70 
These same filters were tested with a bouillon culture of Staphylo¬ 
coccus pyogenes aureus in a reverse manner to fig. 2. From three to 
four hours were taken in passing 150" cc. of the bouillon culture 
through each filter. The filtrates remained sterile after twelve days 
in the incubator. 
The Berkefeld filters were rigged with the glass cylinders which 
come with them and a vacuum was used. About four hours were 
required to run 150 cc. of the bouillon culture through each filter. 
THE FILTRATION OP CERTAIN VIRUSES. 
Peripneumonia of cattle, rinderpest, foot-and-mouth disease, South 
African horse sickness, exudative typhus of chickens, mosaic disease 
of the tobacco leaf, yellow fever, epithelioma contagiosum of fowls, 
hydrophobia, clave!ee, and hog cholera have each been shown to be 
due to a virus which passes the pores of certain porcelain and diato- 
maceous filters which hold back the ordinary bacteria. 
With the exception of peripneumonia we know nothing of the 
character of the infective agent in these filtrates, which by direct 
microscopic examination and by cultural methods have yielded no 
morphological entity. 
The outlook for finding in the body fluids, the specific cause of any 
one of these diseases, by the microscopes in present use is encouraging 
as long as we can say that we have a filter which will not allow the 
virus of the disease to pass, but which does allow the particles of 
some test substance to be plainly seen in the filtrate. 
It is far more important to know what particular filter a certain 
virus can not pass through than it is to know what brands of filter 
it does pass through. Given a filter that will not permit the virus 
of a disease to pass through its pores, and if on testing that filter 
we find that we can put through it visible particles of. some test sub¬ 
stance, there is plenty of hope that the infective agent of that virus 
may be visible with our present oil-immersion systems. 
On the other hand, if we can find a filter which will not transmit 
particles of microscopic size and yet will allow the virus of a certain 
disease to pass into the filtrate, we can not expect to see the individual 
entities in the virus. 
Several factors influence the filterability of a virus, namely, the 
kind of filter used, the character of the menstruum in which the 
virus is suspended, the degree of pressure or vacuum used, the amount 
of time allowed to the process, the temperature, the motility of the 
particles, and other factors. Unfortunately, in the study of the 
literature of the various filterable viruses we sometimes fail to find 
exact data on all these points. 
