116 The Mechanism of Evolution in Leptinotaksa 



criterion that it is either homozygous in action or in composition, and its 

 apparent homozygous action may well be the product of inability to make mani- 

 fest the conspicuous differentiating characters, and so the series is passed on as 

 pure-breeding extractives, when in reality the thorough testing of such races 

 is necessary to determine their true characters. In the great majority of the 

 instances I have had to deal with, the usual tests of present neo-Mendelian 

 operations are not sufficient. 



The most important internal agent is the Ac determiner and its values, and 

 the most important external one is the water-relation, as measured in terms of 

 evaporation from the body. These animals obtain water only from their food 

 and as hygroscopic water that is absorbed from the air through respiration. 

 They are extremely sensitive to changes in the water-relation, and it is sug- 

 gestive that a relation has been found between the water-loss, the determiner Ac, 

 and the recombination of the V determiner in the operations of F^. The 

 modifiability of the internal mechanism represented by the Ac determiner, its 

 rapid change to a balance with the conditions of the medium, and its readiness 

 to shift its values within limits, suggests most strongly not any substantative 

 thing in the organism, but rather a general state of the entire mass of the system 

 in the relations of its component atomic or molecular elements; that is, pro- 

 ductive of rapidity of reactions in one instance or their prolongation in others. 

 Within any species or group of species this agent is capable of change only 

 within limits, and I have not in any instance been able to modify it beyond those. 

 Thus, in L. signaticolKs, 29 days is the absolutely low limit, and one that can 

 not be maintained, while 37 to 40 days is the lowest that can be maintained with 

 any success. The upper limit is 63 days as the condition that can be held in the 

 strain for purposes of work. In L. diversa, the lower limit is about 50 days, the 

 upper 70 days or thereabouts. Even though plastic in its nature, this general 

 condition in the mass of the system is specific within limits for that mass, and 

 the specificity of the condition segregates with the mass of each gamete with 

 certainty and distinctness. None of the extracted diversa gametes or zygotes 

 in these experiments take on the rate found in signaticollis at its more rapid 

 limit, nor is signaticollis made to react as slowly as can diversa, but in the cul- 

 tures both may and do react perfectly Avithin the mid-limits of their ranges, and 

 when the two are reacting in these values, crossing of the two species is always 

 accompanied with the most regular and stereotyped monohybrid Mendelian 

 reaction. Divergence in the rates of action of this agent in the gametes is 

 productive of the changed arrangement in the agents that are productive 

 of elytral characters, although why it should be these and not others 

 it is not possible to state. It may well be true that there are others that are 

 not observed that are changed, and that added investigation will disclose them. 



Precisely the same series of events occur if the cross is made in the other 

 direction, diversa female and signaticollis male, so there is no discovered rela- 

 tion of the action to sex, or sex-linking. 



Decidedly interesting relations of this reaction to the external conditions in 

 the medium have been developed as the outcome of this series of experiments 

 and its analysis thus far. These experiments show a relation of the operation 

 of the internal agents to the conditions in the medium that is most suggestive. 



