GENESIS OF FAULT-BARS IN FEATHERS. 339 



When adult ring doves were starved for periods of one to three 

 days, those portions of the feathers groivn during those days showed 

 well-marked fault-bars, one (exceptionally two) for each day of 

 growth. In such experiments the length of the feathers was 

 measured at the beginning and again at the end of the starving 

 period. For this experiment a control bird of similar age and 

 condition was kept and fed normally in a cage alongside the 

 starved bird. Measurements of the feathers of the control being 

 taken also. The effect of three grades of feeding on the rectrix 

 of a dove is shown in PI. XIII., Fig. 6. 



Many experiments have been made upon feathers which were 

 replacing others that had been purposely removed. In these 

 cases, the rectrices were pulled on the same day from two or 

 more similar birds. After their new feathers had started to grow 

 they were divided into experimental and control birds, and the low- 

 feeding commenced. In PI. XIII., Fig. 6, the effects of several 

 days of starvation on such a feather have just been noted. This 

 feather shows an absence not only of most of the barbides of the 

 affected region but of the distal ends of the barbs as well. Care- 

 ful inspection showed an occasional defective area in the control 

 also. 



The effects obtained by starving chicks — old or young — are 

 in every way comparable to those just stated for doves, and a 

 separate description therefore need not be given. Any consider- 

 able reduction of the food of doves and chicks will invariably 

 produce well-marked fault-bars in many of their growing feathers. 



The Feeding of Sudan III. — While the fat stain, Sudan III., 

 was being fed to some young chicks for a quite different pur- 

 pose, these were found to be producing defects similar to those 

 produced by starving. This led to a careful study of what the stain 

 could accomplish in the way of producing these abnormalities. As 

 a result of this study it can be said that when Sudan is fed in large 

 amounts fault-bars are laid down in much the same manner as in 

 the starving experiments. The chief difference being that with 

 Sudan the defects much more frequently take the form of con- 

 strictions (type 4) than of defective areas (type 1). Examples of 

 the latter type are not uncommon, however, and were even large 

 enough to appear in photographs (PI. XIII., Fig. 13). That the 



