260 PROCEEDINGS OF THE AMERICAN ACADEMY. 



nerve cord without indicating metamerism or symmetry. In the sub- 

 a3sophageal ganglion, however, they show an approach to symmetry. 



4. Confirmation of the conclusion that the substance of Leydig's 

 fibre is uniform throughout and does not represent a bundle of nerve 

 fibrilloe ; further, that the fibre does not show any indication of being a 

 degenerate structure. 



5. In their relation to ganglion cells, Leydig's fibres, if real nerve 

 fibres, as maintained, stand in strong opposition to the neuron theory of 

 Waldeyer, von Lenhossek, Edinger, and other writers. The relation 

 of the process of one giant cell with that of another appears to be one of 

 direct continuity, not simply of contiguity. 



6. The cells which give rise to Leydig's fibres show peculiar structural 

 conditions in the possession of a nucleus always eccentric in position and 

 in the possession of other structures more central in position, the centro- 

 some and sphere. 



7. The presence of a centrosome in these ganglion cells argues for a 

 function of the centrosome other than that of an organ for cell division.* 

 No division of nerve cells has been observed in this worm, and, besides, 

 cell division is generally believed never to occur in case of normal 

 ganglion cells. 



8. Although not an orsran of cell division, the centrosome in these 

 ganglion cells is evidently a centre of mechanical activity. This is 

 indicated, (a) by the eccentric position of the nucleus ; {V) by the 

 flattening or indentation of the nuclear membrane, frequently observed 

 on the side toward the centrosome ; (c) by the concentric arrangement of 

 the pi'otoplasm around the centrosome ; and (c?) by the radiations which 

 extend from the centrosome. 



9. The presence of a centrosome in ganglionic cells is an argument in 

 favor of its being a permanent cell organ. 



10. In some of the giant cells two centrosomes and spheres are pres- 

 ent, without there being any evidence of approaching nuclear division. 



IL Parts of the peripheral nervous systems of the two Maldanida?, 

 Axiothea torquata and Clymene producta, terminate in multicellular 



[* The identity of the centrosome here described with that which usually 

 accompanies cell division is liere assumed ; hut it lias not been proved. An im- 

 portant problem for tlie future is to determine if a genetic connection can be 

 established between tliis so called centrosome of nerve (and other) cells and the 

 centrosome which exists during cell division. — E. L. Mark.] 



