When he returned home, Metchnikoff immediately resumed his work. He continued, with his collaborators, researches on the normal intestinal flora and on the microbian poisons which provoke arterio-sclerosis. They were able to ascertain that certain microbes of the intestinal flora, such as the _bacillus coli_ and _Welch's bacillus_, produce poisons (phenol and indol) which are reabsorbed by the _normal_ intestinal walls and which provoke arterio-sclerosis and other lesions of the organs. A part of those poisons is eliminated by the urine, and the quantity found therein allows one to estimate the quantity contained in the organism. An exclusively vegetarian or carnivorous diet increases its production, while a mixed diet reduces it. During the rest of his life Metchnikoff made systematic and periodical analysis of his own urine in correlation with his diet. From certain facts and certain experiments he concluded that the reciprocal influence of microbes might be utilised to attenuate or to eliminate the noxious action of some of them. Thus, by cultivating the lactic bacillus in the presence of those microbes which produce poisons belonging to the aromatic group, the decrease in quantity and even the disappearance of phenol and indol is observed. All those facts confirmed anterior results which Metchnikoff had obtained, and indicated the route to be followed in his struggle against those toxins which gradually poison the organism and induce premature senility. Having thus elucidated certain questions concerning the part played by microbes in a normal organism, he studied the pathogenic intestinal flora. He began by infantile cholera because this question is simplified by the fact that new-born children are fed exclusively on milk. It was then believed by practitioners that this intestinal disease of infants came from their mode of feeding, from summer heat, and other external influences. Metchnikoff, however, succeeded in demonstrating that the contents of the intestines of infants suffering from "cholera" always included a special kind of microbe, the _B. proteus_; he was also able to give the disease to young anthropoid apes by making them ingest food soiled by the intestinal contents of sick infants, thus establishing the infectious character of infantile cholera. He then attacked another intestinal disease, typhoid fever, of which the microbe (_Eberth's bacillus_) had been known for some time, but had not been studied experimentally, ordinary laboratory animals being refractory. Metchnikoff had again recourse to anthropoids, and succeeded in infecting a chimpanzee by making him eat food soiled by the intestinal contents of a typhoid patient.