102 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 



maiider) ; Gulland ('96 ; leucocytes) ; Spuler ('96 ; connective-tissue 

 cells) ; Child ('97 ; ovarian stroma cells of mammals) ; Heidenhain 

 ('97 ; red blood corpuscles of duck embryo) ; Eawitz ('98 ; sperm 

 cells of Scyllium canicula) ; Niessing ('99; lymphocytes). 



Reviews of the centrosome question, with lists of literature, have been 

 given from time to time by various authors, among them Flemming 

 ('93), Moore ('93), Hacker ('94), Heidenhain ('94), vom Rath ('95), 

 Henneguy ('96), von Erlanger ('97), Heidenhain ('97), Kostanecki 

 und Siedlecki ('97), Btihler ('98; centrosome in nerve cells), and 

 Meves ('98). 



III. Methods of Investigation. 



The identity of the so-called centrosome of nerve cells with the cen- 

 trosome of mitosis cau be established by a series of observations showing 

 that the centrosome of the last mitosis of the embryonic nerve cell 

 persists through the development of the nerve cell and becomes the 

 structure seen in tke resting cell ; or, if the cells can be induced to 

 divide by mitosis, it may be possible to show that the centrosome of 

 the resting cell gives rise to the centrosomes of the mitotic figure. The 

 problem, then, is open to attack from two directions. Given, material in 

 which the presence of a centrosome in the differentiated nerve cells can 

 be demonstrated, — as it has been in many vertebrates and inverte- 

 brates, -— and in which the centrosome is also to be found in the embry- 

 onic neural tissues, — as Heidenhain und Cohn ('97) have found it iu 

 the neural epithelium of bird embryos, — then we may seek to establish 

 the genetic relationship of the organ in the embryonic cell with that of 

 the differentiated cell. 



A second method of investigation is offered by the process of regen- 

 eration. If any cells of adult nervous structures are capable of under- 

 going mitotic division, regeneration would seem to oflFer the most 

 favorable opportunity for the exercise of such power. If, however, 

 the regenerated nerve cells do not arise from old nervous tissues, the 

 development of the new nerve cells from other tissues would still offer 

 opportunities for determining whether the " centrosome," if present in 

 the regenerated cells, had persisted from the last mitosis in the history 

 of the cell. 



The question arises as to whether processes which take place in the 

 embryonic origin of nerve cells will be repeated in the regenerative 

 development of similar cells. To show that a centrosome in a regen- 



