02 ^r. GlSLKN, 



Sp. 7 (St. 451 (!. XII i)- 10; 2,:,- 4 mm. Arms X, 20 mm. some- 

 what .smoollier (liaii in the ])rece(iing sp. — P, 7 T, P^ (i '\\ P., and 

 P3 wanthig. — Mouth a little tlisplaecd. y\iial cone stcongly |)a|»illosc. 



Sp. 8 (St. 4-')) C. XIV S-IJ; 2.Ô-4 mm. Anns XII, Ki mm. II 

 Br-s 2. - P] 7 T. P2 and P^ missing. Anal cone |)a])illated. 



Sj>. 9. (St. 45) ('. XVli 10—14; 3-4,.-. mm. 'A'' and 4'" segment. 

 L = 2 br. Arms X, 15 mm. uï serrate type. P, 5 high T. P., and P^ 

 absent. 



Sp. W (St. 45) C. IX 7—9; 2—2,5 mm. Arms X, 12 mm. — Pj + 

 T. P2 and P.j wanting. Mouth central. 



Sp. 11 (St. 45) C. XVI 10. Arms X. JO mm. P^-P^ wanting. 



Sp. 12 (St. 45) C. XVII 8—9. Arms X, + 15 mm. P^ 9 T. P2 mis- 

 sing, P3 without a comb. — Anal tube smooth. — Perhaps Commwsia 

 ignota minuta. 



Sp. IS (St. 45) C. XIII 9—10; 3 mm. Arms X, broken. R-s con- 

 cealed. — Pj + 6 T, P., rudimentary. 



Above I have brought together all the young Comasterids from 

 the Bonin Island that I have not been able without doubt to identify to a 

 certain species. With regard to some specimens, howevei-, I am rather 

 sure as to which species they belong by comparison with somewhat 

 larger, systematically determinable specimens h'om the same locality. 

 (This is for instance true of Comaster se/ rata sp. juv 1). 



At the same time as the investigation made above I also tried to 

 establish some chai'acteristics of systematical value for the determina- 

 tion of the young Comastei'ids. Even a hasty reflection tells us at 

 once that 'this is a very difficult task. For the systematical charac- 

 teristics in the Comasterids, as in all Comatulids, are taken from the 

 length, breadth, appearance and order of magnitude of the calcareous 

 ossicles which combine to constitute the animal, and these ai'e very 

 different in young specimens and in full-grown ones. — If guided by 

 the figures and statements given above one examines the characters, 

 the following facts are evident. 



1) The cirri in all young forms correspond to a type about 

 the same as the one in Comaster serrata or Gomaietla hracJii/cirra, that 

 is to a number of segments of +10, where the 3'' and 4"' segments 

 are the longest, L = 2 or 3 X bi-. 



