186 



PHYSIOLOGIC GENETICS 



small parts of the four-dimensional field of mean selective values {W) expected from the 

 selective values in figure 30. In addition to selective peaks and pits in the corners, 

 there are shallow selective cols between the lowest peak (abCD) and the intermediate 

 one that is shown (aBCd), and between the latter and the highest peak (ABcd). Figure 

 33 shows the mean selective values along the most favorable path connecting the lowest 

 peak with the intermediate and this with the highest (as well as a less favorable direct 

 path from lowest to highest through a four-dimensional col not shown in figure 32). 

 The mean selective values of the gene-frequency systems of figures 32 and 33 were 

 calculated from the selective values in figures 29 and 30, using formulas 16 and 19 

 in a 1935 paper 1444 with respect to interaction effects. 



Fig. 32. Trajectories of gene-frequency systems. 



ABCd ABcd 



1.000 — *~-r* »*l.250 



aBCD 

 0.875 



t£99C^ v^ 



t/ 



1.000 

 abCD 



0.875 

 aBcd 



-0.750 

 abCd 



Trajectories of gene-frequency systems on surfaces of selective values on two faces of 

 the four-dimensional field defined in figures 29 and 30. 



Random processes need shift mean selective value only 8 per cent as much counter to 

 selection as would be involved in fixation of gene B. This figure brings out the reason 

 why I have held that fixation from unbalanced random drift (the Hageodorn effect) 

 is of no importance in progressive evolution. As this is the point on which my position 

 in the 1931 paper has been misinterpreted most frequently, I will give another quota- 

 tion from it. 1422 The reference is to an extremely small, completely isolated popula- 

 tion. "In too small a population, there is nearly complete fixation, little variation, 

 little effect of selection and thus a static condition, modified occasionally by chance 

 fixation of a new mutation, leading inevitably to degeneration and extinction." 



It should be said that the conditions of balance are such that passage directly from 

 one selective peak to another is not likely to occur except from arrays of genes with 



