37 
any method, unless it be b}^ the precipitin or other biologic tests. 
These experiments were repeated on other samples of urine con- 
taining blood with like results. 
BLOOD IX SALIVA. 
Similar results have been obtained with saliva, as may be seen 
from the following: 0.0036 gram of human blood was dissolved in 
100 c. c. of fresh saliva, and the solution labeled (1) ; some of the 
same saliva was kept for comparison and labeled (2). One c. c.of 
solutions (1) and (2) were mixed with 2 c. c. of water and tliree drops 
of alumina cream. The mixtures were then thoroughly shaken and 
filtered, and the residues washed with redistilled water. Small 
V . 
amounts of these residues were added to 2 c. c. of the phenolphthalin- 
hydrogen peroxide reagent. After standing a few minutes at ordi- 
nary temperature, solution (1) became deep purplish red, whereas 
(2) showed only a faint pink coloration. Small amounts of the two 
residues were then added to alkaline phenolphthalin containing no 
hydrogen peroxide. At the end of five hours solution (1) had be- 
come pink, whereas (2) remained colorless. At the end of twenty- 
four hours (1) was dark purplish red, while (2) was faint pink. The 
unfiltered and filtered sahvas were also tested for blood directly by 
means of alkaline phenolphthalin and hydrogen peroxide, and also 
by means of alkaline phenolphthalin alone, without previous treat- 
ment with aluminium hydroxide. With the unfiltered salivas, 1 c. c. 
of (1) with 2 c. c. of the alkaline phenolphthalin-hydrogen peroxide 
reagent developed a deep pink color on standing twenty minutes, 
whereas (2) showed only a faint trace of pink under the same con- 
ditions, and the filtered solutions gave similar results. The two 
salivas, filtered and unfiltered, were also tested with alkaline phe- 
nolphthalin alone, using 1 c. c. of the saliva and 2 c. c. of the alkaline 
solution of phenolphthalin. Solution (1) showed a fight pink color 
after twenty-four hours, while (2) showed only a trace of pink. 
It is possible, therefore, to recognize 36 parts of blood in 1,000,000 
parts of saliva by direct test with the phenolphthalin reagent, but 
much more satisfactory results are obtained by previous treatment 
with aluminium hydroxide. 
ADSOEPTIVE POWER OF SOIL FOR BLOOD. 
In view of the highly adsorptive power of aluminium and ferric- 
hydroxides for blood, it occurred to me that possibly soil would also 
exhibit adsorptive power towards the blood pigments. In order to 
test the matter, 0.0026 gram of human blood was dissolved in 25 c. c. 
of water and 1 gram of soil added. This mixture was weU shaken 
and a portion of it filtered. Both the filtrate and the residue showed 
the presence of blood when tested with the alkaline phenolphthalin- 
