[iOl] APTERYGOTA 99 



(No. 60) ; nine, Sitka, June, 1S99 (Nos. 64, 71) ; one, Yakutat Bay, 

 1899 (No. 69) ; fifty-five, including intergrades with niger and amer- 

 icauus, Muir Glacier, west side, June 12, 1S99 (No. 63). 



In arcticus there is a strong tendency toward a double series of dental 

 spines, a peculiarity limited to arcticus, so far as I know. The nature 

 of the doubling is shown in fig. 51, in which certain of the proxi- 

 mal spines are each laterally accompanied by an extra spine. The ad- 

 dition of spines begins at the base of the series and proceeds distally ; 

 in fig. 52, only the basal spine of the right dens is repeated, the re- 

 maining spines being single. 



There is no question about the identity of these specimens. Four 

 is the normal number of teeth for a superior claw, as Schott says ; on 

 one pair of hind feet I found five on the right and four on the left foot, 

 a variation mentioned by Schott. He neither describes nor figures a 

 tooth for the inferior claw ; such a tooth was distinct on most of the 

 Harriman specimens, however, although occasionally obscure or even 

 absent, especially on small individuals. The dental spines, rarely 

 eight, in a normal series, were as often six as seven. The number of 

 spines increases with the size of the individual. I may add that the 

 dental scales disagree with Schott's figure by being acuminate instead 

 of rounded. 



The preceding description is based upon nearly typical specimens. 

 To describe the varieties of arcticus would be to describe nigcr and 

 amcricanus again, as arcticus varies into both those forms. The 

 presence of an accessory spine beside each scale and a slight reduction 

 in the number of teeth for the superior claw, variations which actually 

 occur in the specimens from the Muir Glacier — transform arcticus 

 into niger. The variations leading into americanus are given be- 

 low and I have already shown (p. 99) that between americamis and 

 niger proper, no natural distinctions exist. 



Since its discovery by the Vega Expedition in 187S-79, arcticus 

 has never been recorded. The types occurred in colonies at Pitlekai, 

 Chukchi peninsula, eastern Siberia. 



TOMOCERUS NIGER Bourl. var. AMERICANUS Schott. 

 (PI. vni, fig. 53.) 



Tomocertis americatius Schott, Proc. Cal. Acad. Sci., VI (2), p. 172, pi. 16, 

 figs. 6, 7, 1896. 



Among the many specimens of typical arcticus from the Muir Gla- 

 cier are several which agree with arcticus in every respect save that 



