XVir. SPONGJ.K 
12 Spoil g. 
and Spongilla are also described. In all observed species the segmenta- 
tion is more or less as follows : — As the result of the first two cleavages, 
four equal blastomeres are formed, standing meridionally, and at right 
angles to one another. The third furrow is horizontal, but divides the 
four blastomeres unequally, and, from that point, we have unequal 
segmentation, the difference between micromeres and macromercs being 
distinct. A segmentation cavity, which appears already at the eight-cell 
stage, exists only for a short time. The micromeres, at the anterior pole, 
go on dividing more rapidly than the macromeres at the posterior pole, 
and soon partly overgrow the latter. At that stage the micromeres con- 
tain very little yolk, and have a nucleus with a close framework of 
chromatin ; tho macromeres contain a good amount of yolk, and 
have a vesicular nucleus. The micromeres aro transformed into flagel- 
lated columnar colls, and cover tho anterior and lateral portions of 
the oval larva. The macromeres are differentiated in the way described 
by Maas in his last year’s paper [see Zool. Rec. 1892, Spongke , 
p. 12]. Spicules are developed at a very early stage. Two types of 
larvae can be distinguished amongst the true Cornacaspongke. In the one 
type, the flagellated cells simply stop at the posterior end, and, whilst 
they are coloured red, the internal mass of cells, protruding at the 
posterior end, shows no colour but that of ordinary protoplasm. Esperia , 
Myxilla , Desmacidon , Cluthria , Dictyouella , Axinella , belong to this 
group. Iu the second type a circle of specially long flagella is found 
where the ordinary flagellated cells stop, and they are supported by 
larger cells. The posterior pole of the larva is pigmented — especially the 
large flagellated cells, which form a ring of violet, browm, or yellow 
pigment — whilst the rest of the body remains white. Iteniera , Challnula , 
Gellius , Pachychalina , Toxochalina , belong to this group. This shows 
that there is a close relationship between the Ho morrh aphklui and Ilete- 
rorrhaphidai , on the oue hand, and betwoen tho Desinacklonkhv and 
Axhiellkhe on the other. In the larvie of the Ceratosa ( e.y , Hlrcinki) 
there is not a ring of flagella, but a circular area covering the posterior 
pole. These larvee are to be regarded as belonging to the second type, in 
which the ring of flagella became extended over the whole posterior end. 
Similar are the larvte of the SpongillkUe. Tho larva of Aplysilla, as 
described by Delage, seems to be allied to the first type. This would 
point towards a polyphyletic origin of the Ceratosa from the Cornacu- 
spongice , therefore not in the sense meant by Lendenfeld, who places 
Aplysilla , &c., in a special group ( Hexaceratina), and regards it as closely 
allied to the Hexactinellida . 
After pointing out the homology between the chief stages of the 
development of the silicious Sponges and those of Sycandra , and com- 
paring their development with that of the Ccdenterata and other Metazoa , 
Maas comes to the conclusion that either the germinal layers of Sponges 
cannot be homologized with those of other animals, or else that tho 
flagellated cells of sponge embryos correspond to the ectoderm, the 
large cells to the eudoderm. The systematic position of the Sponges is 
