116 BULLETIN': MUSEUM OF COMPAEATIYE ZOOLOGY. 



The results of investigations upon other groups of animals are t : : « e U 

 known from recent reviews to make it necessary to repeal them here. 

 It should, however, be s: :iiat Suzuki '98 in elasmobranchs and 



von Kortf ("99 i in Helix have shown that the proximal portion of the 

 axial filament arises merely as an elongation of one of the centrosoiL-;. 

 But in the vast majority of animals this has been proved to be nc : true. 



1'. Centrosomes. 



The behavior of the centrosomes during the spermatid transformations 

 is so very different in different animals that I will mention in this con- 

 nection only observations made upon chilopods. 



. >Jophilus the observations of Collin (:Ol) and of P. et It Bouin 

 ■ 03) working in the same laboratory are quite a: variance concerning 

 the behavior of the centrosomes. Collin sees the centrosome first upon 

 nuclear membrane, and from this the axial filament runs out. A: the 

 close of the --.- :ond spermatocyte division in Geophilus linearis, according 

 * he observations of P. fc! M. Bouin, the centrosome lies upon the inner 

 face of the cell membrane. The "centrosome" contains two closely 

 united " centrioles," which are oriented perpendicularly : - 

 brane. The li centrosome " is a mass of cyl .. : substance around 



the '• centricles. 1 ' The nucleus moves toward the "centrosome" and 

 comes into contact with it. " On voit alors apparai::- but le centriole 

 le plus externe ou distal une tres mince expansion. elle-ci a 'e rnd 



entre le centriole distal et la membrane cellulaire." The nucleus and 

 centrioles now move toward the centre of the cell. " Ce premier rudi- 

 ment du filament axile ne provient done pas de l'aliongement du 

 centriole distal puisque ce dernier se distingue toujours avec son con- 

 genere, au niveau de son extremite.'' When the axial filament has 

 ined a considerable size and the cell has begun to lengthen, '• Le 

 centriole proximal penetre dans le noyau et fait saillie a la surface interne 

 de la membrane nucleaire. Puis il grossit pen a pen et prend une taille 

 gigantesque." It remains connected with the external " centriole ' 

 means of " un pedicule plus ou moins long et grele." T:.e nucleolus 

 sometimes applies itself to the mass, which it seems to augment, but 

 it never fuses with the mass, The body thus formed is gradually 

 dissolved, but a delicate rod connecting with the distal "centriole " still 

 remains. The distal " centriole " "se dedouble ensuitr et lee _ranules- 

 filles s'ecartent legerement Tun de Vautre." 



The account of the br::_rr: Bouin is veiy :.ef and being unaccom- 



