No. 2.] THE ORIGIN OF THE SEX-CELLS. 20% 
ABNORMALITIES. 
In the cleavage of the egg many peculiar forms of segmen- 
tation appeared, but it was impossible to demonstrate that they 
all formed planulae. Those that were followed did not; yet 
as very frequently even those having normal cleavage died 
while under observation, we cannot conclude that under no 
circumstances do these abnormal forms become hydroids. 
The most common abnormality is that arising at the first 
cleavage, in which the plane does not pass entirely through 
the egg, but buds off two, three or four small cells at the 
upper pole, leaving one large cell at the lower pole. Dr. 
E. B. Wilson mentions this form of cleavage in Renilla. 
Thus in enumerating the forms of cleavage found in this 
animal, he states “cleavage may begin at one pole of the egg 
with the formation of four or five small spheres, and usually 
after a quiescent period the remainder of the vitellus breaks 
up at once or progressively into spheres approximately the 
same size as those first formed, and the egg passes into the 
sixteen-sphere stage.’’ Again Brauer also notes for the Tubu- 
laria Mesembryantheum Allm.: “Im allgemeinen verlauft sie 
auf zwei verschiedene Weisen, entweder folgt jeder Kern- 
theilung auch eine Zelltheilung, oder es vermehren sich 
zunachst nur die Kerne und es beginnt dann eine allmah- 
liche Abfurchung, welche am Richtungskorperpole anfangt und 
dann nach der entgegengesetzten Seite fortschreitet.” I have 
once thought I detected the large cell breaking up into small 
cells, but at other times I have watched this form for three and 
four hours without any change taking place in it. 
In starting to form the four-celled stage the blastomeres 
frequently undergo a movement, at times taking a position in 
which instead of lying opposite one another they lie beside 
each other, their furrows apparently cutting in from above as 
in the first cleavage (Pl. II, Fig. 33). In this case by further 
segmentation a chain of loosely detached cells is formed 
(Figs. 34, 35). I did not, however, trace the development 
beyond this stage. Metschnikoff has found this plan of 
