198 AN INTRODUCTION TO MODERN GENETICS 



obtained, and Goldschmidt discovered modifiers which affect the dose- 

 effect curves of the mutants and thus alter their dominance relations 

 compared to wild type (he has coined the word "dominigenes" for 

 such dominance modifiers). Moreover, just as in the Ephestia case, the 

 phenotypic effects can be produced as phenocopies by suitable tem- 

 perature shocks. 



Lillie and Juhn^ have shown that the mechanism of pattern-formation 

 in the feathers of fowls also depends on an underlying pattern of 

 developmental rate like that discovered by Goldschmidt in Lepidop- 

 tera. They stimulated pigment formation by injections of thyroxin or 

 female sex hormone into Leghorns in which the females are more 

 darkly coloured than the males, and observed the deposition of pigment 

 in new feathers regenerating from follicles from which the original 

 feathers had been plucked. Different parts of the individual feather 

 differ in growth rate, and so do feathers from different regions of the 

 body. In both cases, the slower the rate of growth the lower the con- 

 centration of hormone which is effective in stimulating pigment forma- 

 tion. But these differences in thresholds are not the only factors in- 

 volved, at least in the individual feather. Although the slow growing 

 regions (those near the shaft of the feather) react to a lower concentra- 

 tion of hormone, they also have a longer latent period between the 

 time when their threshold is reached and their reaction begins. If a 

 large dose is given, and is gradually absorbed, the time taken for the 

 concentration in the feather germ to reach the threshold of the slow- 

 growing parts, and then for the slow-growing parts to begin reacting, 

 may be just as long as the whole time necessary to reach the high 

 threshold of the fast growing parts which begin reacting almost at once. 

 In such a case, all parts of the feather will begin depositing pigment at 

 the same time. By varying the dosage, one can vary the relative impor- 

 tance of threshold and reaction time, and in this way obtain different 

 patterns of pigment. It is then fairly easy to see how some pattern- 

 controlling genes might work, though none have yet been fully worked 

 out. 



The pattern of feather growth rates in different parts of the body is 

 genetically controlled, and becomes determined quite early in develop- 

 ment, so that pieces of skin from newly hatched chicks, when trans- 

 planted to other sites in the body, develop the same kind of feathers as 

 they would have done in their place of origin. ^ Wright^ has suggested 



^ Lillie and Juhn 1932, Juhn and Fraps 1936. 



2 Danforth 1929. ^ Wright 1917. 



