34-0 OSCAR RIDDLE. 



Sudan probably acts by reducing the actual nutrition of the bird 

 is a conclusion that will be referred to again. 



Birds for these experiments were kept under similar condi- 

 tions in the three compartments of a specially constructed brooder. 

 One lot of chicks served as control ; another was fed a small 

 amount of the stain ; another a maximum amount. The number 

 of fault-bars produced in any feather stand in this order (the num- 

 ber of feathers grown by the birds in the reverse order). The 

 stain was fed in creamy milk (all the birds were given the milk). 

 It was found that one had to feed the stain with the milk at the 

 first time the milk was offered ; otherwise the birds avoided it. 



Effect of Mechanically Crumpling the Feather-germs. — In carry- 

 ing through the experiments already described, a considerable 

 amount of handling of the birds was unavoidable.- It, therefore, 

 seemed necessary to learn whether any of the defects, and par- 

 ticularly the few showed by the control, could have been produced 

 by this procedure. This was tested by slightly marking in vari-, 

 ous ways one of the two birds of a brood ; the marked bird (or 

 in other cases the unmarked one) was then occasionally taken 

 from the nest, its feathers measured, etc., as had been done in 

 the earlier experiments. It was found that this ordinary treat- 

 ment zvas not followed by the production of evident defects. When, 

 however, the feathers were strongly crumpled or broken in the 

 region of growth, fault- bars resulted., and by this means it was easy 

 to produce diffuse fault-bars at all levels of the feather and this 

 quite independently of the usual time element involved. That is to 

 say, these large defects were laid down at irregular intervals, the 

 space between two groups of defects depending upon the fre- 

 quency with which the germs were crumpled. Crumpled prima- 

 ries from the right wing are shown along with the corresponding 

 ones from the left wing of a ring dove in PI. XIII., Figs. 20-21. 



General. — Several experiments of various kinds were next 

 tried in order to learn whether the defects were in any way related 

 to other conditions attending the birds in their nests. To this 

 end, one of a pair of young birds were repeatedly taken from the 

 nest and left exposed to cool air more than was the other ; some 

 birds were reared in foul nests, others in clean ones ; some birds 

 were infected with bird-lice, others not, and so on through the 



