158 MICBO-OBGANISMS AND DISEASE. [CHAP. 



skin of guinea-pigs produced by setons, is in no case followed 

 by tuberculosis, has conceded that in his earlier experiments 

 there must have entered some error in the use of the materials. 

 Cohnheim has conceded the same. It is clear that these ob- 

 servers, while working at the same period with both tuberculous 

 and non-tuberculous matter, must have had, in the course of 

 experiments with the latter substance, accidental contamination 

 with the former, and hence had the guinea-pigs inoculated by 

 them with non-tuberculous matter nevertheless affected with 

 tuberculosis. Dr. Williams, who had no contamination to fear, 

 working with non-tuberculous matter only, had consequently 

 no accidental contamination. This shows us how dangerous, 

 as regards reliability of results, it is to work in one laboratory 

 with different infective materials at the same period. 



I have myself experienced some very curious results bearing 

 on this very point. During the last year I have seen the 

 following cases of accidental contamination occur. I work in 

 the laboratory of the Brown Institution, which comprises a 

 suite of rooms. Although working extensively on anthrax, I 

 generally limit myself to one room only. A friend of mine, 

 who one day injected into a vein of a guinea-pig blood taken 

 from a blood-vessel of a dog suffering from distemper, found, 

 to his great disappointment, the guinea-pig dead after two days 

 under the typical symptoms of anthrax, the blood of this 

 animal teeming with the characteristic bacilli. The hypo- 

 dermic syringe used in this experiment for injection had^ not 

 been previously used by me in my anthrax experiments, since 

 I never use a syringe in my inoculations, but only glass 

 pipettes freshly made and drawn out into a fine tube. The 

 experiment was performed in the room adjoining the one in 

 which my anthrax investigations were being carried on, but I 

 was in the habit of making every day a good many specimens 

 of anthrax cultivations and spores, so that there must have 

 been a good many of these spores distributed on the table and 

 floor, and probably found their way into the wound of the 

 guinea-pig at the time the above experiment was made. 



Another gentleman working in the laboratory of the Brown 

 Institution intended to inoculate several guinea-pigs with 

 human tubercles. For this end he mashed up in saline solu- 

 tion, in a clean mortar, a bit of human lung studded with 

 tubercles. He did this in my room on the same table on 

 which I was working with anthrax. One of these guinea- 

 pigs, inoculated with human tubercle, died before the second 

 day was over of typical anthrax. Its blood was teeming with 

 the bacillus anthracis. Such an accidental anthrax of a 



