eand: nervous system of lumbricid^. 143 



some of the cell, and that it is not an oi'gan sui generis, like the 

 nucleus. Centrosomes and microsomes, he believes, are mechanical cen- 

 tres. A centrosome may arise at any point when needed, by the aggre- 

 gation of microsomes. Similar views were expressed by Watase ('93). 

 Mead ('98) shows that in the maturation of the egg of Chsetopterus a 

 large number of asters arise in the cytoplasm. They are finally reduced 

 to two, which become the asters of the maturation spindle. 



The presence of the secondary and tertiary centres in the nerve cell 

 is not inconsistent with the belief that the primary centrosome is the 

 centrosome left from the last mitosis. The continuity of the centro- 

 some of mitosis with that of the resting cells has been established in 

 many cells other than nerve cells. That a centrosome may in some 

 cases arise de novo must be granted in view of the results obtained by 

 Mead in Chsetopterus. Where the centrosome is continuous it may be 

 regarded as a permanently differentiated microsome of special functional 

 importance. After a mitosis it pei'sists in the resting cell and marks the 

 starting-point of the centred system of the mature cell. As the cyto- 

 plasm of tlie growing cell increases, the primary centrosome becomes 

 connected, by means of fibrils, with other granules of the new cyto- 

 plasm. With continued growth of the cytoplasm, the radiating network 

 develops from the centre outwards, being composed of " centrosomes " 

 and radiating fibrils of lower and lower degrees of importance according 

 to their distance from the primary centre. This is substantially the 

 centred system described by Heidenhain. 



That the function of such a system is a mechanical one appears most 

 probable. The centrosome is generally believed to possess an important 

 role in connection with the motor activities of cells, whether of the cell 

 as a whole (leucocytes), or of appendages of the cell (cilia ; flagellum 

 of the spermatozoon), or in the movements observed in mitosis. In the 

 absence of known motor activities in mature nerve cells, the most 

 likely function for the centred system is that of mechanical support. 



I have found no evidence whatever that the primary centrosome of 

 the nerve cell ever resumes its mitotic functions. In all of the regen- 

 erating worms examined, no typical nerve cells with processes were 

 found showing any change that would suggest the possibility of their 

 dividing. 



As already mentioned (pp. 78-80), masses of dividing cells are found, 

 occasionally in the old ganglia of a I'egenerating worm, but no such cell 

 proliferation is to be found in the cord of a normal worm. These cell 

 masses probably arise from " indifferent " cells which, in the early 



