70 KARYOKINESIS. 



during the cycle of division, i. e., both are chromatic at one stage and plasmatic at 

 another. I have already remarked upon the fundamental similarities between the 

 nuclear reticulum and the centrosomal reticulum, and also upon the parallelism in 

 the cycle of changes which both structures undergo. While therefore I recognize, 

 together with the authors named above, the resemblance of the metazoan centro- 

 some to the intra-nuclear spindle of the Protozoa, i. £•., the micro-nucleus minus the 

 chromatin, I maintain with E. Hertwig that one need not go so far as the Protozoa 

 to find structures homologous with the centrosome since such may be found in the 

 formed achromatic substance of metazoan nuclei, i.e., in " nuclei Without chro- 

 matin " (basichromatin). 



3. Spheres. — The large, densely staining sphere which persists through the 

 whole of the resting stage in all of the gasteropods which I have studied, represents 

 in the main the cortical zone of the attraction sphere, though remnants of the outer 

 zone of the centrosome vasij be found in it after the origin of the new amphiaster. 

 In each cell-generation the new centrosome and, perhaps, also the new cortical zone 

 arises within the old centrosome {cf. figs. 25 and 70), so that the anlagen of both 

 centrosomes and spheres (at least in some mitoses) come from the previous centro- 

 some. Achromatic substance from the nucleus plus hyaloplasm from the cell body 

 fills the cortical zone and swells it into the enormous sphere of later stages of divi- 

 sion. After the new amphiaster appears it moves out of the old sphere, and the 

 latter may persist for a long time as a degenerating structure. 



In the eggs of all animals spheres are usually present during mitosis,^ but they 

 usually disappear at the close of division, and in no other case can I find any account 

 of bodies with so compact a structure persisting throughout the whole rest, or even 

 through the following mitosis.^ 



In spermatogenesis, however, bodies undoubtedly similar to these sjjheres have 

 been described by several authors, particularly by Moore, '93, Meves, '94, '96, 

 '98, Rawitz, '96. Meves finds in the resting spermatogonia oi Salatnandra at the 

 end of summer that the sphere becomes a heap of granules ; in the spring the sphere 

 is reconstituted out of these granules and in the summer a "consolidated sphere" 

 bounded by a sharp line or even a membrane is found. This sphere probabl}^ con- 

 tains a centrosome at all times, though it is not always visible. The daughter cen- 

 trosomes and central spindle arise within this sphere. The sphere is composed of a 

 cortical and a medullary zone, and the former, at least, breaks up and is scattered 

 through the cell. Masses of these granules collect close under the cell wall, and at 

 the equator of the cell, and when the cell body divides they lie along the newly 

 formed cell membrane. These granules are frequently present for some time in the 

 daughter cells, but usually disappear before the next division. In all stages of this 



1 Carnoy and Lebrun ('97) totally deny the existence of spheres of any kind. They say : 

 "Boveri's arehoplasm is a part of the cell-substance; van Beneden's atti-action sphere, Guignard's 

 directive sphere, Vejdovsky's periplast, the earlier kinoplasni of Strasbiirger do not exist as such. 

 These myths belong in the legends." . . Such sweeping statements are chiefly valuable as illustrating 

 the worthlessness of dogmatic biology. 



' Bolles Lee ('95) finds in the testis cells of Heli.v that the spindle remnants are preserved from 

 generation to generation, the old spindle remnants fusing with tlie new ones. 



