48 BULLETIN 34, UNITED STATES NATIONAL MUSEUM. 



The second basibrauchial elemeut is transverse, generally augulated 

 forwards, and is well ossified. The other parts of thehyoid apparatus 

 that are ossified are the proximal portion of the ceratohyal, the ante- 

 rior part of the first basi branchial, and the free extremity of the first 

 cpibranchial. 



The larva3 are characterized by the long, slender processes of the 

 three branchial laminoe, which bear the vascular firabriie, rather than 

 the laminae themselves, as in Chondrotns. The internal nostrils are con- 

 fined between the maxillary series of teeth and the palatine arch, which 

 is concentric with the former and near to it, and is continued backwards 

 on each side in Hue with a similar series on the pterygoid bones. A 

 relation of nostrils to palatine teeth similar to the above is permanent 

 in Amphiuma, and one intermediate between it and the adult condition 

 of Amblystomai of groups iii and IV characterizes Cryptobranchus. 



The tail and back have a free dermal margin, but there is none on 

 the limbs or digits. The tail is short and deep. 



The following are some of the most readily observed characters which 

 are assumed by the Amblystomae at the period of their transformation : 

 (1) The series of teeth on the splenial bone is shed ; (2) the carpus and 

 tarsus ossify ; (.'5) the tail narrows and lengthens; (4) the branchiae dis- 

 appear; (5) the tongue enlarges and covers the floor of the mouth; (6) 

 the pterygopalatine series of teeth becomes more nearly transverse; 

 (7) brightly colored pigment is deposited in the chromatophorse of the 

 derm. These changes are stated in the order of their occurrence. But 

 in some of the protean species this order is not exactly observed in all 

 individuals, and in consequence of the assumption of one or the other 

 character of maturity in advance of another the number of species has 

 been supposed to be greater than it is. The same irregularity in the 

 successional appearance of structures is well known in the earlier pe- 

 riods of embryonic life, as stated by Von Baeriu the scholia of his Ent- 

 wickelungs-geschichite. In the chick, different portions of the vertebral 

 axis and the abdominal plates may or may not appear in the usual 

 order of succession. 



In Amblystoma the approximation of the period of reproduction to 

 that of transformation varies with the species, and it is evident that 

 the closer this approximation, under the above principle of variation, 

 the more protean will the species be. As we know from the experi- 

 ments of Hogg, Duuieril, and others that metamorphosis is greatly has- 

 tened or delayed by the conditions of temperature and light, what would 

 not be the effect on individuals of such a protean species of a change 

 of topographical situation, such as the elevation or depression of the 

 land f And I have no hesitation in saying that if the peculiarities of 

 series of individuals of A. tigrinum, in the resjiects above eiiumerated, 

 were permanent, they would characterize those series as species as 

 completely as any that zoologists are accustomed to recognize. For 

 the evidences on this head, see the discussion of this species below. 



