450 WHITMORE. 



and 20 were in the third stage of tlie disease. None of tlie eases sliowed 

 any signs of syjDhilis. 



The tubercle bacillus was found in tlie sputum of each patient before 

 he was subjected to tl'eatment, and \'on Tjrquet's skin test was made on 

 each. During the progress of the work, every man's temperature was 

 taken twice a day, all were weighed once a week, and the sputum was 

 examined once a month. 



AYith regard to diet, clothing, habits of living and sleeping I'ooms, 

 the patients were all under practically identical conditions. Hygienic 

 and dietetic treatment are carried out with all patients in the hospital, 

 and in April, 1909, a roof ward was opened, where the patients have 

 protection from the sun and rain but are ex])osed to unobsti-ucted air 

 circiilation from all sides. 



An attempt was made to determine whether or not the method of 

 complement fixation offers a means of differentiating between infection 

 with the bovine and with the human type of the tubercle bacillus. 



Twenty-eight of the 100 cases of pulmonary tuberculosis, 2 cases of 

 bone tuberculosis, and 3 strong, healthy men were tested in regard 

 to this point. Tubercle bacilli were present in the sputum of all the 

 cases of pulmonary tuberc\ilosis, and all of the cases of tuberculosis gave 

 a positive skin reaction. The skin test was negative on all of the normal 

 men. The blood serum of each man was tested for complement fixation 

 Avith both human and bovine tubercidins (old) in the usual way with 

 the usual controls (old tuberculins were used, as filtrates were not at 

 hand ^). 



In the preliminary tests it was found that our tuberculins in dilution 

 of 1 to 14 never gave fixation of complement when mixed with inac- 

 tivated nonnal human scram diluted 1 to 5, using a hsemolytic system 

 of goat corpuscles (5 per cent emulsion), anti-goat-rabbit serum and 

 fresh guinea pig complement diluted 1 to 10. 



Tlio ha;molytic amboceptor was of such strength that 1 cubic centi- 

 metei' of a 1 to 1,200 dilution completely hsemolyzed 1 cubic centimeter 

 of corpuscle emulsion in the presence of 0.1 cubic centimeter of fresh 

 guinea pig complement. As I Avas Avorking with human sera it was not 

 practical to obtain large amounts of serum, so I used 0.5 cubic centi- 

 meter of the sei'um dilution as suggested by Simon.* The above dilu- 

 tions were employed throughout the work, using two dissolving doses of 

 the hjcmolytic amboceptor. 



'Detre's differential skin test is made with filtrates of bouillion cultures of 

 human and bovine tubercle bacilli. {Wien. klin. Wchnsch. (1908), 21, 173, 1410.) 

 'Journ. Eaoijcr. Mrd. (19^)8), 10, R73. 



