The early development of Eiidendrium. 265 



devoid of all trace of nucleolar substance with but slight trace of 

 chromatin granules. Very soon after this phase the nucleus becomes more 

 and more transparent, completely losing its chroraophilous properties, 

 and only rarely being distinguishable in even the thinnest sections. 

 As intimated above no trace of polar bodies has been found, and 

 I am constrained to the opinion that following their forma'tion soon 

 after the dissolution of the nuclear membrane they are probably re- 

 sorbed by the cytoplasm, as Harm (1902), has shown in the case of 

 Clava, and as in other well known cases. 



Early Development. 



Observations made upon living material of several species, more 

 particulary jE. ramosum and E. capillare, showed phases of cleavage 

 comparable with Figs. 8 — 13, drawn under the camera from life of 

 E. ramosum. It was indeed possible to follow in some cases the 

 progress of the cleavage, tho not with any considerable detail owing 

 to the opacity of the eggs and the difficulty of orientation, since not 

 only the individual gonophore but a considerable portion of the hydroid 

 colony seemed necessary in order to maintain normal conditions even 

 for a short time. By a comparison of several specimens it was not 

 difficult however to recognize that in surface aspects they bore every 

 semblance of ordinary cleavage phenomena. Upon fixation, however, 

 by whatever reagent employed every trace of cleavage disappeared, 

 and subsequent sectioning failed likewise to aiford any internal evi- 

 dences of definite blastomers or other cleavage characters. It was 

 only after repeated experimentation with every available reagent that I 

 was forced to the conclusion that in all probabilities the surface 

 phenomena were not true cytoplasmic, but yolk-cleavages, comparable 

 with those of Crustacea and Insecta. That they are not however 

 wholly independent of cytoplasmic, or nuclear activity seems more or 

 less evident since sections show the presence of nuclei in various 

 phases of activity, mitotic figures being demonstrable and resting 

 nuclei quite common. These were however, for the most part wholly 

 independent of the yolk, the latter being more or less evident in 

 early stages as pyramidal masses, somewhat similar to yolk-pyramids 

 in Crustacea, as shown in Fig. 15. Again the irregular distribution 

 of nuclei suggests conditions more or less common in insect cleavage. 

 It would seem therefore that we have in the heavily yolk-ladened 

 eggs of the Eudendridae cleavage features comparable in many respects 

 with those found among similar eggs in other groups, more evident 



