468 TAdK SKOi;sHKRG 



bristles. The s e v e n t h limb has more than twelve bristles. In addition this species differs 

 from other known forms of this genus in being quite without lateral eyes. 



Of those species that I have had no opportunity of investigating myself A. inermis, 

 G. W. MtlLLER, 1906 b, seems to be very closely related to this group. 



For the relation to this group of the following species that have not been investigated 

 by me see the remarlv on^. Grimaldi: Cypridina Mariae, W. Baird, 1850 c, C. oblonga, E. Grube, 

 1859, Cylimlroleheris Mariae, G. S. Brady. 1868 b, Asterope oblonga, G. 0. Sars, 1887, Cylindroleberis 

 oblonga, G. \X. MuLLER, 1894, Asterope Mariae, G. S. Brady and A. M. Norman, 1896, Cylindro- 

 leberis Mariae, J. A. CUSHMAN, 1906, C. Mariae, Ch. Juday, 1907, C.oblonga,^. W.Sharpe. 1909. 



The only species among those described below that has not been mentioned so far, A. aberratu, 

 occupies a somewhat isolated position. It seems to be most closely related to the MiUleri group. 



The characters in which it differs from this group are as follows: — 



The elliptical shape of the shell. 



First antenna: — The marked reduction of the third and fourth joints. 



Mandible: — Second protopodite joint : The backward pointing process : The triaena l)rist- 

 les have from five to eleven secondary spines proximally of the distal pair of spines. Two dwarf 

 bristles are developed on this process. At the middle of the dorsal side of this joint there is one bristle. 



The sixth 1 i m I), unlike that of other known forms, has no posterior ventral bristles. 



The seventh 1 i m 1) has less than twelve bristles. The teeth of the end combs 

 are finely and uniformly pectinated. 



A. elliptica. G. 0. Sars, 1887 is probably rather closely related to this form. As to the 

 position of A. elliptica, A. Philippi, 1840 see below p. 509. 



With regard to the systematic position of those of the species not investigated by me 

 tliat have not so far been mentioned I shall not try to put forward any opinion, as, on account 

 of the incompleteness of the descriptions, this would be so uncertain that its scientific value 

 would be exceedingly small. 

 Which sj/rnrs are the Wh\c\\ of the species SO far known are to be considered the most primitive? 



pnnii ive. ^^ ^j^^ present moment this question can scarcely be discussed. It can only be saitl that 



in one respect — the development of the d-bristle on the first antenna — the Quinquesetae 

 g r o u ]) is more primitive than the others. lAHiether it is als(j to be considered as more primi- 

 tive in (jther characters cannot be decided with any certainty, though it does not seem to lue 

 to be impossible. 

 The lypr species oj The first spccics of this geuus to be described was A. elliptica, A. Philippi, 1840. As this 



'■■> b""*. form — as is shown by the historical sketch, p. 433 — must be said to be unidentifiable as a 



species, it can scarcely be convenient to consider it as a type species for this genus. I suggest 

 instead as the type species A. elliptica. G. 0. Sars. 1887, a species of which, it is true, we cannot 

 say with absolute certainty that it is identical with the form described by PHILIPPI, but which 

 is, however, presumably very closely related to it. (G. W. MULLER makes this identification, 1912, 

 ]). 46, but adds a ((uery.) The form described by G. 0. Sars certainly needs to be re-described, 

 but it must be denoted as one that is identifiable as to its species. It may be noted that tlie 

 type-specimen of this form — according to a written comnumication to me from Professor 



