254 BULLETIN OF THE 
formation of the first cleavage, when we have an 8-cell stage. From 
this we pass into morula stages, in which additional cleavage planes 
were not successfully traced as they originate, and in which the primary 
and secondary planes could not be recognized as such. 
Before leaving the stage (fig. 7) in which the egg was found at 3h. 
45m. p.M., let me mention an appearance in the egg which was not un- 
derstood, but which may have a significance in the embryology of these 
animals. At the point in the egg adjacent to the break which has taken 
place in the direction of the primary furrow, a depression is formed 
which resembles an opening leading into the interior of the ovum. 
From the arrangement of the cell walls in the immediate vicinity, it 
seemed as if this opening was formed by the drawing apart of the walls 
of the cells, but whether it is the result of decay or not cannot be at 
present stated. The single egg in which it was observed, however, 
afterwards died before passing into advanced larval conditions. 
It is at about this time in the development of the Agalma egg that 
some of the most extraordinary examples of protoplasmic elevation from 
its surface were observed. The resulting changes in external form often 
baffle all attempts to observe accurately the normal outlines of the cells 
of the segmented egg. These rhizopodal prominences are most clearly 
marked in those eggs which have been in long captivity, and seem 
wholly different in different ova. 
Before closing our account of the segmentation, let us compare our 
observations with those of other naturalists on the same or closely allied 
genera. The poverty of our knowledge of the segmentation of the egg of 
the genus Agalma is so great, that I find few descriptions in the writings 
of others available for comparisons. Metschnikoff, although not figur- 
ing the segmentation of the egg, evidently observed it, as the following 
mention indicates. He says,* “Die Dotter zerkliiftung, resp. Larven- 
bildung findet auf dieselbe Weise statt, wie ich oben fiir Hpibulia au- 
rantiaca angedeutet habe und wie sie bei allen von mir beobachteten 
Siphonophoren als Regel gilt. Was aber die Vorgange der Organbild- 
ung betrifft,” he continues, “so finde ich die meiste Analogie mit den 
von Haeckel untersuchten Crystallodes rigidum und Athorybia rosacea, 
obwohl auch in dieser Beziehung Agalma Sarsii manches Eigenthim- 
liche darbietet.” Turning for further information to his account of the 
segmentation in Epibulia we find him devoting a few significant para- 
graphs to this interesting process. He says,f “‘ Die bald auf das freie 
Ablegen (es gelang mir nie kiinstlich aus dem Schlauche befreite Eier 
* Op. cit., p. 49. + Op. cit., pp. 40, 41. 
