334 OSCAR RIDDLE. 



that one sees all possible intergradations, that each marks off a 

 day's growth, that when the area of type I occurs it always falls 

 in the place for the line, that a certain part of the line only may 

 be transformed into the obviously defective area, etc. Type 5 is 

 probably caused by a protracted defective nutrition of a segment 

 of the circle of growing feather-elements in the germ. 



In connection with these fault-bars we may mention another 

 condition met with in some extreme cases in which abnormalities 

 extend almost continuously over several millimeters of the feather's 

 length. There occurs not only a weakening of the parts, but also 

 a lack of differentiation of the feather-elements. This is well 

 shown by feathers from the nape of a chick which had been fed 

 with Sudan III. In this case the inner and outer sheaths failed 

 to separate from the barbs, and a banded condition results. 

 Another example of such a structure grown in nature by an Eng- 

 lish sparrow is shown in PI. XIII., Fig. 1 1. The lack of differ- 

 entiation, curiously enough, occurs in only two rectrices and these 

 were corresponding ones on opposite sides of the tail. PI. XIII., 

 Fig. 12, shows also an adjoining rectrix. The latter betrays by 

 its similarly placed fault-bars the fact that they were both produced 

 by the same cause. I have found similar bands on feathers from 

 the head of the cardinal. They were being produced simulta- 

 neously with deep constrictions in the other feathers of the bird. 



The Feather-germ. ^ — The more prominent defects have been 

 observed in the feather-germ, both in their initial stages and im- 

 mediately before the breaking away of the containing sheaths and 

 the unfolding of the feather-elements. The less conspicuous 

 defective lines (type 3) have, however, escaped all observa- 

 tion in the germ, as indeed they have done in the adult feathers 

 when the microscope was the means of observation. This ap- 

 parently means that the modifications of such regions are ex- 

 tremely slight, not localized with extreme sharpness, and consist 

 chiefly in slightly different powers to absorb or reflect light; this 

 difference in reflecting power being better seen when the field of 

 vision is large and the contrast between the parts of a large area 

 plays a part. 



The appearance of extreme constrictions on an expanded 

 feather-germ has already been cited. The further fact that at 



