18 ART. 9. 1. TJIMA : HEXACTINELLIDA. 



5. The synapticular fusion of spicules does not extend to 

 the upper end of the body. 



6. The strongest parenchymal principalis are predominantly 

 stauractins ; more seldom triactins and diactins. 



7. The dermalia, which vary much in size, should have all 

 the rays running out to a fine point. 



8. The noricome, according to my computation from the 

 figure in the Challenger Report, should measure about 100 y in 

 diameter, which size is about the same as in G. eleyans but 

 considerably larger than in G. speciosa. 



9. Of the tw r o forms of the discohexaster the ordinary seems 

 to be by far the commoner as in G. speciosa. Both forms should 

 be of about the same size. For the hexactinose form, F. E. 

 Schulze has given 170/' for axial length. This indicates the 

 size of the discohexaster to be larger than in G. speciosa, and 

 considerably smaller than in G. elegans so far as the hexactinose 

 form is concerned. 



Heterotella corbicula (Bowerbank) 

 (Plate, fige. 16-23.) 



Euplectella sp. Bowerbank, '58, PI. xxv, figs. 37 and 38 

 {probably also figs. 35 and 3G and PI. xxvr, fig. •">). 



Alcyoncellum corbicula. Valenciennes in Uteris, Mus. d 'hist, 

 nat. Paris. — Bowerbank, '62, pp. 1103, 1104. — Bower- 

 bank, '64, p. 176. — Bowerbank, '67, p. 358. 



Alcyoncellum sp. Bowerbank, '64, PI. vir, fig. 187 {prob- 

 ably also figs. 185 and 18G) and PI. vni, fig. 188 

 (? also fig. 195). 



