THE TREATMENT OF TUBERCULOSIS. 597 



the toxic substances extracted from very virulent cultures of human 

 tuberculosis and afterwards a quantity of tuberculin ; those of Babes 

 and Proca, who successively injected avian and human tuberculin, 

 followed by dead bacill of avian and human origin. 



According to the investigators who prepared them, certain of the 

 sera thus obtained had a real antitoxin action ; they prevented the 

 development of bacilli and secured recovery from experimental tuber- 

 culosis. But as a whole the results seemed very uncertain, and despite 

 the optimistic conjectures indulged in, the future of serum treatment 

 for tuberculosis remained undecided. 



The experiments of Gilbert, Roger, and Cadiot were commenced 

 in 1892. Observed facts and experiments had shown that certain 

 animals previously regarded as refractory to tuberculosis, among others 

 the goat and dog, were in reality moderately sensitive to it ; but, 

 struck by the resistance of the Gallinacese to mammalian tuberculosis, 

 these observers studied the action produced by the defibrinated blood 

 and serum of birds on its evolution. In order to obtain sufficient 

 quantities of blood and serum they chose the turkey, in which 

 puncture of the humeral vein yields a considerable quantity of blood, 

 and which can be bled two or three times per month for a consider- 

 able period. 



Several series of experiments made on the guinea-pig showed that 

 the serum and defibrinated blood of birds injected under the skin or 

 into the peritoneum have no greater action on tuberculosis than 

 similar fluids obtained from mammals. They in no way check the 

 course of the disease ; and, in fact, a certain number of the animals 

 treated died before the controls. 



They afterwards studied the action of serum from turkeys 

 previously injected with tuberculous material by way of the veins. 

 For several months these birds were periodically injected with one 

 half to one fluid drachm of an emulsion prepared with tuberculous 

 products, most frequently from the dog. The injections were repeated 

 every week or fortnight, the total number of virulent injections vary- 

 ing between six and twelve. Some weeks after the last, when bacilli 

 had disappeared from the blood, the defibrinated blood or serum was 

 used. In a certain number of guinea-pigs treated at the therapeutical 

 laboratory of the Faculty of Medicine with serum thus obtained, the 

 development of tuberculosis was manifestly less rapid than in control 

 animals. 



In another series of experiments serum-yielding animals were pre- 

 pared by injecting into the veins or peritoneum doses of fifteen to 



