The Germ-cells of Lophius. 585 



frequent in cells with two small ones (figs. 22, 23, 26-28). It is 

 also important to note that it is not infrequently seen in the cyto- 

 plasm of cells with one large plasmosome and one small one (figs. 

 21 and 25). Such observations as these give good reason to be- 

 lieve that the body is actually an extruded plasmosome, because 

 at the time of its appearance the nuclear plasmosomes become 

 smaller. Fig. 26 shows a cell in which the process of extrusion is 

 apparently just completed. The extruded body lies close to the 

 nuclear membrane, and directly opposite it, just within the nu- 

 cleus, lies one of the small plasmosomes, as if the extranuclear part 

 had but recently been separated from it. Fig. 25 shows a case 

 somewhat similar, with this difference, that but one of the nuclear 

 plasmosomes is of small size, and the extra-nuclear one lies 

 opposite this. It is also important to notice that there was no 

 clear case observed of a cell which had two of these bodies in the 

 cytoplasm. 



From such conditions as the above it appears that during extru- 

 sion of material, each nuclear plasmosome moves to the nuclear 

 membrane and there gives out part of its substance into the cyto- 

 plasm. The fact that in later stages the plasmosomes of germ-cells 

 are in contact with the nuclear membrane gives support to this 

 view. It is also evident that the reduction of both plasmosomes 

 is not simultaneous, but that each gives out part of its substance 

 at periods sufficiently far apart to allow of the disappearance of 

 the first one before the extrusion of the second. The other pos- 

 sible mode' of extrusion is that within the nucleus there occurs a 

 division of each plasmosome followed by the extrusion of one frag- 

 ment of each from the nucleus. The only instance I have observed 

 which may possibly indicate such a division within the nucleus 

 is drawn in fig. 24. In this nucleus there is a third body which 

 may possibly be a fragment of a plasmosome about to be cast out 

 of the nucleus. This single case, however, is not clear enough to 

 be of great significance. 



A study of the distribution of the cases as shown in table 

 2, page 584, indicates that the process of extrusion is most active 

 at stage 7, because at this stage there are more cells with 

 an extruded plasmosome than at earlier or later stages. It is 



