396 THE AMERICAN NATURALIST.  [Vor. XXXIV. 
able. The highest powers of the microscope revealed nothing 
which seemed quite comparable with the so-called spinning 
movements of protoplasm, to which Andrews has directed 
attention. 
So anomalous are these phenomena that during the first sev- 
eral series of observations the conclusion was unavoidable that 
they were strangely abnormal or, perhaps, pathologic. Accord- 
ingly, the first series were wholly discarded, except a few speci- 
mens which had been isolated in watch glasses and set aside 
more out of curiosity than otherwise, and left overnight. An 
inspection the following morning revealed several apparently 
normal larve. The observations of the following night were 
to the same effect, though at this time a considerable series of 
all sorts were isolated with pains to eliminate possible error, 
and with results quite assuring, in that in a very large propor- 
tion of the cases perfect larve resulted and continued to de- 
velop. Following this, systematic collections were made and 
painstaking observations and records kept of every feature 
associated with development. 
Occasionally an egg would segment with a fair degree of 
regularity into the two, four, and eight cell stage, as shown in 
Pl. II, Figs. 7, 8. But beyond this point it was difficult to fol- 
low any order in the cleavage, though it might continue with 
more than ordinary regularity as compared with the average of 
its fellows, It was utterly impossible to trace anything like a 
definite lineage of cells, notwithstanding repeated and careful 
attempts. By no means was it possible to predict the direction 
or course or rate of any division beyond the first or second 
phase of the cleavage, and even then only occasionally. 
While Wilson (84) has noted a considerable degree of indi- 
vidual variation in the cleavage of Renilla, he is still able to 
reduce it to some half dozen types. With Pennaria, however, - 
while it is possible to recognize some few rather predominating 
types in the earliest cleavage, as, for example, a centripetal, a 
centrifugal, and a vertical, yet they are not of. sufficiently pro- 
nounced character to constitute well-defined types which are 
distinguishable as such. In no case were there any clearly 
defined and symmetrical equatorial phases recognizable, though 
