MEXICAN TAILLESS AMPHIBIANS 75 



mining what is a family character. The phylogenetic importance of 

 the various characters that have been selected from time to time to 

 express relationships of supposedly allied groups of genera is in most 

 cases rather doubtful, for the grouping is based solely on recent 

 genera. In view of the admitted deficiency in the paleontological 

 record of an undoubtedly old group like the Anura, a classification 

 based upon structures that have been either remodeled or modified 

 may fail to eliminate all cases of convergent and parallel develop- 

 ment. By definition the limits of families can be extended or 

 restricted as the taxonomist so desires. Difficulties familiar to all 

 are encountered when one attempts to set up arbitrary limits or 

 structural features that will invariably characterize all genera of the 

 family. If we grant that families and genera are derived from older 

 types, then we may expect to encounter overlapping structural 

 features in related groups. 



It may be conceded that any classification based on dentition 

 does not adequately express true relationships. There seems to be 

 an evolutionary tendency in some groups toward the elimination of 

 teeth and equivalent stages in this process do not necessarily indicate 

 relationship. Engystomops and Bufo lack maxillary and vomerine 

 teeth. There are, however, important differences in the mode of life 

 history. Engystomops and Bufo are very much alike externally, bat 

 the former lacks the characteristic protruding parotoid glands of the 

 latter, and furthermore their life histories are quite unlike. The toads 

 of the genus Bufo lay their eggs in long or short strings in the water, 

 while the leptodactylids lay theirs in masses or clusters either on land 

 or floating on the water's surface. 



According to Noble,^^ Syrrhophus "is a polyphyletic group of species 

 of Eleutherodactylus which have independently lost the vomerine 

 teeth." Granted that this is the case, a stage in this general process 

 is represented by Eleutherodactylus mexicanus, in which all steps from 

 toothless to a normal vomerine series may be encountered in a series 

 of individuals. Both Eleutherodactylus and Syrrhophus have T- 

 shaped terminal phalanges and the same type of sternal construction, 

 including a cartilaginous style. In Leptodactylus the mesosternum has 

 ossified. Tomodactylus seems to be related to Eleutherodactylus and 

 Syrrhophus, since it has T-shaped terminal phalanges and a cartilagi- 

 nous mesosternum, but lacks vomerine teeth. Engystomops is more 

 distantly related to the Mexican leptodactylids, for it lacks both 

 vomerine and maxillary teeth and possesses terminal phalanges -wath 

 nodular extremity, as well as a straplike bony mesosternum with 

 each branch of distal bifurcation furnished with a cartilaginous plate. 



The egg-laying habits of these leptodactylids indicate that they 

 have taken advantage of the limitations of their mode of life. Of 



" Noble, G. K., Bull. Amer. Mus. Nat. Hist., vol. 46. art. 1, p. 18, 1922. 



