196 SOURCES OF ERROR (IN THE PROCESS) 



evaporates unevenly in the incubator. Take, for 

 instance, Cases 6 and 8. In both cases the serum 

 alone nearly reaches the limit, i. Then, should 

 the dialysate, in the experiment organ + serum, 

 evaporate more strongly, or should the correspond- 

 ing dialysate become more strongly concentrated, 

 during boiling, than that of the relative control 

 experiment, then we shall get a positive reaction 

 owi-ng, entirely, to the concentration ; in which case 

 we shall get a wrong result. These examples may 

 be a warning to those who make use of a particular 

 technique in an imperfect manner. 



It is easy to understand that errors in diagnosis 

 have often occurred, and that, on the other hand, 

 brilliant results have been reported. 



As a matter of fact our limit value, i, is seldom 

 attained. Unfortunately, this occurs just when 

 carcinoma, myoma, salpingitis, exudates, suppur- 

 ations, or haemorrhages are present, that is, just 

 when the method should diagnostically give the most 

 valuable differential results. It is obvious that the 

 investigation of such cases requires double care. 



The performance, under absolutely equal conditions, 

 of a particular experiment, and its control, is of deci- 

 sive importance in regard to the results obtained. In 

 the first place, absolutely pure distilled water must be 

 used. Water, which gives an acid or alkaline reac- 

 tion, leads inevitably to erroneous results. Ninhydrin 



