which I shall now describe from West Africa. This skull (Plate 28, figs. 6, 7) greatly resembles the last but one (figs. 1, 2) ; and the thought that occurs is this—Are the African Rappie dwarfed “races” of various species of the “ Polypedatidz ;” types which attain so large a size in the Oriental region ? This is a rather long skull ; the length and width are equal, and it tends towards the triangular form. The condyles of the quadrate (q.c.) do not reach so far back as in the last but one, they end in front of the fenestra ovales. The occipital condyles (0c.c.) are similar—they are large and posterior ; but the basal emargination is less in this species. In conformity with the greater breadth across the ears, the simply oval form of the capsules is lost, and there is a considerable tegmen tympani (fig. 6). The general form of the mid and fore skull is very much alike in both kinds, but they are both a little broader (relatively as well as really) in this small adult; and the nasal roofs are like those of a metamorphosing Rana temporaria (see Phil Trans., 1871, Plate 8, fig. 1). The four bony tracts of the hind skull (pr.o., e.o.) are like those of the same of Young Frogs; whilst the - girdle-bone is like that of quite old individuals, and the skull, altogether, is a curious mixture of old and young characters. The notochord (figs. 6, 7, nc.) is persistent ; there is, as in the last but one, only one large fontanelle, the tegmen cranii being very short in front, short behind, and not present along the sides in most of the orbital region. Below (fig. 7), the prootics and ex-occipitals slightly overlap the wings of the parasphenoid (pa.s.); but the cruciform synchondrosis is very large. Below, the ex-occipitals reach half way between the vagus-passage (X.) and the stapes (st.) ; x 2 156 MR. W. K. PARKER ON THE STRUCTURE AND above, they climb up to the inside of the posterior canal (fig. 6, p.s.c.), which they flank ; there they are wide apart. The prootic (pr.o.) runs forward to the foramen ovale (V.), and outwards as far as to the tegmen tympani (fig. 6); they show mostly in front, and to a very moderate extent either above or below. Three-fifths of the orbital region is unossified, and the optic fenestra (II.) in this type is larger than in the other (fig. 2). The girdle-bone (eth.) is more developed than in it, and it has ossified exactly all the cartilage that belongs to the ethmoidal region, for it has carefully avoided the anterior sphenoidal, the ethmo-palatine, and the true nasal, regions. The nasal region is very instructive ; the proper roofs (fig. 6) are large cartilaginous pouches that have coalesced with the ethmoid behind, and with the large intertrabecula (nasal septum), (s.7.) in the middle ; they even exaggerate the large size of the nasal region in the Polypedatidee, for they are very tumid, postero-laterally ; a clean chink separates them from the ethmo-palatine bars (e.pa.). In front, the septum grows out into a well-marked prenasal rostrum (figs. 6, 7, 9, pm). (In figs. 6 and 7, both this and the pro-rhinals (p.rh.) are drawn as dissected out, and shown beyond the outer bones ; they were examined by transmitted light.) The form of the larval cornua trabecule is clearly seen below (fig. 7), as they first bend inwards at their first coalescence, and then spread out in front ; their secondary cornua—the pro-rhinals (figs. 6, 7, 9, p.rh.)—are long, slender, and pedate. The secondary upper labials (w./'.w./*.) are well developed ; and the outer segment is an almost perfect “annulus,” with an outer lip over the external nostril: this passage is far from its fellow, as in this “Family” generally ; three-fourths as far as the internal nostrils (7.n.), which, however, are very much larger than the outer. The facial “bow” is not so strongly bent as in the last, and here we have an arrested condition of the applied bony tracts ( pa., pg.), which are extremely feeble, and like those of very newly curtailed Common Frogs. The cartilaginous palato-suspensorial bar is well developed ; the pedicle ( pd.) is large, and the quadrate region (q.c.) is thick, and of the normal length. The condyles, however, are very peculiar; they are obliquely saddle-shaped, and have not the usual resemblance to a kidney. But one of the rarest characters, for an adult, turns up here: this is the retained orbitar process (07.p.), a structure which is large in the Tadpole, in all the Anura, and is to be seen in Polyodon (see BripaE, Phil. Trans., 1878, Plate 57, or.p.); I find it also in the Sturgeon. Tt remains here as a round leafy lobe (07.p.), at the point where the quadrato-jugal and squamosal (9.j., sg.) meet. The Eustachian tubes are large and reniform ; they turn outwards and a little forwards. The annulus (a.ty.) is rather small, and does not unite above ; the stylo- hyals (st./.), coalesce with the auditory capsule. The stapes (fig. 10, st.) is large and oval ; the inter-stapedial (7.st.) is large and semi-osseous : the medio-stapedial (m.st.) is a pistol-shaped rod of bone, which ends in a tongue-shaped cartilage, the extra- stapedial (e.st.) ; it has no ascending process, See gl ee DEVELOPMENT OF THE SKULL IN THE BATRACHIA. Lag Here, again, in the large bony inter-stapedial and the suppressed supra-stapedial we have a curious mixture of what is arrested and