DEVELOPMENT OF SOME STAIiEISHES. 563 



or less definitely in longitudinal rows. The Asteriidse, on the 

 other hand, have inconspicuous marginals, and the ossicles of 

 tlieir abactinal skeleton form a reticulum. 



Stichasier 7'oseus is by no means abundant in the Firth of 

 Clyde, and only one facultatively ripe specimen, a female, was 

 available to me last summer (1915). 



The ovaries resemble those of Asterias riibens or ^4. glacialis, 

 each being a single, simply-branched, much lobulated sac lying 

 free within the body-cavity except at its root, where the single 

 efl'erent duct joins the body-wall. 



The eggs are small like those of A. rubens, pale in colour, but 

 with a reddish tinge. They were unri[ e when first shredded out 

 into sea-water, but underwent the maturation changes shortly 

 thereafter. 



Cross fertilization with Asterias glacialis sperm was ti"ied, and 

 proved so far successful that about 1 per cent, of the eggs exhibited 

 membranes of fertilization, and a number proceeded to develop 

 until the earliest bipinnarial stage was reached. From what we 

 know about crossing in starfish, we may infer that the normal 

 development of Stichaster will follow the same course as that 

 taken by the young hybrids. 



In these segmentation was of the total and equal tj^pe, though 

 a slight difference in size between the cells of the upper and lower 

 pole was present as in Asterias rubens and Porania. The blastula 

 was a hollow sphere, its wall being formed by a single layer of 

 cells, while its central cavity contained no mesenchyme. It pro- 

 gressed with the upper pole in advance, and rotated at the 

 same time in the solar or watch -hand direction as viewed from 

 this pole. 



Gastrulation was by simple invagination as in J. rubens, and 

 the gastrula progressed and rotated in the same manner as 

 the blastula. When invagination was nearly complete, stellate 

 mesenchymal cells began to be budded oS" from the blind end of 

 the archenteron. Meantime the gastrula was elongating, and 

 there next appeared on what was to be its ventral side a slight 

 depression bounded in front and behind by the developing trans- 

 verse portions of the preoral and postoral ciliated bands. These 

 bands became defined in the same manner as in A. rtibens, a,nd 

 there was a similar correspondence in the mode of formation of 

 the enteroccelic sacs, the hydropore, the month, and the regions 

 of the alimentary canal. An interesting point noted was that 

 the posterior enteroccelic outgrowth, which is represented more or 

 less definitely in A. rttbens, A. glacialis, and Porania (see p. 553), 

 could not be made out. Unfortunately my young bipinnarise did 

 not differentiate further. "We may, however, infer that had they 

 done so, the result would have been a brachiolarian larva with 

 sucker-attachment during metamorphosis.* 



* A pure culture was obtained this j'ear (May 1916). Development proceeded 

 along the lines described above, but again I did not succeed in getting the later 

 bipinnarial stages. Many of the larvic showed (9th day) posterior enteroccelic 

 bodies similar in origin and fate to those oi Asterias rubens. 



