86 BULLETIN: MUSEUM OF COMrAEATIVE ZOOLOGY. 



I. Introduction. 



For several years after the description by van Beneden, in 1883, of 

 the mitotic polar structures to which he gave the names, " spheres 

 attractives " and " corpuscules polaires," such structures were observed 

 only in cells undergoing mitosis, and were believed to be visibly present 

 only during the mitotic process, although it was suggested by van Bene- 

 den himself that they nught be preformed in the cell, becoming evident 

 only during mitosis. It was therefore a matter of some interest when, 

 in 1891, Flemming described structures in certain resting cells of the 

 salamander similar to those found at the poles of the . mitotic figure. 

 Flemming was the first to describe these structures minutely, although 

 the observations of Solger ('89) and Eabl ('89) had been made previously. 



This discovery led to the search for the centrosome in the resting 

 cells of various tissues, and interest in the question culminated with 

 the announcement by von Lenhossek, in 1895, of the discovery of a cen- 

 trosome and sphere in the nerve cells of the frog. Von Lenhossek was 

 quickly followed by other authors who described similar structures in 

 the nerve cells of other animals, and at the present time the " centro- 

 some " is known to exist in the nerve cells of many vertebrates and 

 invertebrates. 



It is a matter of some importance, as has been pointed out by several 

 authors, to determine whether or not the " centrosomes," " spheres," 

 and radiations seen in the cytoplasm of fully differentiated and pre- 

 sumably functional nerve cells are related to the polar structures of 

 past or future mitotic divisions of the cell. Accepting the general belief 

 that functional nerve cells do not divide, we are limited to proving either 

 that the centrosome and its accompanying structures of the last mitosis 

 of the embryonic nerve cell persist as the centrosome and radiating sys- 

 tem of the differentiated nerve cell, or that, on the contrary, they do 

 not persist, and that we have to do in the differentiated nerve cell with 

 structures which have arisen in the cytoplasm independently of the last 

 mitosis of the cell. That certain parts of the radiating system of the 

 resting cell could have been left over from the last division, and other 

 parts not, is, of course, a possibility to be considered. 



The following work was undertaken at the suggestion of Professor 

 E. L. Mark, with the purpose of throwing some light upon the origin 

 of the so-called centrosome of the nerve cell. The work was carried on 



