96 M. Coste on the Formation of Cells. 
the whole of the grains of which are kept agglutinated by a dia- 
phanous interstitial matter, the retraction of which matter gives 
the whole mass the somewhat geometric regularity which it as- 
sumes. 
Soon (a few hours are sufficient for the accomplishment of 
this phenomenon) the vitelline sphere divides into two nearly 
equal parts, each of which, immediately rendered spherical in 
form by the centripetal retraction of the viscosity which retains 
its granulations in union, presents the same aspect and the same 
composition as the whole from which it emanates. 
This primary division is scarcely accomplished before the two 
secondary granular spheres which are thus formed by a primary 
division of the vtellus become in their turn the seat of a similar di- 
vision, and the same phenomenon being repeated during a certain 
time upon each new segment, the vitellus is finally resolved into a 
considerable number of granular spheres of a progressively dimi- 
nishing volume, but always of the same nature. However Reichert, 
who has made some special researches upon the division of the 
vitellus of the Batrachia, believes he has observed that each seg- 
ment is a true cell possessing an enveloping membrane and gra- 
nular contents. According to him, the phenomenon of the division 
of the yolk would then have a totally different signification to that 
which we have given, and would essentially be nothing more than 
an illusion produced by the liberation of the pre-existing vesicles 
which were inclosed one within another. The vitellus, in his view, 
would at first represent a mother-cell, the wall of which, when ulti- 
mately absorbed, would expose to view two inclosed vesicles which 
form its contents; these two vesicles having thus become free would 
be dissolved in their turn, and each of them would allow two other 
vesicles to escape, which would produce an appearance of a divi- 
sion of the yolk into four segments, and so on, until the com- 
pletion of this illusory division arrived. But although this hy- 
pothesis appears to explain a phenomenon until then but little 
understood, and to corroborate the theory of the exclusive m- 
tervention of the cells for the formation of the tissues, it does not 
follow that we must accept it without examination, and solely 
from its being reconcilable with an accredited system. I have 
therefore examined the question with all that care which its im- 
portance demands, and, after the most minute researches, I am 
perfectly convinced that the segments of the viéel/us or the gra- 
nular spheres are not real cells. Consequently Barry and Berg- 
mann were deceived when they admitted the contrary. 
When the subdivision of the vitel/us is completed, a process 
ensues in each of the granular spheres resulting from this divi- 
sion which converts them into true cells. But before arriving at 
this degree of organization, as we have seen, the living matter 
