394 A MANUAL OF BACTERIOLOGY 



in water, and some of the pink colour returns. The treatment 

 with acid and with water alternately is repeated until the pre- 

 paration is nearly colourless when rinsed in water. With sputum 

 this is usually the case after three or four rinses in the acid, but 

 it varies with the thickness of the film and with the number of 

 tubercle bacilli present ; when these are absent the film often 

 decolorises more readily than when there are many. The presence 

 of blood renders the decolorisation difficult. After decolorising 

 and washing, the preparations are stained for half a minute in 

 Loffler's methylene blue, washed in water, and dried. The pre- 

 paration is examined without a cover-glass with the oil-immer- 

 sion after applying a drop of cedar oil to the film, unless a per- 

 manent specimen is desired, in which case it should be mounted 

 in Canada-balsam. 



The tubercle bacilli appear as delicate red rods, oftenbeaded or 

 segmented, on a blue background composed of cells, mucus, and 

 putrefactive or other bacteria. Occasionally here and there a 

 little red colour may be present in addition to the tubercle bacilli. 

 Hair and keratinised material generally, such as horny epithe- 

 lium, and red blood-corpuscles, retain the red colour after the 

 foregoing treatment, and the spores of bacteria are also liable to 

 retain the red somewhat persistentfy. These exceptions are not, 

 however, likely to prove a source of error, for the tubercle bacilli 

 should be recognised not only by their red colour, but also by 

 their characteristic size, shape, and general appearance. It is con- 

 ceivable that acid -fast bacilli not tubercle might be present in 

 sputum, but such an event is a very unlikely one. (See Plate IX, ft, 

 and Plate X, a.) 



If tubercle bacilli are not found, other specimens should be 

 prepared and examined. It is only by repeated examinations on 

 different occasions tJiat the negative evidence, the absence of tubercle 

 bacilli, becomes of any value. 



The tubercle bacillus is occasionally not acid-fast ; probably 

 the bacilli in such cases are degenerate, and, like all degenerate 

 bacteria, fail to stain well. Spengler claims that the following 

 method will stain these and "splitter" forms: (1) Stain with 

 warm carbol-fuchsin by the ordinary method, avoiding over- 

 heating ; (2) pour off the stain without washing and treat with 

 picric acid alcohol (equal parts of saturated aqueous picric acid 

 and absolute alcohol) ; (3) after three seconds rinse with 60 per 



