dunn: amphibia caudata. 447 



7. Amphiuma tridactylum Cuvier. 



Ala.: Greensboro, 4,079. — Miss.: Natchez, 144. — La.: 25 miles N. of 

 New Orleans, 1,629-1,631 (3). 



These are very large, dark above and light below with a distinct 

 line of demarcation. There are three fingers or toes on all the limbs. 



Salamandridae. 



We are confronted with the problem of what name to apply to that 

 mass of species of Salamandridae, which is often called Triton Laurenti 

 (1768), not Linne. The earliest substitute name, as Stejneger has 

 shown, is Triturus Rafinesque 1815. If, however, we subdivide 

 Triton Laurenti. we must ascertain the type, and to it and its allies 

 restrict the generic name Triturus. Fitzinger indicated T. cristatus 

 Laurenti as type of Triton in 1843. 



This aggregate of species is usually separated on the presence or 

 absence of a bony postfronto-squamosal arch. Nevertheless Euro- 

 pean herpetologists, while retaining the genus in its all-embracing 

 entirety, usually recognize certain groups of vicarious species, such 

 as the Euproctus group, the Pleurodeles group, etc. 



The species of these groups which are in many instances scarcely 

 more than local races are not marked by constancy in the supposedly 

 important character of the postfronto-squamosal arch. When it is 

 found that in the Euproctus group, aspcr of the Pyrenees and rusconii 

 of Sardinia have the bony arch, and that montanus of Corsica lacks it, 

 suspicion arises that this character surely has been given a fictitious 

 value. Similarly, such closely allied animals as vulgaris, yalmatus, 

 moniandonii, hoscai, and italicus, which are not at all easy to distin- 

 guish out of the breeding dress, and which are, as Wolterstoff states, 

 in large part vicarious forms, are characterized by no such uniformity' 

 in the temporal arch as we should expect if this character is of great 

 importance. Two, vulgaris and italicus, lack it, the rest have it. 

 It seems that no weight can be placed on a character which separates 

 closely allied forms and throws together forms not at all closely related. 

 All the newts seem to be developing or losing this arch more or less 

 independently. 



If we remove the Euproctus group and the Pleurodeles group the 

 most distinct species is unquestionably Cope's crocatus from Persia. 



