PLATE XXXIV. 



Figs. 5-6. Ctenolabrus 64-cell stage. Fig. 5 from above, Fig. 6'?rom below. Both are in about 



(21-22) the same state of division. In [the preparation used for] Fig. 5 the central cells are stained 

 browner (osmic) than the marginal cells, and their distinction was only somewhat less noticeable 

 in (the view shown in] Fig. 6. For this reason and for convenience of analysis, I have given 

 heavy outlines to the central cells, and dotted outlines to the cells seen through the exposed 

 surface. 



I see no reason to doubt that the lettering in Fig. 6 is precisely as it should be in relation to 

 the lettering of Fig. 4 of the previous stage. In Fig. 5 there is no obscurity except with regard 

 to which ones of the marginal cells represent pc and re. However as pc of the left side is cer- 

 tainly identified, it seems probable that (the numeral] 5 (between 2 and 3 of the anterior mar- 

 ginal cells], represents pc of the right side. If these two are correctly identified then there can 

 be but httle doubt in regard to rc-^c. 



In both figures, assuming the division completed, we have thirty-six cells in the margin and 

 twenty-eight in the center. Of the thirty-six cells, eighteen he above (obliquely) the other 

 eighteen, the eighteen of the upper surface, lying for the most part nearer the center, while the 

 eighteen of the lower layer stretch away farthest from the center, extending out so as to present 

 one half or more of the upper siu'face to view. Although Fig. 6 exhibits considerable variations 

 from the radial direction of the nuclear figures which prevails in Fig. 5, still the direction is such 

 that in the majority of cases the inner pole of the figure lies nearest the upper surface, so that 

 the division is not horizontal, but oblique. The pole nearest the observer is more strongly 

 shaded in both cases. 



It is very interesting to note that the outer poles in many cases lie extremely near the margin, 

 and, particularly in Fig. 6,-one can see that the marginal daughter nuclei lie very near the under 

 surface — yolk surface — of the cap. Also it is interesting to note that the margmal cells often 

 pass into the pellicle with diffuse or blended outline. 



May not the marginal cells already be considered as endoderm and the upper marginal as 

 mesoderm? It is quite certain that all the central cells represent ectoderm. 



It is important to note that in all these stages, the two intermediate cells (i-i) maintain 

 uniform relations and always divide horizontally in passing to 64-cell stage. This confirms 

 the view that the embryonic axis coincides with first plane of cleavage. 



.\gain a portion of re {5 in Fig. 6) hes in the floor while the marginal portion lies mainly above 

 the cells 3 and 4- The left cell re also divides obliquely, one half showing in the floor. In 

 Fig. 5 right re lies wholly in roof, left re Ues wholly in margin. 



In Fig. 5 left pc touches the floor at its right end, while right pc lies in margin. In Fig. 6 

 right pc lies mainly above 2 and 3 and left pc divides obliquely, one half showing in the floor. 



The nuclei when seen frontally , look like a line of large elongated granules. Viewed from the 

 pole they generally present the form of a ring of granules. The achromatic poles are sometimes 

 very plain. Some nuclear rays reach beyond the outline of marginal cells, into the pellicle. 

 |Fig. 6, no. 1 in anterior row of cells]. 



Vacuoles are seen only in the outlines of the roof-cells — nowhere in the floor-cells; they are 

 very neat in both preparations but have been omitted from the figures. 



The cleavage-cavity cannot be traced with certainty; possibly it no longer exists. 



In a 64-cell stage of same set, but 30 min. later, I find two of the margmal cells, at the fore 

 or hind end of the future axis that have taken on the form and appearance of true endoderm 

 cells. 



Most of the preparations on the slide used for Figs. 5 and 6 are in the 32-cell stage. One of 

 these 32-cell stages had a remarkable sjinmetry. 



Th'ere were sixteen cells in the periphery and sixteen in the center. The retro-central and 

 pre-central cells did not lap but stood abreast, like the pairs of intermediate cells, reaching from 

 top to bottom of cap. Thus the cap was two cells deep only in the four central cells. 



Endoderm: — .\lthough I believe the 64-cell stage practically settles the endoderm and that 

 some of the cells in this stage very often assume the endodermal condition, I think that they 

 resume their outlines and keep the marginal position in the cap for some time later. Two 

 hours later than Figs. 5 and 6, I find the endodermal wreath more or less conspicuous in all, 

 but in some the wreath has not yet included more than a part of the marginal cells. 



