Since the results of our work are also negative, and the length of 

 this paper limited, I will give merely a summary of the work done. 



In 167 preparations made from the blood of 49 bleedings from 21 

 tuberculous cattle no tubercle bacilli were found in 110.9 hours' 

 search ; 133 of the specimens, made from gravity sediments of blood- 

 citrate mixture, were searched 88.9 hours ; the remaining 34 slides, 

 made from laked and centrifugalized blood, were searched 22 hours. 



In one instance tubercle bacilli were recovered by both methods of 

 examination from blood drawn twenty-five minutes after inoculation 

 of the arterial blood-stream. 



Guinea-pig inoculations with the blood sediments of twenty tuber- 

 culous cattle failed to produce tuberculosis. 



Not less than thirty minutes was devoted to the diligent micro- 

 scopical search of each slide. The findings were negative. In most 

 instances the work was controlled by having two or more individuals 

 examine each slide. Large numbers of acid-fast bacilli were found 

 in one of the specimens prepared by dissolving the blood clot with 

 pepsin. We believe this to be a contamination. Dr. Rosenberger 

 kindly sent us a slide of blood presumably from a case of human 

 tuberculosis. In this we found numerous acid-fast bacilli, indis- 

 tinguishable from tubercle bacilli. He wrote us that we would 

 find the organisms in our preparations if we would only search long 

 enough; so some of our most promising specimens were diligently 

 searched for several hours by three persons accustomed to the use 

 of the microscope, without finding a single organism resembling a 

 tubercle bacillus. Thirty-seven guinea pigs were inoculated with 

 blood sediments. They were autopsied after two months, and failed 

 to show lesions. Control pigs, inoculated with blood sediments to 

 which small quantities of tubercular pus from one of the cows had 

 been added, were tubercular. 



Introduction of Virulent Cultures of Bovine Tubercle Bacilli Into 



the Circulation. 



In order to make a study of the methods of detecting tubercle 

 bacilli in the blood, it was attempted to produce artificially the tuber- 

 culous bacteriemia which we had failed to find. A cow suffering 

 from advanced tuberculosis (No. 7) and a plump three-year-old 

 steer (No. 18) were selected for experiment. The cow had a sub- 

 maxillary abscess from which tuberculous pus and a calcareous 

 gland had been taken. The steer had been twice inoculated when a 

 calf with attenuated cultures of tubercle bacilli. There was no re- 

 action to the tuberculin test. The animals received in their jugular 

 veins quantities proportional to their weights of a suspension of live 

 bovine tubercle bacilli in normal salt solution. No. 7 received 200 cc. 

 and No. 18, 250 cc. The bacilli had been washed, dried, ground dry 

 for ten minutes in a revolving Erlenmeyer flask containing glass 

 balls, and had been rotated for several hours more after the addition 

 of normal salt solution. About 200 mg. of tubercle bacilli were con- 

 tained in 250 cc. Smears of the suspension mixed with blood per- 

 mitted the rough estimate to be made that there were many more 

 free bacilli, and as many clumps, in the suspension injected into No. 



26 



