1904.] NATURAL SCIEXCES OF PHILADELPHIA. 723 



mid-line. Very soon after its formation 4d divides again. This divi- 

 sion parts it, by a bilateral cleavage, into two equal cells, the primary 

 mcsomcres M and M. This division is illustrated in figs. 28 and 44. 

 The primary mesomeres remain undivided up to the 72-ceIl stage, when 

 they undergo a division of great interest. This division is shown in 

 figs. 33 and 36. By it two small cells are budded ofT anteriorly toward 

 the vegetal pole, and close to the line of juncture of the two mesomeres. 

 This division is, however, not bilaterally symmetrical, hut is, on the 

 other liand, 'plainly dexiotropic, and follows the law of alternating cleavages. 

 The product of the left mesomere thus lies in the furrow formed by 

 the juncture of the two mesomeres and the entomeres 5D and 5d. 

 The product of the right mesomeres, on the other hand, lies between 

 its parent cell and 5D. The next division of the mesomeres (fig. 50) 

 is also not symmetrical. By this division the left mesomere buds off 

 a small cell on its left anterior surface, that is, leiotropically ; on the 

 other hand, the right mesomere violates the law of spiral cleavage by 

 dividing in the same direction as before and placing a small cell to 

 the left of its first product (fig. 50). The next division marks the 

 beginning of trvie teloblastic cleavages. Each mesomere in this divi- 

 sion buds off dorsally and laterally a small cell. Just how many of 

 these divisions occur before the closure of the blastopore I cannot say, 

 but probably not more than two. After the closure of the blastopore 

 the mesomeres begin to shift apart, moving laterally, forward and 

 somewhat dorsally. In the horizontal optical section shown in fig. 57 

 they have reached a position which, I think, is their final one. In this 

 movement they not only have changed their position with regard to one 

 another, but also with regard to the median plane of the embryo. Up 

 to the time of the closure of the blastopore the mesomeres are situated 

 at the left of the ventral mid-line, but, as shown in fig. 57, at the 

 close of this movement the mesomeres are bilaterally situated with 

 respect to the median plane of the embryo. 



The cause of this shifting apart is not clear, but I think it is to be 

 explained by the peculiar shape of the entomeres. These are, at the 

 time the blastopore closes, pyramidal in form, with their apices at 

 the blastopore. This point, as will be seen by reference to figs. 53 

 and 55, is just anterior to the junction of the mesomeres. After the 

 ectoderm of the ventral plate has enclosed the mesomeres, they are 

 subjected to a pressure from the ectoderm which tends to force them 

 inward against the narrow ends of the entomeres. These latter press 

 in between the mesomeres and wedge them apart. They are, however, 

 prevented from passing backwards, and in fact compelled to move 



