No. 401.] DEVELOPMENT OF PENNARIA TIARELLA. 397 
occasionally incipient aspects of it were detected. This may 
be due in part to the membraneless condition of the eggs, but 
perhaps rather to their more or less evident amoeboid condition, 
which asserted itself in the earlier phases. Again, the eggs 
early became somewhat flattened and disk-shaped, and only as 
the external phenomena of cleavage were completed did they 
resume an approximately spherical form. In Pl. II, Figs. 1-6, 
are shown a single series of cleavage phases noted at intervals 
of ten minutes or less. All were sketched from life by the aid 
of a camera, and, as will be noted also, there are presented only 
the earlier phenomena. As segmentation progressed its super- 
ficial aspects became less and less evident, owing to the opacity 
of the eggs, and only in a few cases were attempts definitely 
made to follow it to anything like completion. In these and 
other figures will readily be recognized the amceboid aspects 
referred to above. 
Reference has been made incidentally to the variable rate of 
segmentation. This feature was quite as marked as were 
others, both as to individual eggs and blastomeres. In many 
cases a single blastomere would divide at a rate quite phenom- 
enal, so that it was difficult to sketch adequately successive 
phases, while others might remain in a state of inaction for an 
indefinite period or even be engulfed bodily into the more rap- 
idly segmenting portion. Then, also, in many cases cleavage 
seemed to begin in a somewhat discoidal fashion at a single 
pole and only gradually extend to other portions, as shown in 
Pl. IV, Fig. 1. This could hardly be due to any marked 
inequality in the distribution of the food yolk, for in this respect 
the eggs of Pennaria seem to be quite isotropous, and indeed 
there seems little if any definite polarity of any sort evident in 
these eggs. Occasionally, as shown in Pl. IV, Fig. 1, cleavage 
appeared to advance from a single area, but in the large pro- 
portion nothing of the sort was evident, and, as Conklin (97) 
has shown in molluscan eggs, where the greatest variation in 
matter of yolk distribution is distinguishable, there is none the 
less a perfect symmetry of rate and character of cleavage. This 
would seem to be only another illustration of the inadequacy of 
any known law as an explanation of all cleavage phenomena. 
