STREPTOCOCCUS PYOGENES. 495 
but not from all. Failure to do good in human infec- 
tion cannot, as a rule, be attributed to the variety of 
streptococcus. The serum will in animals limit an in- 
fection already started if it has not progressed too far. 
The apparent therapeutic results in cases of human 
streptococcus infection are variable. In some cases 
the disease has undoubtedly advanced in spite of large 
injections, and here it has not seemed to have had 
any effect. In other cases good observers rightly or © 
wrongly believe they have noticed great improvement 
from it. Except rashes, few have noticed deleterious 
results, although very large doses have been followed 
in several instances, for a short time, by albuminous 
urine and even temporary suppression. 
In suitable cases we are, I think, warranted in try- 
ing it, but we must not expect very striking results. 
For our own satisfaction, and to increase our knowl- 
edge, we should always have satisfactory cultures made 
when possible, and the streptococci, if obtained, tested 
with the serum used in the treatment. In the cases 
where we want most to use the serum, such as puerperal 
fever, septicemia, ulcerative endocarditis, etc., we 
find that it is very difficult to make a bacteriological 
diagnosis from the symptoms, and in over one-half of 
the cases even the bacteriological examination carried 
out in the most thorough way will fail to detect the 
special variety of bacteria causing the infection. This 
is often a great hinderance to the proper use of curative 
antistreptococcic serum, for it, of course, has no specific 
effect upon the course of any infection except that due 
to the streptococcus. 
Care should be taken to get only recently tested 
serum, for after six weeks the best serum is almost 
inert; much on the market is worthless, and as it is 
