DR. E. B. WILSON ON THE DEVELOPMENT OF RENILLA. 
733 
divisions of the nuclei. The egg appears to make an effort, so to speak, at cleavage, 
but has not sufficient energy to complete the division of the vitellus. I shall return 
to this point further on. It is perhaps worth noting that the interval (fifteen to 
twenty-five minutes) between the preliminary change and the first cleavage is nearly 
always considerably greater than the ordinary resting stages (eight to eighteen 
minutes). 
4. In a fourth form of segmentation (figs. 19 to 24), of which a single example only 
was observed, the egg was divided at first into two equal parts by a horizontal 
cleavage, and then incompletely into eight by two partial vertical furrows at right 
angles to each other and to the horizontal furrows (figs. 19, 20). The vertical furrows 
started from the horizontal one at four equi-distant points, and travelled about half¬ 
way towards the upper and lower poles. They stopped abruptly at these points 
however, and the egg passed into a very marked resting stage (fig. 21), during which 
the form was nearly spherical and the furrows could only be seen at the points of union 
with each other. At the next cleavage the egg divided into about sixteen spheres 
of different sizes (see fig. 22). The spheres remained sharply marked and rounded for 
twelve minutes, then flattened together slightly, but five minutes later swelled slightly 
and each divided into two with beautiful regularity (fig. 24). The subsequent 
development was regular and normal. 
5. In one case an egg was observed to divide into two nearly equal parts, and then 
passed into a marked resting stage (figs. 25-27). In several other cases eggs were 
observed divided into four equal parts (figs. 28 and 29). Unluckily, the subsequent 
development was not followed in either case, and I cannot state whether these eggs 
were normal. In any case they are interesting, as filling out the series of different 
modes of segmentation of which the eggs are capable. In view of the great variation 
which certainly does exist it seems not improbable that these forms are capable of 
normal development. 
6. In the cases so far described, the entire mass of the vitellus segments at the same 
time or nearly so. In several instances, however, segmentation began at one pole of 
the egg, leaving a large mass undivided at the opposite pole. These eggs had exactly 
the appearance of undergoing a partial segmentation, like that of Pyrosoma, or some 
Teleostean fishes. Thus in the egg shown by figs. 59 to 62 segmentation began with 
the formation of four small spheres at one pole of the egg, which then passed into a 
very marked resting stage. At the next cleavage (figs. 61, 62) the unsegmented 
portion broke up into about twelve spheres, of which two or three were somewhat 
larger than the others (fig. 62). The egg is now closely similar to that shown in fig. 38 
which was directly derived from the unsegmented egg, and its subsequent development 
calls for no remark. Figs. 63 to 67 represent a similar case. In this instance the 
MDCCCLXXXIII. 5 B 