over-developed. The mandibles (fig. 8) are well developed and quite normal, except that the articular condyle is very deeply suleate, so as to be the exact reniform counterpart of the quad- rate condyle of other types; this is in conformity with the very deep scooping of the part on which it rolls (figs. 6 and 7, q.c.). The cerato-hyal band (fig. 8, ¢.4y.) and the hypo-hyal lobe (h.hy.) are quite normal, but the large front notch, as in the other kind (fig. 3), is unusually short. The whole plate is large, and here the lateral lobe is notched in front, is of large extent, and is only separated from the hinder small lobe by a gentle emargination, there is an approach to this in the other (fig. 3), but here (fig. 8) we have it well developed, and it is a rare condition. The thyro-hyals (fig. 8, t.hy.) are rather large and only moderately divergent, they show age in a remarkable way ; the inner edge of each is produced into a large wing of periosteal bone; moreover, the basal plate (b.h.br.) is all converted into true bone, except at its edges. More or less calcification often occurs and is typical; sometimes trwe bone appears close to the thyro-hyals ; here, however, the bony substance reaches to the front notch, and runs far into the lateral lobes. The investing bones are exactly such as we see in newly transformed Frogs of the common kind. The fronto-parietals (fp.) are quite similar to those of the young Hylarana (figs. 1 and 2), and the nasals (7.) are small, crescentic shells, covering very little of the cartilage. The marginal bones ( px., mz., q,j., sq.) are frail splints, like the palatine and pterygoid ectostoses. The parasphenoid has similar wings to the other, and its main part is short but very broad ; here again we have the characters of a young Frog. The vomers (fig. 7, v.) are very minute toothless crescents, and there are no septo-maxillaries. Many of the characters of this dwarf are those of a young typical Rana, but the intense ossification of certain regions show that it is an adult arrested at certain points. It differs from the “norma” in the following particulars :— 1. The cranial notochord is large and persistent. 2. The occipito-otic centres are arrested. 3. There is only one, and that a very large, fontanelle. 4. The mid and fore skull are very wide, and the latter has the roofs very tumid, as in young species of other kinds. 5. There is a prenasal rostrum. 6. The palato-suspensorial ectostoses are arrested. 7. The quadrate condyle is of an unusual form, being deeply scooped, and its counter- part in the mandible being reniform with a very definite “ hilus.” 8. The retention of the larval “orbitar process.” 9. The intense ossification of the inter- and the absence of the swpra-stapedial. 158 MR. W. K. PARKER ON THE STRUCTURE AND 10, The unique ossification of the large basal plate, and the crested growths of the thyro-hyals. In the next kind of dwarf “ Polypedatid,’”—a smaller species than the last and from another region,—the signs of arrest are more numerous, and some of these stoppages must have taken place early in the larval stage; moreover there is but little sign of compensating growth, in over-ossification and the like, such as is seen in the one just described. 42. Rappia (Hyperolius) bicolor—Adult female; # inch long. Dog-trap Road, Paramatta, Australia. This is a rather long skull, for the length and greatest breadth are equal. The quadrate condyles (Plate 19, figs. 6, 7, g.c.) end opposite the middle of the medio- stapedial (m.st.). The mid skull is one-third longer than the other two regions, which are equal to each other in this respect; they are both moderately broad, but the mid skull is very broad. The interauditory region is nearly as wide as the interorbital, and this latter increases in width forwards. The “tegmen cranii” is® very slight in front, largely covers the hind brain, and there is left a fontanelle (fo.) which is evenly egg-shaped, and has its wide end in front. There is a very narrow band of cartilage bounding the fontanelle on each side, but this is not continued downwards into the wall, which is membranous, except for a short distance at each end. There are two rather large secondary fontanelles in the extensive interauditory roof. The occipital condyles (0c.c.) are small, sub-reniform, and postero-inferior ; they are separated by a straight basal tract equal to both in width. The basal cartilage is equal to one condyle in width, and the upper or supraoccipital cartilage is as wide as the intercondyloid space. The four bones (p7.o., ¢.0.) are larger than in the last; the ex-occipitals reach the fenestra ovalis below, and are wider above ; the prootics cover the anterior canal (a.s.c.), nearly, floor its “ampulla” below, and go round, but not in front of, the foramen ovale (V.). There is a moderate tegmen tympani (fig. 6), and the canals and vestibule bulge out above and below (/.s.c., a.s.¢., p.s.c., vb.). The tegmen cranii becomes a narrow band directly it passes into the interorbital region, for the trabeculze (¢7.) have not enlarged, relatively, since the legs of the Tadpole began to bud out, only the intertrabecular tract of cartilage was added. The optic foramen (II.) is large, but the optic fenestra (or fontanelle, 0,fo.), which is large in several types of small Australian Anuwra, here occupies nearly all the orbital region ; it is one-third the length of the skull. The bulging floor of the skull widens towards the ethmoidal region, and then, behind the true nasal region, there is a com- plete girdle for a short distance—not a complete bony girdle, however, yet there are DEVELOPMENT OF THE SKULL IN THE BATRACHIA. a9 thice tracts of bone, as in the young of Rana temporaria (Phil. Trans. 