238 PHYSIOLOGICAL GENETICS 



I do not think thai these arguments arc valid. We know from 

 all the cases presented above that by temperature effects the 

 phenotype of one type may be changed into that of a mutant 

 type (phenocopy). This does not mean, however, that no 

 pattern is involved in the production of those types but rather 

 that tin 1 process of pattern formation is of such a nature that it 

 may be shitted in different directions through external (tempera- 

 ture) or internal ( modifiers) changes. The cases that have been 

 described illustrate this elementary principle. It is not an inte- 

 gral part of the pattern conception that a pattern-forming process 

 is invariable; it is, indeed just as variable as any other process, 

 and its laws will therefore become better visible in statistical 

 treatment than in the study of individual cases. Child himself 

 finds that at a temperature used by the Russian authors the 

 effect upon bristles agrees with the seriation given by these 

 authors. Apparently, a bristle pattern is actually involved, 

 though it may be shifted by modifying agencies. 



A second point to be clarified is the type of pattern involved. 

 If we think of our former discussions regarding the outlet of the 

 determination stream (called center of diffusion in Plunkett, 

 1926, and Child's papers), the following different situations might 

 be found: (1) There is one point of outlet, and the whole pattern 

 in all cases (barring now the variations) is the consequence 

 of the conditions of the whole system prescribing the bed of the 

 determination stream. (2) There are a number of outlets, and 

 their combined workings with the general conditions of the 

 system in time (time of opening of the outlets, time of flow) and 

 space (direction of the stream, eddies, obstacles, etc.) will control 

 the resulting pattern. (3) Each bristle corresponds to a separate 

 point of outlet, comparable to the different typical spots in a 

 piebald pattern. Then there would be no space element, a flow, 

 but only a time element — opening of the outlet — and a threshold 

 element — sufficient quantity of determining stuff for bristle 

 formation (or bristle destruction). The first two possibilities 

 may be combined with the peculiarities met in our former cases: 

 the bed of the stream might be thoroughly fixed (many moths) ; 

 it might be variable within certain limits (the vestigial series); 

 and it might almost be left to chance (the intersexual wing of 

 Lymantria). In the last case, each individual would be different, 

 though a statistical treatment would reveal an average pattern. 



