156 Heredity and Eugenics 



It is obvious that progress in the solution of the problem 

 can be made only through experiments based upon known 

 materials which must meet certain rigorous requirements. 

 The experiments of many observers with plants and animals 

 show clearly that changes are produced which are inherit- 

 able in following generations, but do not produce accurate 

 data upon critical theoretical points. Thus, for example, 

 Sumner's recent work on mice is entirely of this order and 

 gives only unreliable results. Nor can experiments give 

 true data upon these points unless the following conditions 

 are complied with: 



1. A stock of known character, whose behavior, germinal 

 constitution, variability, etc., have been determined for a 

 series of generations and kept in strictest pedigreed line 

 cultures. The stock for experiment must be clarified and 

 reduced to a homogeneous condition as far as possible, 

 and the presence of minor strains fully determined and 

 eliminated. 



2. It must be known what stages in the development of 

 the germ cells, if any, are capable of being influenced, and 

 how — by any force intended to be used later as a somatic 

 modifier. Further, the behavior in inheritance of these 

 germinal modifications, if any, must be known for several 

 generations. 



3. The somatic change must be induced at a time when 

 the germ has been found to be not sensitive to the stimulus 

 employed, so that opportunity may be provided whereby 

 there will supposedly be accumulated in the modified soma 

 that something, carrying the potentiahty of reproducing 

 the modifications which the soma has acquired and which are 

 believed by many to become incorporated into the growing 

 germ cells as part of their constitution. 