1871, Plate 8, fig. 7, eth.). The broad end of the egg-shaped fontanelle (fo.) is bounded by a very narrow tegminal tract; this tract is ossified, at first, evidently, as in the young of Rana temporaria of the first autumn, by ossification of the fibrous investment, and then by some degree of endosteal deposit. In Dactylethra (Phil. Trans. 1876, Plate 59, fig. 1, s.eth.) the ossified membranous tract is large and T-shaped, but it does not unite with the extensively ossified chondrocranium; but in the phaneroglossal Anura, any external bone there may be is soon lost in the deeper ossification. In this “dwarf” the superior marginal crescent of bone seems to have started in growth from symmetrical points, but of this I am not certain; it may have been single and median at first, and then have become partly sub-divided afterwards—a very common thing in this group. In this narrow ethmoidal girdle the “axils” are ossified (eth.), and their limited lateral tracts unite with the arch of bone above, inside the superorbital band. That band (s.0b.) is double, there is a narrow tract over the anterior angle of the orbital space, and a wider lobe projecting from the ethmoidal wing; these two are separated by a deep rounded notch. The true nasal region is altogether unossified ; it is of a good size both in length and breadth. The septum (s.7.) narrows a little, forwards, from the perpendicular ethmoid, and ends in front in a shortish, rounded rostrum (figs. 6, 7, 9); this rostrum is margined by the nasal roofs, and the front of this region is sinuous, and retires considerably, right and left. The pro-rhinals (figs. 7 and 9, p.rh.) are smallish and falcate. The broad subnasal lamine (fig. 7, on each side of s.7.) grow out into ear-shaped angular processes. The external nostrils (¢.7.) are as wide apart as in other “ Polypedatide ;” they are almost as far apart as the inner openings (i.n.); the valvular cartilages (w.l/.u.P.) are normal. The palato-suspensorial arches are normal, but feeble, so also are the bony tracts applied to them (pa., pg.); both the hinder forks (pd., g.c.) are short, and the quad- rate is not ossified. The Eustachian openings (ev.) are nearly circular: they are two-thirds the size of the inner nostrils; the annulus (figs. 6, 10, a.ty.) is rather small, and imperfect above. The mandible (not figured) is normal, but feeble ; the stapes (fig. 10, st.) is nearly semicircular, and has a long “ boss ;” the inter-stapedial is not cut off from the dilated and forked end of the stout arcuate medio-stapedial (m.st.), but is a massive, emar- ginate lobe of cartilage. The extra-stapedial (e.s¢.) is the short unossified end of this non-segmented columella; it is an inverted saddle-shaped process, and has no ascending ray. This small “key” is embedded in a thick cushion of fibrous and fatty tissue. The stylo-hyal end of the rest of the hyoid arch is not coalesced above ; the descending bar (fig. 8, c.hy.) is of the normal width for the greater part of its length, but widens before it turns back into the hypo-hyal region (/./y.); this part has no front lobe, but on its outer side there is a small oval “ extra-hyal” 160 MR. W. K. PARKER ON THE STRUCTURE AND (ex.hy.). The front notch of the basal plate (b./.br.) is deep, the plate itself of normal length and width, but the front lateral lobe is absent; the hind lateral lobe and the ossified thyro-hyals (¢.hy.) are normal. The investing bones are, on the whole, like those of a very young Frog; but the fronto-parietals (fig. 6, fp.) are small, falcate shells of bone; their ragged inner edges pass a very little way beyond the boundary band of cartilage, behind; they expand over the antero-medial part of the ear-masses, but are wide apart even there. The nasals (m.) are normal, but cover only a third of the roof cartilage; there are no septo-maxillaries, and the marginal bones (px., mx., qg,j., sq.) are feeble and thin, like those of young Common Frogs. The parasphenoid (fig. 7, pa.s.) is relatively large and well developed; all its pro- cesses are pointed at their free ends, and the basi-temporal projections are uncinate. The vomers (v.) are small, arrested, toothless, sub-faleate, and rather ragged tracts of bone, like those of the young of any normal type. The figures show, of themselves, how far short this small Frog comes of the develop- ment seen in its nearest Hthiopian and Indian relations; it falls short of the “ norma” in the following particulars :— 1. The arrest, during an early larval condition, of the chondrocranium, both above (in front) and along the sides. 2. The feeble ossification, especially in the ethmoidal region, so that the girdle-bone is but little developed, broken up into three centres, and widely unfinished, below. 3. The distinct prenasal rostrum. A very large, open, main fontanelle. A bilobate superorbital (and preorbital) tract. General feebleness of facial structures, both cartilaginous and bony. Small Eustachian hole, and annulus tympanicus. 