146. The Philippine Journal of Science 1915 
way, except that a somewhat smaller proportion of bacteria 
was added to the agar. The agar was cooled to about 40° before 
adding the bacteria and was inoculated while still liquid. All 
inoculations were subcutaneous. The needle was introduced 
well into the subcutaneous tissue, since, if the agar is deposited 
just under the skin, necrosis is likely to take place over it. Care- 
ful-asepsis in inoculation was observed to avoid the introduction 
of contaminating bacteria with the agar. In addition to the agar 
mixtures of both strains (‘“Tbhag A” and “Tbhag H” in the 
tables), one animal was given a thick emulsion of the avian type 
without agar. Three cubic centimeters of the agar mixture 
were given in all except one, which received 2.5 cubic centimeters. 
Avian 6032 developed some diarrhceal disease with par- 
alysis of the hind legs and was sacrificed eight days after in- 
oculation. Transfers from the agar lump in this animal to 
. nutrient medium showed a good growth of tuberculosis in pure 
culture. The other animals showed more or less infiltration 
around the agar mass, which in some cases became the site of 
an abscess with creamy pus. Long before the virulent dose 
was given, practically all lesions had disappeared except in 
avian 6035, which had a lump about the size of a filbert at the 
point of inoculation. All animals were healthy, and six of the 
nine had gained in weight. 
The dose of virulent tuberculosis was given one hundred 
sixty-eight days after the avirulent. The dose was prepared 
as follows: The sputum of three patients with pulmonary tuber- 
culosis was mixed, making in all about 140 cubic centimeters. 
The three samples examined microscopically showed in the first 
moderately numerous tubercle bacilli, in the second many, and 
in the third very many bacilli. To the mixed sputum enough 
antiformin was added to make a 25 per cent solution, and the 
mixture was allowed to stand for half an hour to emulsify. 
It was then centrifugalized at high speed. The sediment was 
washed with sterile distilled water and again with sterile salt 
solution. Microscopical examination showed a large number 
of tubercle bacilli in this mixture. All animals, immunized and 
controls, were inoculated subcutaneously in the right inguinal 
region with 0.5 cubic centimeter of this sediment, diluted with 
a small quantity of salt solution. 
The results are given in Table I. The symbols +, +4, 
+-+-+ described roughly the size and number of palpable tuber- 
cles formed in the inguinal region. The first class includes 
tubercles of approximately pea or small bean size, the second 
of lima bean or hazelnut size, and the third larger sizes. 
