2 SIDNEY F. HARMER. 



species is really identical with that which is found near Roscoff, — ■ 

 of which, however, I am not entirely convinced. 



My own specimens do not show the large papillae mentioned 

 by Hincks from which the polypides are said to issue. This, 

 however, may possibly be due to contraction induced by the 

 reagents in which they were preserved. The septa between the 

 zocecia are clearly seen from the surface of the colony, whilst the 

 embryos are invariably aggregated into spherical clusters, each 

 contained in the tentacle-sheath of an individual whose alimentary 

 canal has atrophied, instead of being "scattered singly throughout 

 the polypidom," as in Hassall's description. In spite of the fact 

 that in these respects the Roscoff species does not conform to the 

 diagnosis given by Hincks for A. poly own, it does not appear to 

 me desirable to give a new specific name to the Roscoff form until 

 the correctness of Hassall's original diagnosis has been verified. 



I have observed individuals with twenty tentacles, the number 

 given by Hassall, although in my own specimens twenty-one 

 tentacles are more commonly present. It should be noticed that 

 these numbers are considerably higher than those which characte- 

 rize most species of A Icyonidiiim. 



Most of my observations on the development were made on 

 sections of portions of the entire colony which had been preserved 

 with corrosive sublimate. The best staining was obtained by 

 using picrocarmine and washing successively with very dilute 

 solutions of picric acid in water and various strengths of alcohol. 

 By this method the nuclei were stained red and the yolk- 

 spherules yellow. Hsematoxylin followed by eosin and borax- 

 carmine followed by haematoxylin were also used for differentiating 

 the yolk-spherules from the nuclei. 



The eggs, several of which develop simultaneously in the same 

 zocecium, are large and possess numerous yolk-spherules (similar 

 to those figured in the embryos) distributed uniformly throughout 

 their protoplasm. During the segmentation and early develop- 

 ment of the embryo, these yolk-spherules are met with indifferently 

 in all the cells, and do not in the slightest degree preponderate in 

 the hypoblastic elements. The segmentation (which is preceded 

 by the formation of polar bodies) is of the remarkable type which 



