Dice cant aa See a ¥ ; Bi Brahe . ’ a a 
450 The Botanical Gazette. ' [November, 
crossing of filaments below the two centrosomes, neither of 
which has been observed. 
If the views of those who hold van Beneden’s hypothesis 
are correct, it becomes evident that every centrosome with its 
‘ attraction-sphere must arise from a previous one, all the cen- 
trosomes in an organism arising from the primary one in the 
ovum, or according to the view of Guignard, from the two 
that are inthe fertilized ovum, each of which represents the union 
of a male centrosome with one from the female. And thus 
they can be traced backward or forward from one generation 
to another the same as the nucleus. 
The question naturally arises as to whether these bodies are 
present in cells which divide by amitosis, and if present what 
their action is during the process. Flemming (8) states that 
in leucocytes, where division is both direct and indirect, 
the ‘‘central bodies” are present; but they do not seem to be 
implicated in the fragmentation or direct division of the nu- 
‘cleus. He does not state what becomes of the spheres when 
fragmentation takes place, but concludes that only the pro- 
ducts of karyokinetic division continue to live and multiply. 
Neves (19) has worked upon this subject with the spermato- 
gonia of the salamander. He reports some discoveries, which, 
if they can be substantiated, truly present some very won- 
derful phenomena. He says that he saw the attraction-sphere 
become oblong, and that in various stages of the constriction 
of the nucleus the elongated attraction-sphere was twined in 
a ring about the constriction. In some cases the two ends of 
the elongated body appeared as though they were not yet 
fused together. When the division was complete, the elon- 
gated body appeared like a ring lying between but to one side 
of the two daughter nuclei; but there was only one of these 
bodies to the two nuclei. No further observations were made 
in regard to the subsequent action of the body and the two 
daughter nuclei: so the question of attraction-spheres 1m re- 
lation to amitotic division is yet ina very unsatisfactory state. 
Plant cells espectally.—The work hitherto done with soe 
is as follows: Guignard’s investigations stand as one of te 
most important contributions to the subject. Guignard (10) 
found the attraction-spheres and centrosomes both in 
and dividing pollen-mother-cells of Lilium, Fritillaria, 
€ra, and Najas; in the mother-cells of the embryo-sa¢ ee 
nuclei both at rest and in stages of division; in the cells of th 
