ANATOMY OF THE SHOE-BILL. 647 



There is nothing peculiar in this arrangement ; the diagram 

 resembles closely the similar diagrams that I have given for 

 other diastataxic birds {e.g., 32, fig. 23). There is some difference 

 of opinion as to whether the secondary quills, major and minor 

 coverts, and the feathers nearer the outer border of the wing 

 represent hoiizontal or transverse series. Most writers, following 

 the obvious lead given by the quills and major coverts, have 

 preferred to regard the rows as horizontal. I have taken the 

 other view, seeing in the quills merely the enlarged membei's of 

 the transverse rows which happen to lie along the margin of the 

 wing and to have become the flight feathers, and the points of 

 insertion of the feathers in the plucked wing have always 

 appeared to me to lie in transverse or rather diagonal rows 

 stretching upwards from the quills and reappearing on the under 

 side of the wing in the feathers with reversed surfaces. These 

 diagonal rows were plain in the wing of Balceniceps., but I was 

 surprised to find what I have not noticed in any other bird, 

 although I am by no means prepared to say that it does not 

 occur, that there was a transverse row too many. The row 

 corresponding to the carpal covert and remex cui'ved upwards and 

 backwards ; the next most proximal row had a similar curve and 

 belonged to the distal secondary quill ; then there appeared to be 

 an extra row in front of, and not behind the second secondary 

 quill. More proximally the rows were in regular correspondence 

 with the quills and gradually changed their inclination. I tried 

 to correlate the arrangement with what W. P. Pyciuft has called 

 the "intercalary row " (a ti-ansverse row which in his opinion is 

 associated with the mode of origin of the diastataxic gap by 

 " faulting " of the horizontal rows in development), but was 

 unable to make anything of it. No one appears to have pub- 

 lished any observations on the theory of diastataxy since Pycraft 

 and myself. In our communications to the Linnean Society (28, 

 36) we showed that the condition was not due to the loss of a 

 feather, Pycraft arguing from ontogeny and I from comparative 

 anatomy. I showed that whereas most pigeons were diastataxic, 

 a few were eutaxic and had arrived at this condition by a secon- 

 dary closing of the diastataxic gap. I also showed that the 

 eutaxic pigeons were in other respects more specialized than 

 their diastataxic allies. In later papers (29, 31) I showed that 

 similar conditions existed amongst Kingfishei^s and amongst 

 Gruiform birds. The general inference seems to be clear : that 

 the eutaxic groups are more specialized birds and that in 

 association with their general specialization they have lost the 

 primitive diastataxic arrangement. My argument, however, 

 may be anatomically sound with regurd to the groups presenting 

 both conditions, and yet not applicable to birds generally. All 

 birds may have been eutaxic originally ; certain families may 

 have become diastataxic, and amongst these certain members 

 may have secondarily reverted to the eutaxic condition. Pycraft 

 assumes that the eutaxic condition was primitive, and if he be 