8. Absence of inter-stapedial segment, and supra-stapedial ray, with stunted extra- a stapedial. 9. Modification of hyo-branchial apparatus. «. No hypo-hyal, or front lateral lobe ; and b. A distinct extra-hyal, either a rudiment of the pectinate “ inter-branchial” of the Selachians or a remnant of the first extra-branchial of the Tadpole. Note—That the open window in the orbital space is like what I have described in another small Australian type, namely, Camariolius (“ Ranidz”) (same Plate) ; it is seen in a less degree in some of the small Hyle (Plate 31, fig. 7), and in the little Bombinator Toad (Pseudophryne) (Plate 42, fig. 2); all from the same region. Second Family. “ HyLoprpz.” First genus. Hylodes. 43. Huylodes martinicensis.—Adult female; 1} inch long. Martinique. This skull differs in many respects from that of the Oriental “ Polypedatidee ;” and the species of Hylodes evidently bear the same relation to the Cystignathine Frogs of a ei Ee DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 161 the Nearctic and Neotropic regions, that the Indian “ Polypedatidw ” do to the typical Frogs of their own region. In general development the skull is about equal to that of Polypedates maculatus and chloronotus (Plates 27 and 26), but the ear-organs are more arrested ; altogether, however, it is a lighter and more elegant skull, and is almost exactly intermediate between the skull of a Cystignathus (of the same size), on one hand, and that of such a true Hyla as H. rubra (Neotropical), (Plate 33, figs. 6, 7) on the other. I shall com- pare it with that of Cystignathus ocellatus and typhonius (Plate 16). The length is nine-tenths of the greatest breadth (Plate 29, figs. 7, 8), and the cheeks rapidly narrow in towards the broad, rounded snout. The fore and middle regions are of equal axial length; the hind skull is only two-thirds their length, but it is of great breadth; the ossification, outer and inner, is typical; the whole structure is elegantly light, but of considerable strength. The occipital condyles (0¢.c.) are small, project but little, are directly posterior, and are nearly twice their own breadth apart; they are separated by a gently emarginate line; the condyles for the mandibles reach as far back as the fore edge of the stapes. The upper and lower median cartilaginous tracts are rather wide, and the superoccipital edge (fig. 7, fm.) is almost as far back as the basioccipital. The prootics and ex-occipitals (pr.o., e.0.) of the same side are continuous, and the bone reaches in front to the optic fenestra (II). The bone, outside, reaches beyond the horizontal canal (h.s.c.) and then there is a large lozenge-shaped tract of cartilage which ends in the tegmen tympani (t.ty.), whose margin stretches outwards and forwards. In front of the ampullz of the anterior and horizontal canals (a.s.c., h.s.c.) there is a blunt hook of bone, looking outwards; this process is also developed largely in Siren lacertina (Hux ey, Art. “Amphibia,” Encye. Brit., 9th edit., p. 758; and WIEDERSHEIM, “‘ Das Kopfskelet der Urodelen,” plate 1, figs. 11, 12, V. F., and plate 2, fig. 18, V. F.). I have worked out this part in Siren, and in the correspond- ing stage in the larvee of T7viton cristatus, and I find it to be a foregrowth of the plaster of cartilage developed beneath more or less of the auditory capsule—an outgrowth from the bevelled edge of the basal plate, or “investing mass.” The hinder roof (tegmen cranii) reaches to the optic fenestra; it has two lesser fontanelles in it. The main fontanelle is a long oval, half the extent of the roof. The fontanelle is narrow because the tegminal edges are wide, equal to the scanty tegmen in front—the top of the girdle-bone (eth.); there is a small open space there, and a lozenge-shaped naked tract of the ethmoid. The temporal fossze are large and round, being margined outside by the prootic spurs; then the rather narrow skull steadily widens forwards, as in Cystignathus ocellatus (Plate 16, figs. 1, 2). Infero-laterally, two-fifths of the mid skull is unossified, the large optic fenestra lying at the end of the soft tract, right and left. In this flattish skull the girdle-bone barely reaches the true nasal region, and does MDCCCLXXXI. Y 162 MR. W. K. PARKER ON THE STRUCTURE AND not ossify its own ale; its axille are shallow. Although elegantly arched in front the snout is very wide, and the oblique outer nostrils (e.7.) on each side are sub-tubular and very large ; they have outside them well developed labials (w./!.w.l’.). The roof and floor (fig. 7, al.n.; fig. 8, s.n.) are very wide tracts of cartilage; there is no prenasal rostrum, and the angles of the floor and the pro-rhinals (fig. 8, s.7., p.rh.) are large, and end in out-turned hooks ; the septum nasi, above and below, is thick and clearly defined. Up to the point where the true palatine cartilage (fig. 8) begins, the ethmoidal wing, which is generally thick and somewhat pointed forwards behind the outer margin of the inner nostril, is here unusally developed, as in the genus Cystignathus (Plate 16) (indeed, in C. typhonius it is ossified continuously with the ethmoid, Plate 16, figs. 6 and 7). Here it is not ossified, but is of great breadth, and is seen outside the post-narial spike of the vomer (n,f, v.)