760 The American Naturalist. [September, 
sion see Schultze's short note, ^2, and Roux's rejoinder, 9. I 
think the question whether the first cleavage plane is determined 
by the ovum's structure or not is still an open one. 
As already stated, in the primitive segmentation, both inverte- 
brate and vertebrate, the second cleavage plane is at right angles to 
the first, and also meriodional, while the third plane is at right 
angles to both the first and second, and therefore equatorial. 
In some meroblastic vertebrate ova this regularity is entirely 
lost. 
Differentiation of the ectoderm and entoderm. — As already 
pointed out the essential feature of segmentation is the unlikeness 
of the cells produced ; the manifold variations in the process of 
segmentation depend chiefly on the amount of yolk. 
Minot, in 1877, jj, first established the generalization that «« 
all animals the ovum undergoes a total segmentation during which 
the cells of the ectoderm divide faster and become smaller than the 
cells of the cntodcnn. Compare Fig. 15. There are, however, a 
small and, I think, diminishing num- 
ber of cases, where the process of 
, segmentation is imperfectly under- 
"■■i^ Stood, and which cannot yet be 
'i shown to conform to this generaliza- 
.\ tion. " All the known variations in 
the process of segmentation depend 
merely upon: i. The degree of dif- 
ze between the two sets 
of cells ; 2. the time when the dif- 
erence appears; 3. the mode of 
development, whether polar or by 
delamination,^ either of which may 
or may not be accompanied by axial 
cciuuermai ceii^. infoMiug. In gastropods, planar- 
ians, calcispongc-E, gephyrea, annelida, fish, birds, and arthropods 
the difference is great and appears early. In echinoderms, most 
