30 
GEORGE BROOK. 
Soc. Journ. Zool./ vol. xviii, pp. 273 — 290, and unless other- 
wise specified, my remarks must be considered to refer to this 
species. 
It will he well to divide the question under consideration 
into two branches. 
1st. The origin of the periblast (Agassiz and Whitman) = 
parablast (Klein); Dotterhaut (CEllacher) ; Rindenschicht (His) ; 
membrane intermediaire (van Bambeke) ; yelk-hypoblast 
(Ryder) ; and intermediary layer of American authors. 
2nd. The part played by the periblast in the process of 
invagination. 
It has usually been considered by authors that whatever 
might be the ultimate relationship between the blastodisc and 
the periblast, one thing was clear at least, that they originated 
independently of each other. Hoffmann (2) asserts that he has 
actually seen the first cleavage process take place, and that 
the first “ spindle ” divides equatorially, dividing the egg into 
two parts, the germinal disc and the food yolk with a thin 
layer of protoplasm separating the yolk from the germinal 
disc, and that this thin protoplasmic layer becomes the peri- 
blast. Agassiz and Whitman, however, assert that the first 
cleavage process is meridional, and that the periblast is after- 
wards formed from the marginal cells of the segmenting disc, 
which, when once separated, never again unite with it. I 
cannot say from actual observation in which direction the first 
cleavage is made, but in Trachinus the periblast arises inde- 
pendently of the germinal disc. As the thin protoplasmic 
layer settles down to the lower pole of the egg, the majority of 
it is included in the first two cells of the blastodisc. As if 
not to waste any material the remainder collects around this 
disc, and is afterwards developed into the periblast. I have 
sometimes observed as early as the two-cell stage, when seen 
in optical section, a thin granular layer of protoplasm under 
the blastodisc ; and in later stages I have sections showing 
a lower lens-shaped mass of cells (the lentille of van Beneden) 
differing altogether in structure from those above, and which 
possibly forms a central portion of the periblast, but this is 
