STUDIKS IN THE RETINA. 353 



Any freedom of movement at all must involve a good deal of 

 slipping of the individual filaments, for each node is itself a 

 complicated reticulum. At one time it seemed possible to 

 find some clue to the mechanism of the movement in the 

 extra-nuclear networks, in which there appears to be a 

 temporary dissolution or loosening of the intra-nuclear net- 

 works. But although these extra-nuclear networks appeared 

 in the migrating nuclei shown in Part II, PI. 31, fig's. 26 and 

 27, in most cases the nuclei can be seen passing through the 

 outer reticular layer with clean, hard outlines, and sometimes 

 with clearly defined amoeboid changes of shape. The most 

 remarkable cases of this were seen in the large retinal nuclei 

 of Ceratodus and Protopterus (see fig. 28), but it can frequently 

 be seen in nuclei passing through the outer reticular layer, e. g. 

 in the cat many of them were shaped like large-headed pins 

 (see figs. 20 h and 25) . I do not now think the two phenomena 

 have much more to do with one another than that both are 

 dependent upon a great power of slipping on the part of the 

 protomitomic filaments. In fact, it looks as if the reticulum 

 is the very reverse of rigid ; that, in spite of its complicated 

 nodes, in life its filaments move freely upon one another. 



One more problem relating to the retinal nuclei must claim 

 a moment's attention. What is the origin and the function of 

 the numbers of minute nuclei found, especially in young 

 Mammalian retinas, in the ganglionic cell-layer, and mostly 

 on its inner side ? See, for instance, fig. 25, which shows them 

 in a young cat. Others are shown in figs. 2 b, 3 c, d, and 26 

 (/. They are mostly very minute. If they happen to be large 

 they are usually very much indented, as if they were 

 fragmenting (see figs. 26 g, and 3 c). They are almost devoid 

 of chromatin contents, and frequently look like clear empty 

 vesicles. A comparative study Avhich might throw light upon 

 them must be a work of the future. I have so far only 

 endeavoured to see if they were linked on to the retinal 

 protomitomic system, but without arriving at any positive 

 result. The few filaments I succeeded in seeing coming from 

 them were always lost sight of in the surrounding tissue. If 



VOL. 47, I'AKT 3. NKW SKKIES. Z 



