182 
Dourine 
indicates that 100 per cent, of dourine infected animals, whether in active 
or latent stages of disease, give positive serum reactions, provided that 
an interval of two to three months has been allowed for an incubation 
period in the more or less resistant animals, less than one month being 
sufficient in most cases. 
How does the value of the dourine test compare with the Wassermann 
test for syphilis ? The old name of horse syphilis still clings to dourine 
infections, especially among stock owners and the general public, and 
comparisons have been made both in regard to the nature of the disease 
and the diagnostic tests, tending to lead to mistaken conclusions. 
The reaction in dourine by the method recommended in this paper 
is a specific one. A positive reaction in other diseases or with animals 
in which dourine infection could be excluded, remains unknown to us, 
while in every authentic case of dourine the reaction is invariably 
positive. In my whole experience there is only one case in dispute— 
a negative serum reaction being given where a symptomatic diagnosis 
of dourine was made. However, the symptomatic diagnosis may have 
been at fault; unfortunately, the animal was destroyed before any 
proof or disproof of dourine infection was forthcoming. 
The very few cases on record where a negative dourine reaction at 
a first test was followed by a positive reaction at a second or later test 
can be accounted for by infection taking place only a few days before 
the serum was first collected, or by continued exposure to infection 
between the first and later test. 
In syphilis, on the other hand, negative reactions are of more value 
for prognosis than for diagnosis. A positive reaction may become 
negative after a short course of treatment returning again to positive 
if a cure has not been effected. Further, it is admitted that a negative 
reaction is frequently given in primary syphilis and again at times in 
latent and tertiary syphilis. A source of error, operating in the negative 
direction, is, as Noguchi has pointed out, in that human serum contains 
a variable amount of natural anti-sheep amboceptor, which in some 
cases may be sufficient to hide a positive reaction. Horse serum does 
not contain anti-sheep amboceptor, as I have found by many experi¬ 
ments, so that the anti-sheep haemolytic system can be used in horse 
serum tests with perfect reliance. 
A positive Wassermann reaction may be given in several diseases 
in which syphilitic infection can be excluded, in leprosy, scarlatina, 
certain forms of tuberculosis and carcinoma. The Wassermann 
reaction is not specific. Owing to the great difficulty of obtaining a 
