No. 3.] THE EMBRYOLOGY OF THE EARTHWORM. 397 
which (3, 4) is intermediate in size between the four (or three) 
small (5, 6, 7) and the two (1, 2) largest cells. In some cases, 
however, the large cells undergo this division when only three 
small cells are present, as shown in Fig. 10. At the end of this 
period, therefore, the embryo consists of seven (Fig. 10), or eight 
(Figs. 12, 13), cells, namely: two large equal cells (1, 2, Figs. 
10, 11), two middle-sized equal cells (3, 4), and three or four 
small cells (5, 6, 7, 8). The latter are probably situated at the 
upper or ectoblastic pole. 
From this time forward the divisions cannot be followed in 
detail, but I do not believe that any definite order exists. When 
the embryo consists of not more than twelve or sixteen cells a 
large segmentation cavity appears, and the embryo becomes a 
blastula, the walls of which consist of a single layer of cells 
(Figs. 19, 20). At one pole these cells are smaller than at the 
other, but it is not possible to draw any definite line between 
micromeres and macromeres. The structure of the embryo at 
this period may be understood by a study of Figs. 15-10, 
which are accurate drawings of a single individual in several 
different positions, so that every cell may be seen. There are 
in all thirteen cells, which do not perceptibly differ in the char- 
acter of the protoplasm, and I am unable to say what is the pre- 
cise relation of these cells to those of earlier and later stages ; 
or to recognize the future primary mesoblasts, though it is 
possible that they are already present. 
Comparative. — From the foregoing account it will be seen 
not only that I have had no better success than Kleinenberg in 
following the details of the cleavage process, but that my results, 
as far as they go, do not fully coincide with his; and I can only 
reconcile the discrepancy by supposing that, aside from indi- 
vidual variation, there are some constant differences between 
L. fetidus and L. “ trapezoides.”’ The two-celled and four-celled 
stages appear to be essentially the same in these species; and 
the same is true of the three-celled stage, though it seems to 
occur more rarely in ZL. fetidus. Kleinenberg’s six-celled stage 
corresponds closely with that shown in Fig. 7, but it would 
seem, from his description, to arise in a quite different manner. 
From this point to the full establishment of the blastula I am 
unable to correlate Kleinenberg’s observations with my own. 
A comparison of my figures with those of Whitman on C/ep- 
