GENESIS OF FAULT-BARS IN FEATHERS. 



347 



Table I. 



Showing rate of growth during seven days of starving and in control. The num- 

 bers in the first line at the top indicate the age of the feathers, i. e., the number of 

 days since the feathers of the previous plumage were removed. The first day of 

 starvation is that between the tenth and twelfth days. 



•duce marked effects, some of which have been noted in connection 

 with the production of fault-bars. It has been found that when 

 the starving extends through periods of less than three or four 

 days, that no diminution in the linear growth of the feather 

 results (doves). The effects of such starvation are shown only, 

 or at any rate principally, in that portion of the intermediate cell- 

 layer which, as previously stated, produces the barbules. 



If, however, the starving period is prolonged beyond the third 

 day a marked reduction of linear growth occurs (Table I. gives 

 exact figures for the growth of the rectrices of two starved, and 

 two control birds). This means that under poor nutritive condi- 

 tions the formation of barbide -forming cells is first checked, and only 

 tinder still more unfavorable food conditions will the growth of barb- 

 forming cells be impaired. This is a fact of the highest impor- 

 tance for an understanding of the origin of fault-bars, and also for 

 a proper conception of the basis for the rhythms of pigmentation 

 (formation of the fundamental bars). The proof of this is fur- 

 nished by both microscopic and macroscopic study. We have 

 already referred to Fig. 25 of PI. XV., which shows the halted 

 growth and division of barbule-forming cells in the region of the 

 fault-bars. 



In connection with these facts concerning the region and rate 



1 This bird was found to be in bad condition and died soon after this series of ob- 

 servations was concluded. 



