278 DISCOVERY REPORTS 



S. charcoii was characterized by the radial canals being provided with slender lateral 

 branches with a tendency towards anastomosis. Vanhoffen and Lengerich, however, 

 observed a similar branching of the canals in undissected specimens of S. kergiielensis, 

 but were unable to see any branches in sections ; they were inclined to think, therefore, 

 that the apparent branching of the radial canals was due to misinterpretation in both 

 cases. 



The most characteristic feature of S. hodgsoni is the absence of a continuous ring of 

 nematocysts below the umbrella margin, the ring being divided into isolated patches of 

 nematocysts on the basal portion of the tentacles. According to Lengerich (19226, 

 P- 353) the nematocyst ring assumes a similar appearance in strongly contracted 

 specimens of S. kergiielensis. It is, however, rather improbable that the apparent absence 

 of a complete ring in S. hodgsoni could be explained in a similar way, since the descrip- 

 tion and figures of that species are based on six well-preserved specimens. A re- 

 examination of the type specimens is necessary to decide this question. 



Gilchrist as well as Briggs divided the species known up to 1920 into two groups 

 according to the position of the clusters of nematocysts on the upper branch of the 

 tentacles. In S. charcoti, hodgsoni, and kergiielensis the clusters of nematocysts (ex- 

 cluding the terminal one) are placed on the lateral sides of the tentacle, whereas in 

 S. vallentini, capensis, and haszvelli they are median in position, placed on the upper 

 (aboral) and lower (oral) sides of the tentacle. Gilchrist and Briggs both emphasize 

 that the median (oral and aboral) position of the clusters of nematocysts was constant in 

 all the numerous specimens of S. capensis and haswelli examined by them, and in the 

 specimens of 5. vallentini collected at the Falkland Islands in 1901 the position of the 

 clusters is likewise found to be constant. The lateral position seems to be equally 

 constant in S. charcoti, hodgsoni and kergiielensis, so that there is every reason to suppose 

 that the position of the clusters of nematocysts is a suitable character for the distinction of 

 species in this genus. In the four species from Hawaii described by Edmondson (1930) 

 the position of the clusters of nematocysts is very peculiar. In S. oahuensis they are 

 confined to the upper (aboral) side of the tentacles ; in S. alternota they are likewise 

 placed on the upper side, but in two alternating rows; in S. bilateralis there is one 

 cluster on the upper side and two on the lateral sides, and in S. acuminata there are two 

 on the upper side, one on the lower side, and one on each of the lateral sides. This might 

 seem to diminish the taxonomic value of this character, but the absolute consistency, 

 pointed out above, in the arrangement of the nematocyst clusters in each of the southern 

 species must lead to the conclusion that 5. charcoti, hodgsoni, and kergnele?isis are 

 specifically distinct from 5. vallentini, capensis, and haswelli. 



The three last mentioned species are evidently closely related and are only dis- 

 tinguished from each other by characters which are variable in all of them. The number 

 of radial canals is six to seven in S. vallentini, six in capensis, eight (sometimes six or 

 seven) in haswelli. In S. vallentini the number of tentacles does not seem to exceed 

 twenty-four; in S. capensis "the number of tentacles varies very considerably, from six 

 in the youngest to about forty in the largest"; the largest specimen of S. haswelli had 



