H. W. Acton and W. F. Harvey 
267 
reaction. We may conclude from these facts that nucleolar matter is 
not directly converted into specific structures (i.e. Negri bodies etc.) but 
only indirectly, by an interaction between it and the cytoplasm. This 
fact we consider has an important bearing on Negri body formation. 
We look on Negri bodies as the result of interaction between extruded 
nucleolar particles and the cytoplasm. This view will become more 
apparent from the following observations. We have found that usually 
the nucleolus of nerve cells in the rabies-brain instead of staining black 
by the iron haematoxylin and Van Gieson method stains a greenish 
colour in some animals and in some terracotta, indicatins' that some 
change has probably occurred to alter its original staining characteristics. 
With the methyl-blue eo.sin method of Mann it still however continues 
to stain the original red jM’obably because this stain has a marked 
affinity for any chromatin structure even remotely derived from 
chromatin (e.g. the erythrocytes). 
Instead of the nucleolus being always a single spherical body, there 
is frequently evidence that fragmentation is occurring in rabies-brains. 
In the rabbit this is most marked (PI. XI, tigs. 5, 6): and the nucleolus 
breaks up into a number of small particles (3 to 6) and these are then 
probably discharged into the cytoplasm in much the same way as in 
the case of the nucleolar discharges referred to above. These discharged 
fragments are in all probability the small Negri bodies seen in this animal 
and apparently do not increase in size with length of incubation. In 
the guinea-pig the nucleolus usually stains a greenish colour with the 
Van Gieson’s mixture; it is generally irregular in shape and fragmen¬ 
tation may be seen. In this animal the Negri bodies also stain a 
greenish colour and are frequently cap-shaped, the base of the cap 
being hollowed out to receive the nucleus (PI. XI, fig. 9). This 
appearance resembles the archoplasmic vesicle very closely, as seen in 
the sperm cell of this animal, and as we have said before, it was this 
fact which led us to undertake this investigation and connect the Negri 
body formation with nucleolar transformation. In the dog and jackal 
a few of the Negri bodies stain green, like the nucleolus in these 
animals, but the majority stain terracotta with Van Gieson’s mixtures. 
Evidence of nucleolar fragmentation occurring in the ganglion cells of 
the hippocampus major of a dog is seen in PI. XI, figs. 13, 14. We 
have only once seen this process so definite in the dog. The nucleolus 
as well as the particles outside still retain their iron stain, but the large 
Negri bodies (PL XI, fig. }.Sb') stain terracotta and show minute iron 
staining granules. 
