186 BOTANICAL GAZETTE [ MARCH 
stained with eosin and iron alum haematoxylin. After this cell 
was seen the material similarly treated was re-examined for bodies 
which might be interpreted as centrosomes, but with negative 
results. 
The centrosome-like body was stained red by the eosin, as 
was the cytoplasm. As the centrosome of animal cells in all 
cases that I have observed stains black by the iron alum haema- 
toxylin method, like the chromatin, the staining reaction argues 
against the interpretation of this body asa centrosome. The 
presence of the nuclear membrane still intact, together with the 
nucleolus within the spireme, shows conclusively that this body 
cannot be interpreted as a nucleolus. Whenever a nucleolus 
cast out into the cytoplasm is treated with iron alum and haema- 
toxylin it takes on an intensely black stain (jig. 36). 
On the examination of other cells on the same slide the cyto- 
plasm was seen to contain bodies with similar staining reaction. 
These, however, showed no signs of division and their position 
in reference to the nucleus was inconstant. /%g. 375, 0f two cells 
in distinctly different phases of division, shows these appeat- 
ances in the cytoplasm and indicates that their presence is not to 
be considered significant, but is caused by some imperfection in 
preservation, or else is due to some abnormal conditions in the 
cells such as the presence of parasites. Such appearances were 
not found in other material treated by the same or different 
methods. 
Guignard (2) maintained the existence of centrosomes 45 
dynamic organs in the cells of higher plants; Schaffner (13) 
described them as existing in Allium root tips; but Strasburget 
(14) and many other investigators find no evidence of their 
presence and have concluded that the centrosomes described as 
existing in the higher plants are in reality nucleoli, microsomes, 
or fixation products. Ch. Bernard (16) affirms the existence of 
centrospheres in Lilium candidum, but admits that they are vatl- 
able inappearance. His figures showing bodies in the cytoplasm 
with clear spaces about them resemble mine of nucleoli degenet 
ating in the cytoplasm. 
The results of this investigation upon Allium can furnish no 
