THE CRASPEDOTE MEDUSA OLINDIAS AND SOME OF ITS NATURAL ALLIES. 9 



consists exclusively of ectoderm cells, since the boundary line between the two cell- 

 layers is always distinguishable with a good objective (Zeiss apochromatic oil-immer- 

 sion), and there is no sign of proliferation in the endoderm. On the contrary, the 

 latter is usually found pushed in by the otocyst rudiment just the contrary of what 

 one would expect if the endoderm took part in its formation. The first rudiment 

 just described forms, as will appear in the sequel, the otolith and the parts that di- 

 rectly surround it. 



In the next stage (Fig. 16) the endoderm is generally found pushed in somewhat 

 more by the otocyst rudiment, and around the latter on the outer side has been formed 

 a cup-shaped cavity lined by a layer of cells, some of which are flattened, while 

 others can hardly be distinguished from the surrounding ectoderm cells. This cavity 

 is the beginning of the vesicle, and the layer of cells around it becomes its lining epi- 

 thelium. In this stage the cells that formed the first rudiment are all alike. 



The next change that takes place consists in the enlargement of the cavity and 

 the differentiation of the central cells into two sorts, namely, those that are trans- 

 formed into the otolith and those that form its investment and stalk. In the section 

 reproduced in Figure 18 (PL III) this differentiation has hardly begun, and the deeply 

 and faintly staining nuclei are found mixed together, although some of the more cen- 

 tral ones are larger and clearer than others. It may also be noted that in this section 

 many of the peripheral nuclei are vesicular in appearance. These differences taken 

 by themselves appear to me to afford only a dubitable clue to the destination of the 

 respective cells, depending as they do on the changing conditions of the nuclear 

 substance. For example, even in one and the same otocyst some nuclei of its 

 lining epithelium are clear and vesicular, while others are darkly stained. In a later 

 stage the cells that are destined to be transformed into the otolith and those that 

 remain as its investment and the stalk are clearly distinguishable, the central nuclei 

 being generally larger, clear and vesicular, while the peripheral ones are deeply stained 

 and are more or less flattened. It must be borne in mind, nevertheless, that these 

 differences are by no means absolute. Sometimes, as is seen in Figure 17, the dif- 

 ferentiation of the central cell mass appears to be effected at a very early stage, and 

 in such a case the difference between the two kinds of cells is very striking. 



In the last stage described the central cell mass was sessile, resting as it did directly 

 against the endoderm of the circular canal by a broad base. If double staining is 

 resorted to at such a stage, the cells that are destined to become the otolith stain of 

 the same color as the endoderm, while the lining epithelium and the investing cells 

 stain like the ectoderm; for example, with fuchsin and methyl green the endoderm 

 and otolith cells are violet, and the epithelium, investing cells, and ectoderm are green. 



