234 
NOCARD AND ROUX. 
We had given up all attempts when the paper of MM. Metch- 
nikoff, Roux and Salimbeni, upon the choleric toxine and 
antitoxine, was published.* The results that they obtained 
with cultures in vivo^ in bags of collodion, gave us hope of 
success. In a few words, the principles and technic of this in¬ 
genious method of culture are as follows : 
Small bags of collodion with very thin walls are prepared: 
when sterilized in the thermostat, a few cubic, centimeters of 
bouillon are introduced into them, with a little of the virulent 
liquid to study ; the bags are well closed and then introduced 
into the peritoneal cavity of a susceptible animal—guinea pig, 
rabbit, dog, sheep, cow, etc. All these manipulations are 
quickly learned, and on no occasion does the animal seem to 
suffer, either from the operation or from the presence of the 
bags in the peritoneal cavity. 
After a varying length of time, from a few days to several 
months, according to the nature of the microbe that one studies,, 
the animal is killed ; the bag is found lodged in some part of 
the peritoneal cavity, surrounded by a more or less thick enve¬ 
lope of fibrine and cells, or of young fibrous tissue, from which 
it is easily enucleated. 
When the animal of experiment and the liquid of culture 
have been properly selected, surprising results are obtained, 
which can, however, be easily interpreted. 
The wall of collodion offers an insuperable obstacle to the 
microbes as well as to the cells ; the former cannot get out of 
the bag, but can multiply in perfect security, because the latter 
cannot enter it; they are protected against phagocytosis. Be¬ 
sides this, this wall, inaccessible to microbes and cells, is per¬ 
meable to the liquids as well as to substances in solution; it 
forms a perfect osmotic membrane ; through it exchanges take 
place which deeply modify the primitive composition of the 
inclosed liquid; substances elaborated by the microbe can be 
diffused outside, and when they are sufficiently active or when 
the animal is sufficiently sensitive, they may produce the death 
* Annales de V Institut Pasteur, 1896 , p. 257 . 
