BACILLUS TYPHOSUS. 429 
would not. Thus, Schultz has reported lately that 
among 100 cases of non-typhoid febrile diseases 
apparently positive results were obtained in 19 with 
dilutions of 1 to 5, in 11 of these with 1 to 10, in 
7 with 1 to 15, in 3 with 1 to 20, and in 1 a very faint 
reaction with 1 to 25; whereas in as many cases of true 
typhoid he never failed with dilutions of 1 to 50. In 
these experiments it must be noted, however, that the 
time-limit was from one to two hours. A faint re- 
action with a 1 to 25 dilution with a time-limit of two 
hours indicates less agglutinating substance than an 
immediate complete reaction with a 1 to 10 dilution. 
From an experience with the practical application 
of the serum test for the diagnosis of typhoid fever 
extending over three years, it may be said that this 
method of diagnosis is simple and easy of perform- 
ance in the laboratory by an expert bacteriologist, but 
it is not to be recommended for routine employment 
by practising physicians as a clinical test unless they 
have had experience; that with the modifications as 
now employed, and due regard to the avoidance of all 
possible sources of error, it is as reliable a method as 
any other bacteriological test at present in use; and 
that as such, though not absolutely infallible, the Widal 
_test is an indispensable aid to the clinical diagnosis of 
irregular or slightly marked typhoid fever. 
The Isolation of Typhoid Bacilli from Suspected Feces, 
Urine, Blood, Water, etc. In the bacteriological study 
of typhoid infection for diagnostic and other purposes, 
attempts have been made to isolate the specific bacilli 
from the blood, rose-spots, sweat, urine, feces, and by 
spleen puncture. Although the results obtained by 
puncture of the spleen have been encouraging and have 