—a part which is also, here, four times its usual size. Beyond the notch which separates the ethmoidal from the palatine regions the cartilage is rather feeble and narrow, and so is the whole palato- suspensorial arch, except in its quadrate region behind. In conformity with this, the two pairs of bones (pa., pg.), although quite normal, are slender and rather feebly developed. The pedicle (pd.) is a small osseo-cartilaginous “ foot,” gliding on the sub-concave facet outside the prootic spur. The angle between the forks of the pterygoid is more than a right angle, and in it we see the large, oval, Eustachian opening (ev.) The cartilage of this arch is nowhere obliterated except above the joint (q.c.) where the quadrato-jugal (q,j.) has freely grafted itself. The condyles are large and normal, the inner convexity being much the larger of the two. Outside the suspensorium and its T-shaped splint (sq.) there is a smallish erescentic annulus (fig. 11, a.ty.); it is three-fourths of a circle, its width and concavity moderate, and its hinder horn two-fold. The stapes (fig. 11, st.) is oval, and has a boss; the columella is slender, the partly ossified proximal end is smaller than usual, and there is no inter-stapedial segment. The bony rod (m.st.) is almost straight, slender, and partly unossified distally ; at that part segmentation has taken place, cutting off the symplectic element, or extra- stapedial. This cartilage (e.st.) is like the bill of a Spoonbill, and has no supra- stapedial fork. The mandible (fig. 9) is quite normal, the condyle (a7.c.) is reniform, the eoronoid crest (ar.) is definite, the dentary (d.) is nearly half the length of the ramus, and the mento-Meckelian (m.mk.) rather long. The stylo-hyal (st.h.) margining the Eustachian opening is confluent with the auditory capsule ; it passes into a slender cerato-hyal (fig. 10, chy.) ; outside its flexure there is a small extra-hyal, and from the arch there is a long slender hypo-hyal horn (h.hy.), which curves towards its fellow in front of the great pre-basal notch. The basal plate (b.4.br.) is of the average width, and withal, very long also ; its lateral lobes are narrow, the hinder longer than the front pair; these grow outwards and forwards, those outwards and backwards. DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 163 The thyro-hyals (t.hy.) are as long as the basal plate at its middle ; they are slender, straight, moderately divergent, and well ossified. The investing bones are thoroughly Cystignathine, and very elegant; the fronto- parietals (fig. 7, fp.), are dilated in the temporal and ethmoidal regions, are hollowed out to form the temporal fossa, and spread well over three-fourths of the hind skull, where they end by a straight margin. They widen steadily in the orbital region, and taken together, their fore-edge is crescentically emarginate—so much so that the great fontanelle is not quite covered, and a large lozenge-shaped tract of the girdle-bone is left naked. The diverging outer angle of each lies on the corresponding nasal. The nasals (7.) have the normal shape; they are large, convex shells, and each touches the septum nasi by its inner edge; their bluntish fore ends leave the snout uncovered for some distance. The marginal bones ( pz., mx., q.j., sq.) are well developed and perfectly normal ; the squamosal, however, only lies on the edge of the tegmen tympani; its post-orbital process is bent outwards, and is curved and long; the descending part is normal. There is a very small, sesamoid septo-maxillary (fig. 7, between w./?. and al.n.). The parasphenoid (fig. 8, pa.s.) is long, well-developed, and normal; the bone is thickened at the cross, and the transverse processes are carinate, the thickening running along them as a ridge; the “handle” is unusually long and slender. The vomers (fig. 8, v.) are very large triradiate plates; there is a post-narial, but not a pre-narial spur; and the front part is triangular, running under the edge of the dilated subnasal lamina (right and left of s.n.); these, right and left, are far apart. The hinder half of each bone, on the contrary, converges towards its fellow, each almost reaching the point of the parasphenoid. From the hind margin a thick rib grows, covered with a rasp of retral teeth; these lobes reach outwards further than the inner edge of the inner nares (7.7.), their thick end is outside, and they are furthest forward there. This hind margin is slightly arcuate, so that they run across the fore palate as a crescentic rasp, just broken at the mid-line; they are quite Cystignathine. ? Most of the divergencies from the “norma” in the skull in this species are very gentle, and yet quite appreciable ; this isa true Neotropical Frog, with the toes dilated. Its main points of difference are as follows :— 1. The nasal region is extremely large, both wide and long; and the orbital region steadily widens towards it. 2. The hind skull is unduly ossified. 3. The prootic region has a projecting spike, as in Siren lacertina, and the larva of Triton cristatus. 4, The palato-suspensorial arch is very slender, 5. The quadrate is largely ossified. 6. The annulus is small and widely open. 7, There is neither an inter- nor a supra-stapedial. y 2 - 164 MR. W. K. PARKER ON THE STRUCTURE AND 8. The processes of the basal plate are all very slender, and there is a pair of small extra-hyals. 