352 
Mr. Walter Rothschild on 
shews no important differences, I regard it as only sub- 
specifically distinct. If, on the other hand, a form, however 
closely allied to another, occurs in the same geographical 
area with it, I consider it a distinct species. In the case of 
Paradisea apoda and its allies I may be accused of incon¬ 
sistency in not also treating P. minor and P. decora as 
subspecies, but I consider that I am right for the following 
reasons: while P. apoda , P. novae guinece , P. raggiana, 
P. intermedia , P. granti, and P. augustae-victoriae replace 
one another absolutely geographically, have the same 
structure, and intergradate completely in colour, P. minor 
occurs in the same geographical area as P. augustae-victoriae, 
and the ornamental flank-plumes have a different structure, 
the barbules being wider and closer together, so that the 
feathers are less disintegrated and silkier in appearance. 
As regards P. decora, it is true that it has a separate geo¬ 
graphical area, but the ornamental flank-plumes are totally 
different in structure and the female is so different that its 
specific distinctness is incontestable. All the other forms 
in various genera which I have placed as subspecies replace 
one another geographically and do not present structural 
differences. 
In my “ Paradiseidae” in the f Tierreich ’ I united all the 
forms of Diphyllodes, with the exception of D. gulielmi-tertii, 
as one very variable species. I have since got more material 
and, after a careful examination, have come to the con¬ 
clusion that Diphyllodes magnificus should be separated into 
three subspecies as follows :— 
Diphyllodes magnificus magnificus (Penn.). 
Inner secondaries clay-colour.— Arfak Peninsula. 
Diphyllodes magnificus chrysopterus Gould. 
Inner secondaries orange, head greyish brown.— Jobi 
Island and the opposite coast, east of Geelvtnk 
Bay. 
Diphyllodes magnificus hunsteini Finsch & Meyer. 
Inner secondaries orange, head rufous.— German and 
British New Guinea. 
