12 
Cases from wliicli blood was taken in first week 7 
Positive 2 = 2S. 5 per cent. 
Cases from wliicli blood was taken in second week S 
Positive 2 = 25. 0 per cent. 
Cases from wliicli blood was taken in third week d 
Positive 3 = 75. 0 per cent. 
Cases from which blood was taken in fourth week 3 
Positive 2 = 66. 6 per cent. 
Cases in which the time of disea.se was not stated 5 
Positive 0 = U. 0 per cent. 
Of the 27 cases examined probably a definite proportion were not 
typhoid at all, for many of these tests were made at the request of 
physicians in doubtful cases to clear up difficult diagnoses. Three of 
the cases, after being reported, were determined clinically not to have 
been tyjihoid fever. Further, it was not practicable to make more 
than one examination from each case. T\T now know that 5 c. c. of 
blood is not always sufficient to demonstrate the presence of the 
typhoid bacillus at a single examination. In one of our cases 10 c. c. 
of blood was taken and divided into two portions. . One portion 
remained sterile, while the other gave a pure culture of the typhoid 
bacillus. 
In each instance the identity of the bacillus was determined by 
complete cultural and agglutinative tests. 
Excluding the three cases clinically determined not to have been 
typhoid, the percentage of instances in Avhich the bacillus was recov- 
ered is brought up to 37.5 per cent. Judging by the results of other 
investigators, the method followed in making these blood examina- 
tions should have demonstrated the typhoid bacillus in a larger per- 
centage of instances if all of the 27 cases had been typhoid fever. 
The number of cases of typhoid fever in the District of Columbia 
which are unrecognized and not reported probably exceeds consid- 
erably the number reported as typhoid fever under mistaken diag- 
nosis. Thus, error in diagnosis alone makes it impossible by epi- 
demiological investigations to account for a considerable number of 
cases. All data obtained in regard to cases reported as typhoid 
fever but which are not are misleading and so obscure the general 
problem. Spread of infection directly by contact or indirectly by 
i^ersons. flies, milk. etc., from cases unrecognized and unreported 
cannot be determined, especially in a large community. 
From the standpoint of prophylaxis, the unrecognized cases are, 
of course, more im^^ortant than the cases erroneously reported as 
tyj)hoid fever. In order that the actual prevalence of typhoid fever 
in a community may be determined and that proper measures to j^re- 
vent the spread of the disease may be carried out, the correct and 
early diagnosis of all cases is of fundamental importance. 
