So DISCOVERY REPORTS 



"This arrangement of the gonads holds true in all specimens examined, namely: 



" Leptychaster prophiqims, 2 males and 3 females; station 4788. 



"L. anomalus, i male, station 4233; i female, station 4280. 



" L. pacificus, I male, station 3223; 2 females, station 2862." 



I have ascertained that in L. kergiielenensis the ovaries are not as in the northerfi species, 

 but are serially arranged close to the marginal plates; and that L. magnificns and L. 

 accrescens also have both ovaries and testes in series (although not close to marginal 

 plates). 



The northern forms have not been observed to brood their young. They differ also 

 in having somewhat longer, less compressed adambulacral plates with a curved furrow 

 margin, rather than the sharply angular one of kergiielenensis, accrescens, and magnificns. 

 The furrow armature of these three southern species is composed of 3 spines which 

 generally stand upright in a co-ordinated group. In the northern forms, even in 

 arctictis, it is a comb of spines (3-6 in different species, usually 3 or 4) followed by 3 

 longiseries of 3-5 similar spines (Fisher, 191 1, pi. 5, figs, i, 2). The southern forms have 

 relatively larger tube feet. 



In the specimen of L. kerguelenensis the papulae are confined to the margin of the 

 abactinal area, where the plates have very short lobes. Elsewhere the plates are roundish. 

 But in accrescens and mag?iificus the papulae extend completely across ray and all the 

 plates are stellate or strongly 6- or 5-Iobed (in tnagnificus reduced to 4 near margin of 

 area). 



I have discussed (191 1, pp. 52, 53) the difficulty of distributing the north Pacific 

 species between Glyphaster Verrill ( = Parastropecten Ludwig) and Leptychaster . 

 Verrill (1914, p. 327) did this, but his diagnosis of Glyphaster is not correct. He states 

 that "the inferomarginal plates are convex, not carinated". As a matter of fact they 

 are strongly carinated, especially on the lateral face. The deep, narrow, superomarginal 

 and inferomarginal fascioles are lined with very fine spinelets. In Verrill's system, 

 Leptychaster propinqnus would fall in Glyphaster, yet the inferomarginals are as short and 

 oblique as in Leptychaster pacificus and the superomarginals are small. Since the 

 marginals are short, the disparity between the number of inferomarginals and adam- 

 bulacrals is not so great as is emphasized by Verrill for the type anomalus. In an adult 

 specimen of anomalus there are 16 superomarginals to 29 adambulacrals ; in prop in quits 

 there are 22 superomarginals to 29 adambulacrals; in pacificus {Leptychaster in Verrill's 

 system) there are 25 superomarginals to 29 adambulacrals. 



The curious difference in the male and female gonads, common to the 4 species, is 

 further evidence that they are very closely related. The diversity in body form is no 

 greater than that found in Odontaster penicillatus. 



On the other hand, the 4 species differ from the southern forms in the distribution of 

 the gonads and in having the less compressed adambulacrals with a rounded instead of 

 sharply angular furrow margin. The inner end of the marginal series of mouth spines 

 lacks the conspicuous decurrent angle found in Leptychaster accrescens and flexuosus, 

 and less prominently in magnificns and kergiielenensis. After observing the amplitude of 



