14 
EMBRYONIC DEVELOPMENT 
and 28 exhibit the first step of this process. Sometimes, as in fig*. 26, when the yolk is 
divided into two unequal halves, a third smaller, sphere may be seen making its ap- 
pearance between them and growing larger, while one of the latter is diminishing. 
Therefore, the fact that the yolk is sometimes divided into three spheres, during 
the process of the division, only indicates a modification of the general rule, and not 
a new rule. 
Next we come to the number four, and instead of four equal parts, we have two 
small spheres, appearing like hernise in the middle and on both sides of the tw r o halves 
(fig. 30). Figs. 31 and 32 exhibit two other stages of the same phase. Occasionally 
the two halves elongate and repeat the process of lateral depressions and grooving 
(fig. 33). 
Then comes the case with five spheres : a small one, with four of equal size (fig. 
36). Six and seven will follow (fig. 37), then eight, which coincides with the mathe- 
matical number. 
I have not seen any case with nine spheres; these, undoubtedly, have escaped my 
notice. We have ten (fig. 41), and eleven (fig. 42). Beyond these numbers it be- 
comes difficult, if not impossible, to follow out the irregular process. Even in the 
regular course of the division, the number of the spheres is difficult to ascertain be- 
yond sixteen (fig. 43), for, soon appears the mulberry shape (fig. 44), in which the 
now very small spheres of division are either equal or unequal in size. 
Soon afterwards the mulberry shape itself vanishes, when the eggs reassume a form 
and an appearance similar to what it was previous to the division, with this difference, 
however, that the vitelline sphere has become a little larger (fig. 46). 
It has been a matter of some controversy as to whether the spheres of division were 
surrounded by a membrane or not. When the egg was divided into four of these 
spheres, the latter could be distinctly separated from each other (fig. 35). The same 
was effectuated when eight spheres were present (figs. 39 and 40). There can be no 
doubt, therefore, that the spheres of division are perfectly independent of each other. 
Under a heavy pressure (fig. 40), the peripheric line of the spheres, or else their 
membrane, seems to penetrate into the mass where the central transparent spaces, 
although not clearly circumscribed by a membrane, are still present. I am inclined 
to believe that there exists a temporary cellular membrane, forming and disappearing 
alternatively during the labor of the division of the substance. For, the vitelline mem- 
brane in an egg in process of formation, is itself not more conspicuous, still the yolk’s 
sphere is perfectly circumscribed, though not yet surrounded by either an albumen 
or a chorion. 
