THE MUTATED GENE 219 



individual barbs, viz., nearer the tip in the barbs which grow out 

 more vent rally as compared with those growing out toward the 

 shaft. Lines of equal age are therefore parallel to the base of 

 the collar and stretch transversely across the feather. Now, 

 injections of thyroxin produce pigmentation in an otherwise 

 light-colored feather. Lillie and Juhn showed that with smaller 

 concentrations, pigment appears only near the shaft but stretches 

 across the feather with higher concentrations. Thus, there is 

 w T ithin the feather germ a different threshold of reaction to 

 hormones, lower near the shaft and increasing toward the tips 

 of the barbs. These tips have, however, a shorter period of 

 latency and react therefore faster to the proper concentration of 

 hormone. By using these reactivities and injecting different 

 concentrations of hormones at different intervals, many beautiful 

 types of pigment pattern may be produced. Lillie and Juhn 

 believed that the decisive difference in regard to the threshold for 

 response to the hormone was the different rate of growth of the 

 barbs at different points of the collar, which would have reduced a 

 part of the problem to the problem of differential growth. But the 

 recent mathematical analysis of Fraps and Juhn (1936a, 6; Juhn 

 and Fraps, 1936) has demonstrated that differential growth is not 

 involved in the case. It is very difficult to decide whether or not 

 these results throw light upon the formation of feather patterns 

 controlled by genes. Of course, where purely hormonic differ- 

 ences are involved, as in the patterns distinguishing both sexes, 

 which can be shifted at will in regenerating (or transplanted, 

 Danforth), feathers by adding the proper sex hormone, hormonic 

 action, together with thresholds and some yet unknown rhyth- 

 mical process may account for the pattern. This is different. 

 however, for genetically controlled patterns like the Plymouth 

 Rock barring of feathers, which does not respond to sex hormones 

 (see Danforth's, 1929, transplantations). Montalenti (1934) has 

 tried to attack this problem. He found that substances in the 

 blood cannot be regarded as responsible for the rhythmical 

 pattern. But there is one interesting relation: feathers in 

 different body regions have a different velocity of growth, and the 

 breadth of a double bar varies proportionally. If we think of the 

 facts regarding the velocity of differentiation of wing scales in 

 butterflies as related to the pattern formation (see page 241), we 

 might conclude that a similar process might take place in the 



