I 



I 



214 BULLETIN :fl, UNITED STATES NATIONAL MUSEUM. 



bale, wall osseous internally. Preniaxillaries consolidated. Occipital 

 condyles on cylindrical pedestals." 



ruoTONOPSiD.i;: ">"o anterior axial cranial bone. * * * Parietals 

 and pref'ontals prolonjjed, meeting aiid enibraciniLj IVontals. Wall of 

 vestibule membranous internally, rremaxillaries separated. Occipital 

 condyles sessile." 



The following observations were made on tlu^ Ampliiumid:e: "Tlie 

 (>ccipital condyles and temporocervical tendon are quite i>s in Desmojjfua 

 thus; they have not been previously described.* In Amphlnnut iiiians 

 there is a minute non-articulated bone on the suture between the o. o. 

 froutalia and piefontalia in the situation of tlie lachrymal. There are 

 some a])proximations to Ca-cilia in Amphiumida-. It does not appear 

 to have been noti(!ed that the * * * free nuuf'in of the frontal seems 

 to foreshadow the overroolinj^ of the oibit and temporal fossa seen in 

 Ca'cilia. Theic is also a v»'ry iar,i;e foramen or canal i)assinjf throii.uh 

 the o. maxillare from lu'ai' its middle to the orbit, foreshadowiii};- the 

 (■(DKilis trntdciili/crus of C.vv.lVux : a narrow one occurs in (he same situa- 

 tion in Protonopsis. rurther, the jjrominent horizontal anterior inle 

 lior pro(^esses of the vertebral centra are the same in Ami»hiuma and 

 Ca'cilia." 



The characters assigned as above to the two families Amidiiumida- 

 and Cryptobraiu'liida- are abundantly suHifient for retaininj;' them as dis- 

 tinct, t The form of the occi[>ital <H)ndyles mij;ht I»e ('xcepted from this 

 estimate, and the axial bone in front of tiu' parasphenoid proves to be 

 abnormally cut otf in the sj)ecimen then examined. The Protonopsida- 

 a^ree with other I'rodela in all of the characters p:iven, ex(!ept in the 

 exclusion of the fnudals from the supraorbital border, and in the mem- 

 branous characteristic of the internal wall of the vestibule. The Am 

 phiumida' ditfer from other I'rodela in the presence of a huj^'e ethmoid 

 bone (the one referred to .is ? vomer in the diajjuosis above (jiioted), in 

 the presence of temjtoral ridjj^es, and of two anteriorly directeil liypapo- 

 j)liyses of the precaudal vertebra-. 



It is interestinj;' to notice that three of the four eharaeters Just cited 

 are shared by the Ca'ciliidie. The jiresence of the ethmoid is ofi'special 

 importaiu;e, as it is an element constantly wanting in the I'rodela. I 

 have not found it in Desmoifuathus, Anaides, Spelerpes, Aiiiblystonia, 

 H;damandra, nor Cryptobranchus, nor is it present in Xecturus or in 

 Siren. It is, on tlMi contraiy, always present in rjeciliida't (see Plate 

 IX, .'5). The double anterior hyi)apophyses ar«' otherwise conlined to the 

 same family. 



The characrters of the hyoid an^hes also distinj^uish this family from 

 the Cryptobranchida', and they ditler from thos(M)f the Pseiidosauria 



" 'I'licy wcni (Icscrilicd liy l>r. .1, (i. I'isclii'r, Aiiiiloiiiiscli. Ahliiintll. iil>. I'ciciiiii- 

 lir.incli. II. Dcrotrt'iii., lirslcs Ilcfr, p. fit, ixil. 

 t I'rocct'd. AliiiT. I'liilosopli. Sdc, IS.-ii;, p. | |-J. 

 IWu'dcrslicini, Aiiiilniiiii' ilcr (iyiiiiMipliidiirn. .Fcnii. I"'7'.l. 



,K^ 



