252 n. D. GOODALE 



on the belly and much white in the wings. Judging from a study 

 of the other females, there are four points or centers for the origin 

 of these white areas, viz. : At base of lower mandible, neck ring, 

 posterior end of sternum and outer primaries. The white areas 

 themselves may be large or small. They may all be present or all 

 absent, or part may be present and part absent. In the case of 

 the two females under consideration, all are present and all large, 

 thus producing an efTect which calls to mind the pattern of Indian 

 Runner ducks. Both these ducks have nearly yellow bills. 



One female has a noticeably redder breast than the others. 

 Its meaning is uncertain. 



There appears to be a curious correlation between plain head 

 (fig. 4B) {i. e., head without stripes), obscured speculum and 

 pigmented under surface of the wing. There are four cases of 

 plain head among the females, not including the black individuals, 

 two in Fi and two in F2. All had obscured specula and all had 

 the under side of the wing pigmented. All the blacks have 

 obscured specula and the under surface of the wing pigmented, 

 though it is difficult to determine whether or not the head is 

 really striped. 



DISCUSSION 



The data which I have been able to present are too meagre to 

 permit of much discussion. There are, nevertheless, some 

 features which are strongly suggestive. 



The most striking thing about the Fi generation is the appar- 

 ently greater heterogeneity of the females as compared with the 

 more homogeneous males. A description of one Fi male would 

 fit all the other Fi males, though, of course, larger numbers might 

 prove this statement to be without foundation. The evidence 

 in regard to certain characters is, however, fairly strong. All 

 Fi males have a white neck ring and white primaries. Further, 

 leaving out of consideration for the present the summer plumage, 

 no case of striped head has been seen in the Fi males. There is 

 similar but less conclusive evidence in regard to some other charac- 

 ters. But there is no evidence of an opposite nature. Since in 



