iyi2\ FATE OF TUBERCLE BACILLI OUTSIDE THE ANIMAL BODY 321 



TESTING THE samples were tested almost exclusively by 



SAMPLES inoculation of guinea pigs following the method 



outlined in our previous publication. Stains of 

 smears from the infected manure were, made on a few occasions; 

 but since it is impossible to distinguish between dead and living 

 tubercle bacilli by examining stained preparations, no reliance is 

 placed upon this test. The largest sample possible was used so 

 as not to kill the guinea pigs by acute infection. The amount of 

 sample varied, using the centrifuge sediment from 40 cc. of a I 1 /*, 

 to a 2 l /> percent emulsion of dried feces in 0.8 percent salt solu- 

 tion. The sample from the layers of infected manures was taken 

 by cutting out pieces of the dried layer about one inch square in 

 cross section and taking all the manure in this section down to 

 the ground and a small layer of soil with the sample, so as to be 

 sure to get any tubercle bacilli that might have passed into this 

 top layer of soil just under the infected manure. 



The testing of the sample was carried out with more than 

 usual precautions, keeping in mind all the time the probability of 

 producing tubercles in the guinea pig, with dead tubercle bacilli. 

 To be sure that the tubercles in the infected guinea pigs were 

 produced by living, virulent bacilli, the extent and rapidity of the 

 disease was considered, and also the diseased tissues were tested 

 for tubercle bacilli by staining smears, by cultures, and by inocula- 

 tion of another guinea pig. If the second guinea pig became 

 tuberculous, stains and cultures for tubercle bacilli from its dis- 

 eased tissues were made. If, now, all these tests were character- 

 istic for tubercle bacilli, it was considered reasonably certain that 

 the tubercle bacilli found in these diseased tissues were alive and 

 virulent. But when the guinea pig inoculated from the sample of 

 manure became in a short time severely tuberculous, it was not 

 thought necessary always to inoculate a second guinea pig from 

 the diseased tissues of the former. 



WEATHER ^ e weather conditions during the time the ma- 



CONDITIONS nure infected with tubercle bacilli was exposed are 

 given in Table I2. 1 The weather was not marked 

 with any unusual occurrences. With the exception of the loss of 

 the sunshine records from the third to the seventh of September, 

 inclusive, when the electric sunshine recorder failed to work, the 

 records are quite complete. This table records the date on which 

 each test was made, the number of days of exposure to the time 



J The data for this table were furnished us from the Laboratory of Soil 

 Physics of the Department of Agronomy of this station by the kindness of 

 Professor J. G. Hosier. 



