H. W. Acton and W. F. Harvey 
259 
From the above facts it can be seen that Negri bodies are not truly 
specific of rabies. At the same time we must state that in none of 
the above experiments have the Negri bodies been either so numerous 
or so large as those seen in rabies. As far as these experiments go 
vm should be inclined to assert that—specificity apart—there is un¬ 
doubtedly a very high correlation between the presence of numerous 
Negri bodies and rabies. Thus their presence in marked numbers 
forms a valuable if not absolutely certain test for rabies. If they are 
few in number or absent, no definite conclusions can be drawn. 
Having shown that Negri bodies are not truly specific for rabies, we 
may now proceed to study their so-called life cycle as described by 
Williams and Lowden in the dog. If Negri bodies were of the nature 
of parasites, we would expect them to be similar in structure and 
development, whatever species of animal acted as host. From this 
point of view the following animals were examined: man, horse, mule, 
cow, goat, monkey, rabbit, guinea-pig, dog and jackal. The tissues 
were hardened in Zenker’s fluid, spirit, formalin, picro-formalin and 
acetic bichromate mixtures; Zenker’s fluid was found to be the most 
satisfactory of them all. The tissue was embedded in paraffin wax, and 
the sections stained either by Mann’s methyl-blue eosin (long method) 
or by iron haeraatoxylin (Mallory) and in the latter case couuter-stained 
by the picro-fuchsin mixture of Van Gieson, or Bordeaux red. 
The physico-chemistry of staining reactions is a particularly difficult 
subject. We do not propose to enter upon it here. But it is necessary 
to give some idea of the alterations in staining reactions of the bodies 
we are about to describe. This is particularly necessary in order that 
our terminology may be rendered exact. Normal nerve cells show 
when stained by Mann’s long method a red nucleolus, and with 
Mallory’s iron haematoxylin a black nucleolus. Now, if the cell be 
a normal one, counter-staining with Van Gieson’s stain or Bordeaux red 
does not give rise to any alteration in the colour of the nucleolus. In 
nerve cells of brains affected with rabies the case is otherwise. With 
Mann’s stain the nucleolus stains red or reddish yellow and Negri 
bodies stain red. With iron-haematoxylin and Van Gieson’s stain the 
nucleolus may stain black, or, more usually, it stains greenish or 
terracotta. Precisely similar staining reactions are seen in the Negri 
bodies. The larger Negri bodies may show in addition to the green or 
terracotta some black dots (iron stain). It would seem as if in rabies- 
brains the iron staining of thfe nucleolus was more easily removed by 
the decolourisatiou employed than in normal brains, thus allowing the 
