Villemin's experiments gave the first positive evidence of infectivity. His experiments undoubtedly did much to popularise the idea that tuberculosis is an infectious disease. Koch's experiments demonstrated this fact, and placed Villemin's induction on a solid foundation. This is a convenient point to revert to the instances in which apparently tubercles had been experimentally produced by non-specific inoculation. Klebs suggested that extraneous infection was the cause of these anomalous results, and Frankel and Cohnheim showed that this was the correct explanation. The evidence that tuberculosis is infective is not confined to experimental inoculation. Experimental observations, however, have proved that it can be spread either by the inhalation or the ingestion of tuberculous material. Tuberculosis has frequently been induced in guinea-pigs by making them breathe in an atmosphere containing dust contaminated by tubercle bacilli. Baumgarten, Fisher, and others have shown that tubercle bacilli can pass through the mucous membrane of the intestine without producing any local ulcer.