THE STRUCTURE AND AFFINITIES OF HERDMANIA CLAVIFORMIS. 253 



not at first displace all the cytoplasm. Frequently from their regular arrangement 

 around the nucleus these give the cell a rosette-like appearance (Figs. 176 and 17c). 

 As they appear before the cell has become enlarged at all, and are of considerable 

 size, usually only a little smaller than the nucleus, they form as a rule at this stage 

 a single layer around the nucleus. From this condition the cells increase in size, 

 the increase being associated with the multiplication of the clear bodies. The stain- 

 able portion of the 'cytoplasm becomes less and less until finally it disappears entirely 

 and the cells, two or three times as large as at the beginning, are wholly made up of 

 the nucleus which has not yet changed its size or character, and the clear bodies (Figs. 

 17e and 17/). The cell now breaks easily and the clear bodies are frequently found 

 scattered about in the body space. The freed bodies or granules are spherical in 

 form, are moderately refractile, and have a slight yellow tinge. It frequently happens 

 that the cells filled with granules appear to possess two nuclei, one situated in the 

 interior, the other upon or very near the surface (Fig. 17/). The one situated within 

 is the nucleus proper to the cell. It is the one which undergoes the changes above 

 described. The nature of the superficial one is in some doubt. Almost certainly, 

 however, it belongs to some tissue, probably in the nature of connective tissue, that 

 is present in small quantity within the granule-producing cell-masses. These nuclei 

 persist after the granule-cells and their nuclei have disappeared. 



I have mentioned above that some writers have regarded these granules as 

 fat. They are not of this nature in Herdmania. In the first place they are 

 more definite and persistent in form than one would expect to find in fat granules. 

 But direct test fails to prove them to be of this nature; for example, they are not 

 affected by ether, neither are they colored in the least by Sudan III, nor by osmic 

 acid, at least in any kind of preserved material tried by me. My belief is that they 

 are more in the nature of yolk; this belief rests, however, entirely on their micro- 

 scopic appearance. I have no experimental proof to this end. The fate of the gran- 

 ules I have not followed. We must suppose, however, that they are used up as food 

 by the permanent tissues, probably of both parent and growing embryo. 



The fate of the nuclei of the granule-producing cells is noteworthy. By the 

 time the granules are mostly scattered or absorbed the nucleus has fully doubled its 

 size; and the chromatin elements, which at the outset were distinct, have entirely 

 disappeared, the nuclear substance having become homogeneous. The original 

 globular form of the nucleus is lost, an irregular more or less angular outline being 

 assumed. It still stains distinctly, though less intensely than does the chromatin 

 of the earlier stages. In fact the nuclei come to have much the same size, form, and 

 staining reaction as does the cell-body at the stage immediately preceding the ap- 



