Analysis of Heterogeneity in Complex Characters 261 



the normal range as far as operation is concerned, to be interpreted? This 

 intense " quantitative accumulation " of a character did not in any way change 

 the gametic nature of the strain. Only in a certain portion of the strain there 

 was an accentuated condition of the organism, such that pigment was produced 

 over a wider area of the surface than was usual in the race experimented with. 

 The result is comparable to the quantitative accumulation of any other substance 

 or condition in organisms which may have been recorded. In the experiment 

 described it is clear that increased amount is persistent through many gener- 

 ations, so long as only parents with high pigment-forming values are used as the 

 progenitors of the next population, and it is evident that some germinal constit- 

 uent is exaggerated, but in what respect? The pigment itself is passive, a 

 product of the oxidation of a chromogen, and a capacity for pigment production 

 exists at all times in all portions of the organism. Whatever determines pattern, 

 either itself has, or has associated with it, something which inhibits pigment for- 

 mation over certain areas and allows it in others. In the selective accumulation 

 that factor which limits pigmentation in the normal strain is weakened so that 

 it allows pigment to be formed over wider areas than is normal, but it appears 

 to be impossible to do more than diminish (or increase) this factor by the cumu- 

 lative method, and no fundamental change is made either in the factors present 

 or in their relations, and only a distorted action (a condition) is produced which 

 may be varied at will. 



Another series of experiments similar in import was made by starting with 

 biotype 4 and attempting, by quantitative diminution, to produce biotype 8 in 

 its extreme form. For this material from the same culture of homozygous-act- 

 ing biotype 4 stock was taken and matings were made from the extreme lower 

 limit of the type present. The results of this attempt are shown in figure 47. 

 No modification, no change of pattern, and little decrease of pigment was found, 

 and the whole culture presented the appearance of being against a limit below 

 which it was not able to be forced by mere subtraction of substance. 



At different times I have attempted modification of other biotypes in L. 

 multitceniata, hut never with any success when the method of quantitative 

 accumulation or diminution tvas v^ed. The continued attempts to produce 

 changes by this method have thus far resulted in failure by all workers and in all 

 kinds of organisms. The reason for the failures appears to be because there is 

 nothing in the method of operation which could either remove any factor or 

 change its relation in the germinal complex. These biotype groups provide ade- 

 quately refined materials for such tests, and while with complex populations 

 (mixed) results of supposed quantitative accumulation may be obtained, the 

 results are due entirely to other processes. In refined strains such as I have 

 used or as have been used by Jennings, Johannsen, and others, when properly 

 protected from errors due to improper experimental culture, results of the same 

 kind have uniformly been secured. 



The results show that there is in the method employed no point of entrance 

 produced in the organic system whereby anything can be added to or taken away, 

 or any relation within changed, and these are the only known means of inducing 

 change. In the strains the continued homozygous-acting gametes give in each 

 reproduction and in gametogenesis the same processes over and over again, 

 with only increase or decrease of some factorial action and manifestation. There 

 is no opportunity for metathesis or for possible change following this. The 



