INVERTEBRATA, CRYPTOGAMIA, MICROSCOPY, ETC. 895 



Tlie ectoderm cells next acquire long cilia, and those in the region 

 of one pole of the now elongated embryo, undergo a slight increase in 

 thickness ; the embryo will afterwards become attached by this, its 

 posterior end. But the difterentiation of the ectoderm into cells of 

 two kinds is only temporary : as it further increases in size the cells 

 once more become uniform. In the meantime the granular cells of the 

 mesoderm increase in number, filling up large interspaces between the 

 rosette-cells, which previously were in close contact. 



The cilia next become converted into slowly moving processes, and 

 eventually disappear, leaving the ectoderm as a layer in which the 

 separate cell-bodies are only distinguishable by the aid of silver 

 nitrate : the outermost layer of the ceils, at the same time, becomes 

 firm and cuticle-like. 



After a pause of two days in development, the formation of 

 the canals begins, — Metschnikoff's account of this process being of 

 the highest interest. These cavities are formed by a grouping of the 

 granular mesoderm cells, which become arranged side by side, around 

 intercellular spaces, so that, in this case, the endoderm is actually a 

 secondary product of the mesoderm. 



2. Anatomical Observations on Ascetta. — The author was able to 

 prove conclusively the presence of a true ectoderm, instead of a syncy- 

 tium as Haeckel maintains, the contours of the cells being well brought 

 out by silver nitrate. Haeckel's statement that the lowest sponges 

 consist merely of ectoderm and endoderm is also denied : a distinct 

 mesoderm was found in Ascetta, in the interior of the granular amoeboid 

 cells of which the spicules were found to be produced, and not 

 in the gelatinous interstitial substance. Nothing new was made 

 out as to the structure of the endoderm, but the observations of 

 Haeckel and Carter as to the passage of ciliated into amoeboid cells 

 were confirmed. 



3. Emhryological Observations on Calcareous Sponrjes. — In Ascetta 

 primordialis the process of development resembles closely that de- 

 scribed above in Halisarca. A one-layered blastula is formed, and 

 from its cells large granular mesoderm cells are produced, which 

 accumulate in, and finally fill up the segmentation cavity. Amongst 

 these largo granular cells occur small elements with few granules, 

 probably the foundation of the endoderm. In A. blanca, the formation 

 of mesoderm cells is confined to the lower pole of the ovoidal blastula, 

 the process having some resemblance to one of invagination. But in 

 neither species is there any true gastrula stage; the inner set of 

 cells forms a mere parenchymatous mass, and never a definite epithe- 

 lioid layer ; from this mass endoderm and mesoderm are subsequently 

 differentiated. 



In Sycandra rapJianus, Metschnikofi^ confirms his own earlier and 

 Schulze's later observation, that the ciliated hemisphere of the blas- 

 tula undergoes invagination within the non-ciliated, and not vice versa, 

 as Schulze formerly believed. In the blastula, previously to invagina- 

 tion, a cavity was observed in the non-ciliated hemisphere, dividing the 

 latter into two layers of cells, an internal in contact with those of the 

 ciliated hemisphere, and an external. The former remain large, and 



