STUDIES IN 'I'HE KETINA. 355 



of contracting, and as its filaments are continuous with those 

 connecting the nodes together, we are safe in assuming that 

 the protomitomic fihiments are contractile. 



3. As the rods are protruded from the retina and the 

 filaments are pushed out along them, no matter how long the 

 rods are, the filaments are capable of growth. 



4. The apparent openings out of the nodal reticulum to 

 form the extra-nuclear networks suggest the possibility of a 

 somewhat free slipping of the junctions of the protomitomic 

 filaments. And this has also to be assumed in order to 

 account for the migration of individual nuclei, for the fila- 

 ments which function as the paths for the nerve-stinnili must 

 be thought of as permanently continuous. 



These seem to be the chief functional attributes of the 

 protomitomic filaments ; they are capable of conducting 

 nerve- stimuli, they are contractile, they can grow, and they 

 can move freely upon one another. But it is probable that, 

 by themselves, they can do none of these things, and that the 

 vital phenomena we have mentioned as apparently attributable 

 to the filaments of the protomitomic system, are only so 

 when it is associated with the other principal constituents of 

 protoplasm. These are four. 



Probably the most important, and certainly the most inti- 

 mately associated with the fibrils, is the chromatin. This is 

 found either massed in the close nodal or nuclear networks, or 

 streaming along the filaments from node to node. In no case 

 does it appear to leave the filaments. 



The next is apparently the clear fluid matter which is 

 closely, associated with the chromatin, and makes the nuclei 

 turgid vesicles. 



We then have the granular cytoplasm, which tends to 

 aggregate round the nuclear nodes, probably as nutritive 

 material; and lastly, the clear glassy strands of the syncytial 

 framework, which seem to act as supports. The protomitomic 

 filaments eitlier pass through these mcndiranes, or may be 

 bent by them (see fig. 13 b), or they may run along in some 

 close association, and yet in such a way as not to hinder the 



