6 INTRODUCTION. 



fectly good results, it cannot be too often or too emphatically impressed upon 

 all workers in immunity methods, that the only guard against mistakes and 

 failures in diagnosis is necessarily found in control tests. And especially 

 in doing research work, the latter are indispensable. For, experimental 

 work which involves reasonable possibilities, or has any pretension toward 

 plausibility warrants no true scientific conclusion without the employment 

 of such tests. 



The author has made it a rule, whenever new findings in serum diagnosis are published, 

 always to look for the given control experiments. If these are insufficient, then no matter 

 what the contents may be, the value of the research is slight, for all its claims only may, 

 but not necessarilv must, be correct. 



