REVIEW OF RECENT RESEARCHES 139 



found wanting in this branch of practical application, 

 then it is quite certain that he has not mastered the 

 method, and no special investigations should be under- 

 taken without such a certificate of capacity. No con- 

 scientious student should make use of a particular 

 method without having made sure, by some means 

 or another, that he is master of it; and the diagnosis 

 of pregnancy supplies these means. 



A further important rule is, that we must never 

 allow current theories to mislead us. The assump- 

 tion of a specific action of the defensive ferments has 

 been established on the strength of the facts derived 

 from experiments, but it has still to be supported by 

 means of exhaustive investigations. Nor should any 

 final conclusion ever be expressed in regard to a 

 particular case, until it has been definitely determined 

 by clinical means. It is in this respect that preg- 

 nancy offers such a sure basis. \Ve do not know of 

 any other condition which allows of such a clear and 

 indubitable estimate of method. It either exists or 

 it does not, and is never open to a discussion of 

 11 probabilities.' 1 



Valuable service is rendered to research as a whole, 

 if every case, which gives an unexpected result in 

 respect of its reaction, is thoroughly investigated. In 

 the first place the dialvsation tubes should be 

 changed, or else another organ should be used. If 

 the result alwavs recurs, we still have to exclude the 



