22 George Thomas Hargitt 



abundant in the entire region of the future hydranth, as well as 

 on the stem just below the knob. They were not found in the 

 entoderm as in T. larynx. 



There is the same thickening of the ectoderm and entoderm 

 layers found in sections through very early stages, as in Eiiden- 

 driuiii rainosuin, and this condition is probably the result of cell 

 proliferation. A careful examination of sections shows mitosis 

 to be fairly abundant, mitotic figures being found in both ecto- 

 derm and entoderm (figs. 6i, 62). Therefore, in this form, as 

 well as in Eudendrium raniosuTn/ indications point to mitotic 

 division as an active process in the development of the regen- 

 erating -structures. However, indications of direct division were 

 not lacking. In fig. 63 is shown a nucleus in the process of 

 direct division, accompanied by division of the cytoplasm, a cell 

 wall being partly formed. In other sections nuclei were found 

 dividing directly, but cytoplasm seemed to have taken no part 

 in the process (fig. 64). Furthermore, amitotic division can 

 not be considered merely as a somewhat abnormal condition 

 brought about by artificial conditions. The same thing is found 

 in sections made through the young hydranths developing nor- 

 mally on colonies growing in their natural habitat. Fig. 65 

 shows a considerable number of nuclei in different stages of 

 direct division, and this is in the region of the developing ten- 

 tacle. Such dividing nuclei were limited to the ectoderm. 



Only a few sections through developing tentacles were ob- 

 tained. Fig. 66 shows the appearance of the cells in one case. 

 Since the core of the mature tentacle is made up of a single row 

 of entoderm cells, as in Endendrimn, the same process of devel- 

 opment would be expected. Whether this is the case could not 

 be definitely determined owing to the lack of material of the 

 later stageS; 



THE OCCURRENCE OF AMITOSIS 



Other writers have referred to the occurrence of mitosis in both 

 the ectoderm and entoderm of regenerating hydroids. Therefore 

 on starting the histologic work, wdiich has been described above, 

 I expected to find approximately the same conditions. When 

 traces of mitosis were not found, after staining particularly for 



296 



