98 



Mr. H. N. Moseley on the 



[June 15^ 



and worm-like, measuring \ of an inch in length. They have a thick trans- 

 parent ectoderm, abundantly supplied with the larger form of thread- 

 cells. The spadix in both genera, as the development of the ovum pro- 

 ceeds, becomes divided at its margin into a series of lobes, which lobes 

 subdivide and encroach over the surface of the ovina until more than 

 half the proximal surface of the ovum is thus embraced by the cup of the 

 spadix. The lobes of the margin of the spadix appear just like develop- 

 ing tentacles; aud the spadix of Cry^tolielia was at first supposed to be a 

 developing actinula. The outer, thin, perforated calcareous walls of the 

 ampullae in Errina appear to get thinner as development of the embryo 

 advances, until they fall away or are absorbed altogether, and give 

 free exit to the planula. In Cryptolielia the planulse probably escape 

 through the mouths of the calicles. The endoderm, spadices, &c. are 

 coloured red by a colouring-matter, soluble in spirit, insoluble in gly- 

 cerine, in Polypora, Cryptolielia, and Errina. In the Stylaster resembling 

 Cryptohelia the coloration is dusky green. The green colouring- matter 

 is soluble in spirit, and yields an absorption-band in the spectrum. In 

 Polypora the living layer of coenosarc set free by decalcification is very 

 thick, not merely a thin superficial film as in Millepora ; indeed all but 

 the most central axial regions of the branches of the corals are in actiA^e 

 life. Li the other genera the whole of the coral appears to maintain 

 its vitality, there being no dead region represented by a cavity after 

 decalcification. 



Conclusions. 



Since the observations of Prof. Sars * on the polyps of Allopora oculina 

 it has been to some extent suspected that the Stylasteridae were not 

 Anthozoa, but possibly allied to the JMilleporidee, although the fact was 

 not in any way demonstrated. Milne-Edwards long ago expressed him- 

 self extremely uncertain as to the affinities of Disticliopora, and suspected 

 that it might be an Alcyouarianf. In consideration of the facts now 

 ascertained, there can be no doubt as to the hydroid affinities of the 

 family. The Stylasteridae appear to form a very natural family. They 

 all possess two kinds of zooids. The tentacular zooids are closely simi- 

 lar in form in all the genera ; and in the variations in the forms of the 

 ahmentary zooids all gradations are present. The thread-cells appear to 

 be alike in form in all the genera. In all the gonophores are developed 

 within ampullae. The corals all bear, as far as has yet been ascertained, 

 fixed sporosacs, as do, according to Allman, all deep-sea HydroidsJ. It 

 is possible, however, that forms such as Stylaster sanguineus occurring in 

 shallow water § may bear planoblasts. There can be no doubt that Bis- 

 tichopora will prove closely allied to the other six genera of Stylas- 

 teridae : its well-marked ampullae and two kinds of pores are decisive in ' 



* Sars, Forh. Selsk. Chr. 1872, p. 115. 



t MM. Milne-Edwards and Haime, I. c. t. iii., Appendice, p. 451. 

 I AUman, I. c. vol. ii. p. 155; also 'Nature,' Oct. 28th, 1875, p. 656. 

 § Pourtales, I. c. p. 83. 



