32 The Philippine Journal of Science 1920 
faction cannot be accepted without reserve as final proof of 
scientific accuracy, in the long run it is a very safe criterion. 
In the case of the Wassermann reaction, clinical satisfaction is 
our only criterion. This may be rather distasteful to some of 
us who have been prone to decry empiricism. The Wassermann 
reaction is purely empirical in the sense that it is experimental 
rather than scientific. The failure to appreciate this fact has 
led many physicians astray. It has led to errors both in the 
laboratory and at the bedside. The laboratory worker should 
not fail to check his results, whenever practicable, with the clin- 
ical aspects of the cases. The clinician should never regard the 
Wassermann reaction as an infallible means to the easy diag- 
nosis of syphilis. He is perfectly correct in looking upon a posi- 
tive Wassermann reaction ag the most constant single symptom 
of lues. But before a clinician accepts a report as positive he 
should have done one of two things: either he should have sup- 
plied the laboratory man with all available clinical data, or he 
should have familiarized himself with the particular test which 
is done in the laboratory to which he sends his sera so that he 
is himself in a position to interpret the test. Until such time 
as the test shall have been standardized and made “fool proof,” 
certainly one or the other of these alternatives should be adopted. 
It is well recognized that widely different results may be ob- 
tained in Wassermann reactions by the use of different antigens, 
and by the employment of different incubation conditions even 
with the same antigen. In a communication published ' in collab- 
oration with Dr. Ward J. MacNeal, I reported upon Wassermann 
tests done by me in the laboratories of the New York Post- 
Graduate Medical School and Hospital by three methods, on 
496 identical specimens of sera and spinal fluids from 477 pa- 
tients. In the first method a cholesterin-reénforced antigen was 
employed, and the first incubation was carried out at 37° C. 
In the second method a simple alcoholic extract was used as 
antigen, with incubation also at 37° C. In the third method the 
latter antigen was again employed, but the first incubation was 
carried out in the refrigerator for from four to twenty-four 
hours. The last method proved more sensitive in the group of 
known syphilitics than the other procedures tested. Further- 
more, a positive result thus obtained proved to be more trust- 
worthy evidence of syphilis than did positive results obtained 
with the cholesterin-reénforced antigen and first incubation at 
 Y aa, 
*Journ. Immun, 2% (1916). 
( 