9. The squamosals have a very narrow supra-temporal part, and the vomers are of great size, especially their dentigerous lobe. Second genus. Acris. 44. (A) Aeris Pickeringii—Adult female; 10 lines (% inch) long. Cambridge, Mass., U.S. Although essentially a [Zylodes, this minute Frog is worthy to be put into a sub- genus; I therefore retain for it the term Acris, The skull of this species is very valuable as showing the effect of dwarfing in a remarkable arrest of the chondro- cranium, such as is seen in the dwarfs of other genera, in different and distant regions. The small species of Rappia from Australia, e.g. R. bicolor, and similar minute Oxydactyle Frogs from the same territory, e.g. Camariolius, have the same modification of their skull—almost all windows, with but narrow strips of wall ; and this is also seen, but to a less extent, in the small Bombinator Toads of Australia, viz.: Pseudophryne. Hence it is evident that this economy of cartilage is due to the same cause as the general arrest in size of these species. On the other hand, some of the smallest kinds are most ossified, notably a species close akin to this, and to Rappiu bicolor, viz.: the small Rappia from Lagos ; and some of the small edentulous Anura show the same thing, e.g., Zylaplesia—a Platydactyle type : the smallest of the typical Frogs, Rana pygmea, has also a very bony skull. This skull (Plate 30, figs. 1-5) has its breadth slightly greater (one-twentyfifth) than its length; itis semi-oval in its facial outline, has a broad but rounded muzzle, and its quadrate condyles (¢g.c.) end opposite the middle of the columella, The occipital condyles (oc.c.) are large, reniform, and postero-inferior; they are separated by a space nearly equal to the width of both, and this wide notch is crescentic. There is a wide basal or median tract of cartilage, in which the remnant of the notochord is seen to be enclosed in a feeble bony sheath—a cephalostyle—the rudiment of a basioccipital (b.0.). Above the wide foramen magnum (fm.) the roof-cartilage comes well back, and is seen to be partly calcified, showing the rudiment of a supra- occipital bone. This roof is complete up to the same transverse line as that from which the optic nerves (II.) emerge ; from thence the roof is membranous nearly up to the septum nasi (s.7.). This is the large ovoidal fontanelle (/o.); it is two-fifths the length of the skull, and three-fourths the width of the interorbital region. The ossification of the hind skull is feeble, and by the two pairs of normal centres (pr.o., ¢.0.). The ex-occipitals (¢.0.) flank the posterior canals (p.s.c.) above, and surround the twin nerve-passages (IX., X.) below; the prooties (pr.o.) climb over the ampullx of the anterior and horizontal canals (a.s.c., h.s.c.) above, and half surround the foramina DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 165 ovalia (V.) on their upper edge, reaching from thence to the pedicles (pd.). Beyond the horizontal canal the capsule lessens to one-third its breadth, to form the “tegmen”’ (fig. 1). The fore and hind skulls are equal in length ; the mid skull is one-half longer, and relatively broad ; its sides are sinuous, for it dilates a little both in the temporal and ethmoidal regions. The actual floor of the chondrocranium is only half the width of the roof, and the optic nerves (II.) pass out of the end of a long, oval fenestra, which reaches to within a short distance of the ethmoidal wings. Measured from the hinder margin of the actual optic foramen to its fore end, this fenestra takes up four-fifths of the orbital wall, and the girdle-bone (eth.) does not even reach to the fore edge of the fenestra. A narrow band of cartilage bounds the single, great fontanelle (fo.), and as this widens in front it becomes bony ; the fore margin of the fontanelle is bony, but the tegmen cranii thus ossified is a very narrow selvedge, and this band only reaches to the very definitely marked nasal roofs, whilst the bone runs very slightly into the septum (s.7.). Laterally, there projects a narrow superorbital eave beyond the bone, and the ossification projects slightly into the wings. The nasal septum (s.7.) is thick behind and bulbous in front; its junction with the roof-cartilages is shown in front, where the short decurved ‘“ rostrum” (p.n.) has a small projection on each side of it, the right being the largest ; these are the ends of the erescentic roof-tracts, which are large and wide. Below (fig. 2) the floor is very wide, and ends in front in the primary angles of the cornua trabeculze, and in the smaller, secondary, decurved pro-rhinals (s.7., p.rh.). The narial openings are wide apart, the external nostrils (e.n.) bemg three-fourths the distance from each other of the inner holes (7.7.); the outer openings are protected by the normal valves (w/'.u.?.). The palato-suspensorials are slender, but quite normal; from the ethmoidal wing, which thickens round the hinder margin of the inner nostril, the ethmo-palatine becomes a moderate tape of cartilage, expanding into the spiked pre-palatine band, which runs backwards as a post-palatine, with a rounded hinder lobe projecting into the maxillary, and then becoming pterygoid (pr.pa., pg.). The forks of the suspensorium are apparently nearly equal, as seen from above and below, but the quadrate region and its condyle (¢.,¢.c.) pass further backwards, reaching as well some distance downwards and outwards. The palatine bone (pa.) is the normal falcate thin piece, and the pterygoid (pg.) has the usual shape, but, contrary to rule, it covers the cartilage most above and least below. The bulbous pedicle is most seen above ; below, it is confluent with the stylo-hyal (pd., st.h.)