NO. 2211. GENERA DESMOGNATHUS AND LEUROGNATHUS— DUNN. 421 



Description. — No vomerine teeth. The parasphenoid teeth form 

 two leaf -shaped series which touch at the anterior end and are poste- 

 riorly distant. 



The tail is shorter than the head and body, flattened and finned 

 above and to a less extent below. 



The head length is about four and a half in the length of head and 

 body. The head width is about five and a half in the length of the 

 head and body. In the Roan Mountain specimen the head is rugose 

 Hke grain leather. In the type the head is quite smooth. The legs are 

 stout. In the four from Grandfather Mountain there are two and a half 

 intercostal spaces between the appressed toes. In the Roan Moun- 

 tain specimen there are three. The costal grooves are 13, counting 

 the axillary and inguinal. They are as in most quadramaculata. 

 The color is much like tliat of quadramaculata, save that the dorsal 

 coloration is lighter. The pattern is confused, but usually there is 

 a double dorsal row of light spots without any definite borders. 

 This coloration is intermediate between that of D. quadramaculata 

 and that of D. monticola. The three smaller ones have the belly light 

 while the two larger ones have it black. No sexual differences are 

 evident. 



I refer one larva caught in Grandmother Creek just below the lake 

 to this species. 



Moore's statement that L. marmorata is more aquatic than Des- 

 mognatlius qua,dramaculata, the absence of adults or larvae in small 

 streams near Linville which we searched thoroughly, and this larva 

 from Grandmother Creek, all go to show that this species inhabits 

 large streams. On several occasions we noticed salamanders in the 

 Linville River. All those we could catch, however, turned out to be 

 Desmognathus quadramaculata. 



The larva shows Desmognathine affinities in its stout hind legs, 

 ghstening white gills, and dorsal fin only on the tail. No larvae 

 kno^\^l have aU these characters save those of Desmognathus. It is 

 much too large (63 mm.) for monticola, which transforms at 30 mm., 

 or carolinensis which is even smaller. 



In size and general appearance it agrees with larvae of quadra- 

 maculata collected at Brevard and Linville. It differs, however, in the 

 much longer tail, the shorter and shmmer head, and the slimmer body. 



Compare the proportions of several larvae: 



