504 CHARLES W. HARGITT 
attention only to eggs which exhibited the regulation aspects 
of cleavage, and disregarding, as abnormal, those of differing 
aspects, Just such an account might easily have been made up; 
and this in all probability may have been the method followed. 
It is not strange that under prevailing conceptions as to for- 
mulated ‘laws of cleavage’ this method might naturally have 
been adopted. In the case of Pennaria the present writer delib- 
erately disregarded an entire batch of eggs which were so erratic 
in behavior as to suggest the probability of pathological condi- 
tions. But, by whatever method one may explain the matter, 
certain it is that there is a measure of irregularity in a large pro- 
portion of the eggs of Hydractinia, especially after the third or 
fourth cleavage furrows, which at once takes them out of the usual 
category of geometrical order or symmetry and puts them, if not 
in the Pennaria class of chaotic irregularity, at least consigns 
them to the category of the indeterminate and unsymmetrical. 
However, it is not my purpose, in thus discrediting an account 
which gives so inadequate and misleading an impression, to goto a 
similar extreme in the other direction and convey the impression 
of predominantly erratic cleavage. On the contrary, let it be 
noted that in perhaps a majority of the eggs of Hydractinia echi- 
nata the cleavage, while seldom exhibiting an approach to geo- 
metric order or symmetry, is yet more or less regular and orderly. 
In such cases cleavage begins, as usual, at the animal pole, cut- 
ting vertically downward, and generally divides the egg into sym- 
metrical halves, which adhere to each other by a narrow band, or 
connective of cytoplasm at the lower pole. The second cleavage 
likewise may begin at the upper pole and at right angles to the 
previous division, or may begin at the center and work outward, 
thus dividing each half into symmetrical fourths, giving a fairly 
tvpical four-cell stage. The third cleavage, which is usually equa- 
torial, often begins at the center and extends toward the periph- 
ery, a process more or less common in eggs of hydroids. The 
subsequent phases may continue more or less orderly as in earlier 
stages, but often grow increasingly irregular and independ- 
ent, though resulting in a symmetrical embryo. On the other 
hand, figs. 14 to 22, which are camera sketches of living eggs, 