—a rare modification, and seen again in Phyllomedusa bicolor (Plate 34, fig. 8). Above, in front of the fork, the pterygoid-bone is rough and hollowed out ; it is thin and lathy where it binds on the inner face of the quadrate. That region is only slightly ossified by the quadrato-jugal (q,j.) ; its condyle (g.c.) is large, elegantly bilobate, with the inner or hinder lobe much the larger of the two. The annulus (a.ty.) is rather small and open above; the stapes (fig. 5, st.) is oval, with an emargination antero-superiorly, and an external boss. The pistol-shaped 166 MR. W. K. PARKER ON THE STRUCTURE AND columella is large and bilobate behind, but the unossified part is not distinct from the medio-stapedial (m.st.). The decurved extra-stapedial (e.st.) is spatulate, and it gives off a narrow supra-stapedial (s.st.), which degenerates into a ligament, above. The stylo-hyal (st.h.), from its junction with the pedicle, runs into the cerato-hyal (fig. 4, c.hy.), without change of breadth ; the hypo-hyal region (h.hy.) is narrower, and there is no projecting lobe. The great notch in front of the basal plate (b.h.br.) is wide ; the plate is short and wide, and the fore lobes are uncinate and but little wider than the hind lobes. The thyro-hyals are long, a little more than half bony, and hooked inwards behind. There is an irregular patch of ‘“ endostosis;” between these bars, in front, as the sign of a tendency to form a basi-branchial bone. The mandible (fig. 3) is long and slender, with a pyriform condyle (a7.c.), a long dentary and mento- Meckelian (d., m.mk.), and a low coronoid ridge to the articulare (a7.). The investing bones are extremely delicate, some of them being less than may be seen in the Tadpoles of other kinds, before the fore legs are free. This is true, especially of the fronto-parietals (f:p.), which are delicate styles over- lapping the girdle-bone in front, and only forming a wall-plate to two-thirds of the narrow marginal band of roof-cartilage. In the temples each bone projects, and binds upon the anterior canal; behind, it becomes pedate, not reaching, however, to the small ex-occipital ; thus the hinder, complete “ tegmen” is nearly all naked. The nasals (7.) have a fuller development, but they are more than their own width apart, and are thin lunate shells. The premaxillaries (pz.) are extended in front of the wide snout, and the main bar of each is long; their nasal and palatine processes are moderately developed. Over the junction of these bones with the maxillaries there is a small irregularly radiate septo-maxillary (s.ma.) ; the maxillary (mz.) is of great length, almost reaching to the hinge (q.c.), but it is thin and not high ; the small tooth-like quadrato-jugal (q¢.7.) is partly joined to the quadrate (q.). The squamosal (sq.) runs one-third further forward than the tegmen (fig. 1, left of au.); it overlaps that eave a little, and is split at its free postorbital end ; the descend- ing part is normal. The parasphenoid is thin and like that of a Tadpole, but it reaches from the front of the “ cephalostyle” (b.0.) nearly up to the notch in the margin of the narrow girdle-bone (eth.) ; its processes are all normal, but the lateral wings are rather limited, and the hind lobe is irregular. The vomers (v.) are like bony spicule ; there is a post-narial, but not a pre-narial hooked spike ; the ‘‘ body ” is in two arcs, the hinder ends in a thick lobe covered with teeth, and the front lobe is pointed and underlies the angle of the floor-cartilage (s.7.). The main points to notice in this skull, as compared with the “norma,” are its general arrest at a condition of growth equal to that of a young Common Frog of the first summer. The particular points are :— 1. Rudiments of a basi- and supra-occipital. 2. Its one large fontanelle. 3. The long orbital fontanelles. —— a a i DEVELOPMENT OF THE SKULL IN THE BATRACHIA. 167 . The small girdle-bone. . The definite prenasal with its lateral projections. . The junction of the pedicle with the stylo-hyal. . The absence of a distinct inter-stapedial. . The absence of hypo-hyal lobes. 9, The extreme delicacy and narrowness of the fronto-parietal bones ; this is repeated oN oO oO ee in the other investing bones, but not to the same extent. A4 (continued).—(B) Aeris Pickeringii.