.280 Transactions of ths Society. 



l3roken up, and in such cases it is difficult to identify the remains 

 ■as being a nucleus. From a study of a number of oogonia at this 

 stage it seems as if the nucleus becomes blown up with its con- 

 tained nucleoli, suddenly bursts, and shoots its granular contents 

 into the surrounding cytoplasm. One is impressed at all events 

 by the collapsed and changed condition of the nucleus after it has 

 parted with most of its plasmosome material. 



In many cases it becomes possible to identify a thick halo 

 of differentiated cytoplasm surrounding the nucleus, as shown 

 in PL V, figs. 6 and 10, at H. One is irresistibly induced to 

 believe that this halo is formed by materials squeezed, or at all 

 events passively passing, out of the nucleus. In certain examples, 

 as in PI. Y, fig. 6, the peri-nuclear halo contains vacuoles and 

 granules, and is arranged in a radiating manner from the nuclear 

 membrane. 



Using Benda's alizarin and crystal violet stain, or even acid 

 fuchsin and toluidin blue, it has been possible to sliow that the 

 nucleolar or plastin material undergoes a change when extruded 

 into the cytoplasm, or when it comes into contact with the cyto- 

 plasm. This is shown by such stages as in PL V, figs. 9 and 10; 

 in each case the granules (G or N) after passing into the cytoplasm 

 stain more densely in crystal violet than before. In fig. 10 the two 

 plastin bodies at? X stain more heavily than their parent body in 

 the centre. 



After the extrusion of the plastin materials the nucleus reforms 

 as in PL V, fig. 9. Occasionally some of the extruded plasmo- 

 somes stain less heavily than tlieir fellows, as at X. The nature 

 of this peculiarity is not known. 



That the extruded plasmosomes form the mitochondria of the 

 ripe egg seems to me to be an attractive view. A number of 

 authors (^, 7) have identified these cytoplasmic granules as chro- 

 midia and have traced their origin to the nucleus. Subsequently 

 the extruded plasmosomes break up (divide ?) into smaller pieces, 

 and appear to grow and continue active within the cytoplasm. 



Quite lately I have found that in the oogenesis of Saccocirrns 

 the extrusion of nucleolar materials and the formation and presence 

 of mitochondria in the cell are not related. As is well known, 

 Hempelman and Buchner both describe in Saccocirrus the extru- 

 sion of nucleolar (plasmosome) fragments which later form " yolk." 

 Now these nucleolar fragments in Saceocirnis simulate the staining, 

 if not the fixing reactions of the mitochondria, and for a time I 

 thought that the cases of the sponge and of this nrchiannelid might 

 be similar. This seems to be far from being true ; in Saccocirrus 

 true mitocliondria exist in the (ig^ before the extrusion of the 

 nucleolar fragments. In my last paper on sponge embryology {3) 

 I came to tke conclusion that true mitochondria did exist in Grantia 

 collar cells, and that the so-called "chromidia" in the egg cytoplasm 



