GENESIS OF FAULT-BARS IN FEATHERS. 34 1 



range of factors which might conceivably be acting on the birds. 

 The results seem to fully justify the statement that none of these 

 conditions can account for the defects, neither in those of the ex- 

 perimented birds nor the occasional ones of the control. The net 

 result of all the experiments thus far served apparently to demon- 

 strate that the important factor in the production or non-produc- 

 tion of the emphasized defects is nutrition. It did not, however, 

 seem to be the only factor, for defects might sometimes be seen 

 in feathers of well-fed birds. The difficulty thus presented was 

 largely cleared up by the progress of studies in another direction 

 (discussed under the next division of this paper). These studies 

 had made it certain that normally one defect is laid down in the 

 germ for each day of growth ; while the examination of the vari- 

 ous types- of fault-bars had made it certain that the defective area 

 of type 1 is a true representative of the defective line of type 3. 

 The latter is, therefore, laid down daily as is the defective area. 

 It follows that the proof of a causal relation between nutrition and 

 defective areas is at the same time a proof that nutrition is causally 

 related to the defective lines (that types I and 3 are merely differ- 

 ent forms of the same defect I have already shown). And since 

 the latter are present in all feathers from tip to tip, one for each 

 day of growth, it is evident that the efficient cause (i. e., nutrition) 

 acts rhythmically. The few defective areas in control feathers 

 could then be due, conceivably, to a slight emphasis of this normal 

 internal rhythm. Since nutrition is a proved factor for the most 

 extreme defects, it seemed extremely probable that it was also 

 playing the chief part in the formation of all types of defects, 

 including the extremely faint, elusive, and universal, defective 

 lines. From this it would appear that the internal rhythm is a 

 daily rhythm, and that it is able — like my experiments — to inter- 

 fere with the nutrition of the feather-germ. This suggested blood 

 pressure to me. To test this idea the experiments described in 

 the following section were undertaken. 



The Production of Fault-bars with Amyl Nitrite. — It is well 

 known that amyl nitrite powerfully reduces the blood pressure in 

 mammals. Some preliminary experiments on the effects of amyl 

 nitrite on the circulation of the chick showed that when traces of 

 it are inhaled, an immediate and extensive vaso-dilatation occurs. 



