534 GILBERT 0. BOUKNB. 



carefully for evidence of the formation of spicules or crystals 

 within cells ; but no such evidence was to be found. Were 

 spicules or crystals formed within cells I could hardly have 

 failed to find them, either in Heliopora or the Madrepo- 

 raria in the numerous preparations which I searched with the 

 polariscope. I have found crystals in certain cases both in 

 Heliopora and in Caryophyllia; and have nearly been 

 deceived by them, but they occur in the endoderm and par- 

 ticularly in the zooxanthellse, and I have no doubt that they 

 are aleuron crystalloids, though I have not specially tested the 

 point. 



Moreover I am convinced that the structure of the Madre- 

 porarian skeleton does not admit of a theory of spicular 

 formation. I can take no exception to Mrs. Gordon's draw- 

 ings and descriptions of the numerous sections, and most of 

 the surface views of Madreporaria which she has given. They 

 are faithfully delineated, and constitute an important addition 

 to our knowledge of corals. But I join issue with her on the 

 subject of the ultimate skeletal unit. She has dismissed the 

 " crystalline sphteroid " described by von Koch in almost con- 

 temptuous terms, " For what is this sphseroid ? Is it a 

 spicule? Wc can see and separate the spicule, then why not 

 the sphseroid? It is noteworthy that the sphseroid has been 

 presented from first to last by one or two transverse 

 sections, no longitudinal section has demonstrated it.'' 

 Mrs. Gordon's memory is at fault. The sphaeroid was repre- 

 sented very clearly by von Koch between the base and the 

 surface of attachment in an Astroides larva, and the section 

 was a longitudinal one. Quite recently De Lacaze-Duthiers, 

 in a beautifully illustrated memoir on the development of 

 Caryophyllia Smithii and Balanophyllia regia (24), 

 has completely rehabilitated the crystalline sphteroid or 

 "sclerenchymatous nodule," if that name be preferred. Speak- 

 ing of the larva of the former species, he says, " Lc premier 

 depot de la charpente calcaire se produit sous la forme d'un 

 semis de globules tres refringents." His illustration (loc. 

 cit., pi. ix, tig. IG) leaves no doubt on the subject; the first 



