1884.] 69 [Hyatt. 



publication, but have become of interest as confirmatory, or as 

 supplementary to work published before and since that time by 

 ether authors. We did publish in a popular treatise on Com- 

 mercial Sponges figures of the egg in its early stages, and of the 

 three and six-celled stages and the hollow morula stages of 

 Halichondria and also of a larva of Spongia graminea in Science 

 Guide, No. 3, Boston, 1879, p. 36. In these figures the three-celled 

 stage has cells with nuclei, which we did not see, and the gran- 

 ules around the ova in fig. 23 have nuclei, whereas only the 

 largest of these are ova and should alone have had the nuclei 

 according to our original drawings. 



The mesoderm of Porifera may be represented as a germ- 

 bearing layer in any part of which ova may arise from the 

 amoeboid cells as shown first by Schultze in Sycandra raphanus. 

 (Zeitschr. Wissen. Zool. vol. xxv, Suppl. p. 261, pi. 18, fig. 2.) 

 We have seen almost the exact duplicate of this figure in the 

 mesoderm of Tethya hispida collected early in September, but it 

 is interesting to note, that in our drawings of this and other 

 species in which ova were present, none of the cells with long 

 pseudopodia were observable; all the cells, possibly through 

 the action of alcohol, appeared to have become rounded and ovi- 

 form. 



The ova were more numerous in central parts of the body in 

 all forms, but there was no distinct localization of the reproduc- 

 tive cells, since in some like Suberities they occurred even near 

 the base. This of course does not necessarily apply to buds, 

 especially the peculiar forms known as statoblasts. These are 

 distinctly localized, occurring at the base of fresh-water sponges 

 in many species, though not in all and at the base of the Chal- 

 inula arbuscula and Isodictya * in our salt waters. 



Hyaline plasma was present in ova of Hymeniacidon carun- 

 cula and increased in size faster than the nucleus. The latter 

 became nucleolated, and subsequently granulated, probably by 

 division of the nucleolus, since a double nucleolus, the two parts 

 still in contact, was observed in another species, at a correspond- 

 ing stage. In Halichondria incrustans no nucleolus was observed 

 though it was very likely present since the nucleus became filled 



1 Described since the presentation of this paper at a meeting of the Bost. Soc. Nat. 

 History, Oct. 15, 1884. See Science, Cambridge, 1884, vol. iv, no. 92, p. vi. 



