[99l 



APTERYGOTA 



97 



Excepting these variations, all the individuals are essentially alike 

 and nearly all the variations given are found in one lot of specimens 

 (No. 63, Muir Glacier), among which are also the forms arcticus and 

 americanns. Individuals of class No. 1 are clearly niger {jlavescens) . 

 Those of No. 2 depart from the type in having an extra tooth but are 

 more tvpical than No. 1 by having but one accessory spine. Having 

 admitted No. 2 as niger, how may we exclude No. 3, as regards the 

 claws ? The number of spines is normal on the left, and but one too 

 many on the right dens. Considering the numerical variability of the 

 spines, No. 3 could still be called niger. Notice, however, that No. 

 3 is just as evidently a variety of T. americanus Schott. His diag- 

 nosis (1896, p. 172, pi. 16, figs. 6, 7) provides especially for the three- 

 toothed form. Nos. 4, 5 and 6 are clearly americanus, in which 

 Schott himself found great variability and affinities with JIavescens. 



Any distinction between niger and americanus, then, must be arti- 

 ficial and arbitrary. This is not all, for T. arcticus enters the dis- 

 cussion. Schott (1S94, pp. 43-44, taf. 3, figs. 8, 9) distinguishes 

 arcticus as having (1) four teeth normally on each superior claw (five 

 may occur on any foot, but his statement, " doch scheinen 4 Zahne auf 

 alien das normale zu sein," holds, nevertheless), (2) "Spina? den- 

 tium simplices, septem vel interdum octo, intima parva." The only 

 apparent differences, therefore, between arcticus and americamts are 

 the absence of a tooth on the inferior claws of arcticus and of two large 

 spines beside each dental scale. Now the tooth mentioned was pres- 

 ent on most of the Alaskan examples of arctictcs, although not re- 

 ferred to by Schott, in whose specimens it was very likely absent. As 

 to the accessory spines, one such is indeed mentioned by Schott and 

 several of the Alaskan specimens, which occurred with typical arcticus 

 and were unlike it in no other respect, had two well developed acces- 

 sory spines. Therefore, arcticus and americaitus merge together. 



Comparing arcticus directly with niger, the former, when it has 

 one accessory spine, agrees to that extent with the latter ; the teeth on 

 each superior claw of arcticus are not known to be less than four, and 



