Reactions and Peoducts in Interspecific Crosses 121 



An interesting complication is introduced into the series by the reactions 

 shown in figure 8, case 4, in which the output from a single pair of parents may 

 be altered, if the parents have the proper Ac values, by the application of the 

 climatic complexes X and Y to the parents at the time of fertilization of the 

 gametes and in the early stages of development. In the figure is shown the 

 process of production in the same pair of parents of the two discordant re- 

 actions of cases 1 and 2 in the progeny, merely by the use of the proper climatic 

 complexes at the appropriate moment. In this there is nothing that is not in 

 harmony with the other findings in this series of experiments, but it checks 

 some of the results of the other cases, showing that the differences produced 

 are really the product of the differing Ac values, the conditions of the medium, 

 and the nature of the two gametic systems. In all, it is purely a dynamic pro- 

 duct of the interaction of agents, within and without the organic mass, and in 

 no wise anything but purely physical in conception and operation. In the 

 laboratory the order in this type 4 can be varied so that the role of any possible 

 age affects strength of parents, " orthogenetic," " ontogenetic progressive," 

 gametic change, and kindred "interpretations" may be checked and effectually 

 eliminated. No trace of the operation of any of these supposed agencies has 

 been found. In table 12 is given data from four experiments of this sort, 

 showing the average results that are obtained from type 4 reactions. 



I have presented in this digest of the crossings of these two species the re- 

 sults of much more work than appears upon its surface. From 1907 to 1914 

 the series has been in continuous operations, with a minimum of 4 generations 

 per year, and more than 100 matings per year that were productive of progeny. 

 Throughout, the fraternities have been strong and often have been purposely 

 limited in size to save labor and food, so that the total numbers might easily have 

 been far larger than they have been, but no purpose would probably have been 

 served by this increase in mere numbers. I have never had the total numbers 

 compiled that have passed through our hands in this study, but breeding not 

 less than 100 fraternities per year, with on the average of 50 progeny per 

 pair, gives an estimated number of about 25,000 that have in one way or 

 another been consumed in this investigation. Many of these are either routine 

 tests of the material or tests of the different products of experiment, and 

 have been used merely as indicators of the composition in the manner shown. 

 The counts of a considerable part of these fraternities have been made by 

 Mr. Kuehnel or others, and I have in many and in all critical instances checked 

 the determinations, suspicious cases being bred oat and otherwise tested to 

 decide their nature. The earlier series from 1904 to 1907 are regarded as 

 preliminary tests and do not enter into this account to any extent. 



The role of a reaction of this kind in a state of nature is obviously important. 

 It is not difficult to discover in nature, or to conceive of conditions and acci- 

 dents producing exactly the array that is shown in case 1, where the result 

 would be, while in all aspects and classification a signaticollis race in the 

 location, one with potentiality for change, per saltum, on the realization of 

 the proper agent entering into the history of the race, producing diversa or 

 some or all of the other types described. In the laboratory, if perchance the 

 reaction seen in case 1 were all that had been seen, the immediate conclusion 

 would be that the production of a fixed hybrid that breeds true for generations 

 had been accomplished. Case 2 in nature might also be a potent agent in evo- 

 9 



