lOG ON THE STRUCTURES AND DISTRIBUTION OF THE 



Carpus and tarsus cartilaginous. 



Vertebrae amphiccelian. 



Occipital condyles sessile. 



Plethodontid^e. — Genera Plethodon, Hemidactylium, Spelerpes, Geotriton, and 

 Batrachoseps. North America; seven north neotropical species, one in Siam, one in 

 South Europe. 



This family, with that preceding and that following, have been regarded by Dr. 

 Gray as subtypes of one family, — his Plethodontid^e. Dr. Ilallowell elevated the 

 Amblystomidai to its present rank in 18-39, and I have followed him, and can now 

 add several confirmatory characters. Nevertheless, tiie frontals are not embraced by 

 the parietals and prefrontals in Onychodactylus, but are as in Plethodon ; I do not 

 know the structure of its tarsi and carpi. In Ensatina also, the premaxillary embra- 

 ces a fontanelle, as in the present family. Eschscholtz correctly represents Batra- 

 choseps attenuatus as without prefrontals. An elongate process of the frontal 

 occupies only part of its place, forming no suture with the maxillary; this is quite 

 different from Desmognathus, where the orbit is completed by the union of frontal 

 and maxillary. In Batrachoseps q u a d r i d i g i t a t u s the prefrontal occupies this 

 depression as an elongate vertical scale. 



In Spelerpes rubra the quadratum presents a small internal anterior ala, which 

 has a superficial resemblance to a pterygoid. In this species there is apparently an 

 azygus bone behind the premaxillaries ; this is, however, only the exposed extremity 

 of their united spines, which are nearly or quite isolated by the approximation of the 

 anterior parts of the nasale. It does not occur in the S. s a 1 m o n e a. 



All the characters of this family are those of low developement, and approximations 

 to the larval condition, except the loss of the pterygoid ; one of the species exhibits a 

 subocular cirrhus, which occurs in some of the Gymnophidia (Coecilia) and Dactyle- 

 thra among Anura. It is probably the persistence of that long subocular tentacle 

 characteristic of the early larval stage of Salaraandrida) and Pleurodelidce {e.g., 

 Salamandra Notophthalmus), and of a later larval stage of Dactylethra (vid. Wyman 

 and Gray), where they resemble the appendages of the Siluridje. They have been 

 called crochets by Rusconi, and homologized with the cylindric cephalic processes of 

 the larval Rana, with what correctness remains to be proven by observations on 

 other types. In the young'larva of Bufo a m e r i c a n u s they are not developed, 

 but the transverse black line, at whose extremities they appear in Rana, instead of 

 vanishing, becomes a fissure, separating two longitudinal lateral lips from an equally 

 prolonged transverse anterior lip, which has much the form of the labrnm of an 

 insect, as, e. fj , Locusta. 



When the remains of the intervertebral cartilages are visible, they adhere to the 

 posterior of two vertebrae, except in the single specimen of Hemidactylium scut a- 

 t u m, which I have examined, where they adhere to the anterior. 