—Larva, 1 inch 2 lines long ; tail, ¢ inch ; hind legs, 4 inch ; same locality. The adult of this species is smaller than that of Camariolius, but its Tadpoles are larger by one half; this specimen is more metamorphosed than the larva of that minute Australian “ Oxydactyle Frog” (see Plate 15, figs. 6, 7). In this larva the chondrocranium (Plate 30, figs. 6,7) is oval in outline, for the hind part of the suspensorial bands and the fore part of the auditory capsules project, and the quadrate condyles (q.c.) are drawn in towards the mid-line. It has gone beyond the Petromyzine stage, for there are two bones, both of them axial and azygous: one intrinsic, the cephalostyle or rudimentary basioccipital (n.c.) ; and the other superficial, the parasphenoid (pa.s.). The first vertebra (c.v'.) is drawn with the crown of its arch cut away ; in it we see the ossifying neural arches, and the relatively huge notochord (nc.) becoming invested with a layer of bony deposit, principally on its upper surface (fig. 6). This is taking place in the mesoblastic sheath, and is two-horned in front. Entering the skull, the notochord suddenly becomes very small, reaches to the same transverse line as the Gasserian ganglia (V.), and ends in a rounded point, which is twisted to the right hand. This styliform apex of the notochord is invested—most above—with a deposit of bone, like that which in the vertebral region forms rudi- mentary ‘“‘centra.” The bony “cephalostyle” has already been described as existing in the adult. The occipital arch is fast finishing, for cartilage is breaking out, above, from the inner edges of each auditory capsule, and this has converted most of the membrane there into the same tissue. Also in front of the notochord there is a long inter- trabecular tract of new, small-celled cartilage (fig. 6, inside t7.), and the orbital walls are beginning to grow over as a super-cranial (teyminal) rudiment ; this is especially seen in front. Between the narial passages (7.7.) the trabeculae have coalesced, converging for this purpose, but there is no appearance here of a distinct mesethmoidal rudiment. The trabecule break free of each other again, to end as free, crescentic, decurved horns, from whose inner edge a crest is growing to fill up the re-entering angle between them. The nasal roofs are only beginning to chondrify, and with the eye-balls, were removed in making the dissection. The auditory capsules began to chondrify directly after the trabeculz and suspensoria, and are now highly developed ; their canals (a.s.c., h.s.c., p.s.c.) bulge largely, and the ampulla of the horizontal canal projects greatly. 168 MR. W. K. PARKER ON THE STRUCTURE AND The tegmen tympani is showing itself, and outside the parachordal floor the oblique fenestra ovalis is nearly filled with the oval stapes (st.), now well chondrified. The unfilled fore end of the fenestra ovalis is waiting until after most of the meta- morphosis is done to receive the columella, which will articulate with the newer parachordal floor at its outer edge. Cartilage has spread from the tegmen tympani into the angular space between the capsule and the otic process of the suspensorium (ot.p.), just as cartilage spreads from the inner edge of the capsule into the mem- branous superoccipital tract—for cartilage, like bone, in many instances, does not heed morphological landmarks. The little cordiate cartilage (fig. 7, sp.c.) which is develop- ing in the opercular membrane, over the 1st cleft, becomes free from the ear-capsule, coalesces with the otic process of the suspensorium, becomes detached again from that part, and then develops into the annulus tympanicus, and is permanently kept as the cartilaginous “ operculum ” of the cleft to which it belongs; the specialisation of both cleft and operculum failing to disturb the original morphological relation of the parts. In like manner the modified hyo-mandibular or columella, will be wedged in between the auditory capsule and the basal plate, exactly as in its simple and generalised counterpart in the Selachians. The suspensorial arch has the normal form; the pedicle (fig. 7, mside sp.c.) is narrow, the otic process (fig. 6, ot.p.) is a thick crest, ascending from the main bar as it suddenly turns inwards to become the pedicle. The quadrate condyle is a small trochlea, and looks inwards as well as forwards, and the terminal or quadrate region (q.) is altogether oblique; it sends inwards an angular pre-palatine spur which. is attached to the cornu trabecule in front of the narial passage (7.7.) by a ligament. The orbitar process (fig. 6, orp.) is large and rounded, with raised edges; the hyoidean condyle beneath it has the usual reniform outline, and is scooped. The wings of the ethmoid are very rudimentary at present, and I find no post-palatine rudiment outside them. The mandibles (fig. 7, mk.) are short, thick, ulniform rods, with the usual hooked angular process ; they are articulated to the lower labials by a wide surface; these latter are each half a thick crescent with a raised lip. The temporary upper labials (u.l.) are large, thin, crescentic flaps fitted on to the end of the cornua trabecule. The hyoid bar (fig. 7, chy.) has the usual form ; it has a large condyle and an unciform stylo-hyal free end ; it has a narrow waist, obliquely ridged, externally, for muscular attachment, and then expands largely, most on its hinder margin, to contract again where it is joined to its fellow, ventrally, by the square, soft- celled, basi-hyal. The branchial arches were not figured in this instance, they were quite normal. The parasphenoid (fig. 7, pa.s.) formed a floor, very exactly, to the newer cartilage of the intertrabecular space ; it had, although it was very thin and young, developed the rudiment of its lateral and posterior angles; the last was directly under the (endoskeletal) bony sheath of the notochord. > ++. —ty