5 jo Humphrey . — On some 
characterize it. That it sometimes exists without an astro- 
sphere, as Boveri believes, still requires proof so far as plant- 
cells are concerned ; but this seems by no means impossible. 
At present we are practically restricted to the presence of the 
astrosphere as a diagnostic character. It is well known that 
the cytoplasmic microsomes are commonly surrounded by 
clear diffraction-areas, which are the wider as the microsome 
is larger. These areas, however, are never so wide as the 
well-developed astrosphere, and their optical properties are 
different. They appear as narrow and empty or watery rings 
about the microsomes; while the true astrosphere is more 
strongly refractive, appearing like a clear thick drop of jelly 
with a darker point, the centrosome, at its centre. The 
attempt is made to show these differences in Fig. 12 , where 
both centrospheres and large microsomes are seen; but it is 
impossible to represent with the pencil the optical difference 
just described. The advantage of conspicuous cytoplasmic 
radiations guiding to the centrospheres, at least in certain 
stages, which are of great assistance in most animal cells, is 
quite wanting among plants with the exception of a few cases ; 
and, even in these, the radiations are far less pronounced than 
in animals. 
Various stains have been said to possess a specific selective 
power for these bodies, or at least for the centrosomes. 
A careful trial of all the important methods proposed for 
staining them which have come to my notice, as well as of 
various unpublished treatments, has entirely failed to yield 
any satisfactory results. Stains which show a strong affinity 
for the cytoplasm often include these bodies in the diffuse 
colouring they impart ; but I have not yet succeeded in 
differentiating them beyond a slight darkening of the centro- 
some in some cases. It would be superfluous to enumerate all 
the methods which have been tried, but it may be well to say 
that neither Heidenhain’s iron-haematoxylin stain, Hermann’s 
process with pyroligneous acid and haematoxylin, nor Rawitz’s 
new method (’95) of inverse staining, has given better results 
than others. The first two of these stains bring out quite as 
