57 
which I have selected above show the dependence of the young 
upon continuous warm weather. 
The past spring and summer were unusually cool, and it 
was not until the middle of July that the weather was warm 
for more than three or four days in succession, and my failure 
to find any floating embryos in the open ocean may be due 
to the fact that they were killed by the cold as fast as the 
eggs were laid. After the middle of July 1 found a few em- 
bryo at the surface of the water of the Sound. 
EXCEPTIONS TO THE NORMAL METHOD OF SEGMENTATION. 
The method of segmentation, as I have described it, is the 
normal method, and is followed exactly by a very large pro- 
portion of the eggs — by more than 60 per cent, of them, I 
should think; but a few eggs in every lot present considera- 
ble variation, especially in the earlier stages. 
Plate X, Figures 54 to 62, shows one of the most common 
variations. If a number of eggs be carefully watched during 
the early stages a few will be found to reach the stage shown 
in Figure 13, Plate I, more directly than the ordinary eggs, 
without going through the process of forming the second mi- 
cromere of Plate I, and then obliterating it. Figure 54 is 
an egg two hours and seven minutes after fertilization, and in 
the stage shown in Figure 4. Two minutes later it had as- 
sumed the form shown in Figure 55, which is very similar to 
Figure 5, except that the three lobes of the trefoil are less 
sharply separated. Two minutes later it had assumed the 
form shown in Figure 56, and one minute later the form shown 
in Figure 57. As shown by these figures, the second micro- 
mere does not become distinct, as in Figure 6, but the faint 
indication of it shown in Figure 55 quickly disappears, and 
the subsequent changes result in the separation of the egg 
into two masses, instead of three. Figure 58 is forty-five 
seconds later than Figure 57, and Figure 59 four minutes and 
fifteen seconds later. A single furrow now extends nearly 
across the egg, from the polar globule, and divides it into two 
nearly separated portions — a small one and a large one. In 
