DIAGNOSIS OF TUBERCULOSIS 455 



the sediment examined by staining with carbol-fuchsin. If this 

 method proves negative, cultures from the sputum may be made 

 after treating it with antiformin for two to three hours. Some of the 

 sediment, after washing to remove the antiformin, should be injected 

 into guinea-pigs. It is usually difficult to find tubercle bacilli in the 

 sputum during hemoptysis, but they are found very frequently in the 

 blood-streaked sputum following hemoptysis. 



Blood. Tubercle bacilli are usually not found in the peripheral 

 blood. If they do occur they are there almost invariably in very small 

 numbers. Occasionally positive results have been reported in examina- 

 tions of stained preparations made directly from the blood when every 

 precaution has been taken to preclude the inclusion of extraneous 

 acid-fast organisms. It is far more satisfactory, however, to inoculate 

 the blood subcutaneously into guinea-pigs. 



The Nasal Cavity. Tubercle bacilli have been found occasionally 

 in the nasal passages of healthy individuals, particularly those who 

 have been closely in association with tuberculous patients. It must 

 be remembered that the acid-fast organism described by Karlinsky 

 as the "nasal secretion bacillus," is found occasionally both in the 

 noses of tuberculous patients and in healthy individuals. This 

 organism grows readily on artificial media and should not be confused 

 with the true tubercle bacillus. 



Pus and Exudates. It is frequently difficult or impossible to 

 detect tubercle bacilli in the pus of cold abscesses. Material derived 

 from this source is very frequently caseous, and although tubercle 

 bacilli can not be demonstrated by staining methods, the so-called 

 Much granules are found occasionallv, which are Gram-positive but 

 not acid-fast. In order to demonstrate the infectiousness of this 

 material it is inoculated into guinea-pigs. Tubercle bacilli may be 

 found quite frequently in exudates, but in order to make the diagnosis 

 reliable the material should be inoculated into guinea-pigs. 



Urine. As a rule it is difficult to find tubercle bacilli in urine, and 

 the probability of contamination with the smegma bacillus or the 

 Lustgarten bacillus must constantly be borne in mind. Sediment ing- 

 large volumes of urine and injecting the sediment into guinea-pigs 

 is the most practical method of detecting tubercle bacilli in this 

 excretion, for it rules out both the Lustgarten and smegma bacilli, 

 which do not produce lesions in guinea-pigs. The slight differences 

 in acid- and alcohol-fastness between these organisms and the tubercle 

 bacillus make diagnosis by the direct smear method of doubtful value. 



