10 J. GRAHAM KEUR. 



As already mentioned, the yolk in the microineres is reduced 

 to the condition of fine granules. These also become reduced 

 in number, and the nucleus tends to be surrounded by an 

 area of finely granular reticular protoplasm, almost free 

 from yolk granules. The transition from the finely granular 

 micromeres to the coarsely yolked macromeres is perfectly 

 gradual. 



Between Stages 8 and 9 there appears an irregular chink of 

 larger size than the others amongst the lower micromeres 

 (PI. 1, fig. 2) . This, the definite segmentation cavity, increases 

 in size, spreading laterally, and at the same time approach- 

 ing close to the upper surface of the egg, being eventually 

 covered in by a roof of comparatively regular thickness 

 throughout. This roof soon becomes composed of two regular 

 layers of cells (figs. 3 and 4). As the segmentation cavity 

 further increases in size these become flattened out, until 

 the roof forms a thin translucent membrane through which 

 in the entire egg the segmentation cavity appears as a dark 

 shadow. The characters of the completed blastula may be 

 sufficiently gathered from fig. 4. 



The blastomeres on the floor and sides of the segmentation 

 cavity are rounded, almost spherical in form, and project into 

 the cavity. Usually, some of these spherical blastomeres 

 appear to float quite free in the fluid of the segmentation 

 cavity. This appearance does not of course prove that they 

 are not really connected up to the other blastomeres by 

 delicate protoplasmic strands; but such connecting threads 

 if present are too delicate to be seen by ordinary observation. 



Grastrulation. — The process of gastrulation in Lepido- 

 siren may for convenience of description be divided into 

 three periods, which I will call a, b, and c. 



A. In this period, which marks the beginning of gastrula- 

 tion, we have to do with a process of true invagination. The 

 commencement of this process is indicated, as I have shown 

 in my previous paper, by the appearance of a row of little 

 depressions of the egg's surface arranged in a latitudinal 

 direction a few degrees below the ec[uator. These depres- 



