WHXTE-RUMPED WOOD-SWALLOW. 
beaks, those from Timor, New Guinea, and from the Philippines, have them 
generally somewhat smaller, but the differences are very small and not 
constant enough for even subspecific separation, while those from the 
Andamans have the beaks of the same small size as those from Queensland ! 
It would therefore, according to the diagnosis of my new subspecies, be 
necessary to call them also A. I . parvirostris. I believe it is very interesting, 
even for those who do not accept this subspecies, to know that A. leucorhyn- 
chus is not so constant as ornithologists made us believe, and that (as in 
some other cases) a form deviates from its average form in the centre (?) of 
its distribution in a similar way in both directions towards the outer limits 
of its area. Other such examples are Cacatua triton, Nyctidromus albicollis, 
Macroptexyx mystacea. I cannot agree with Sharpe and Salvadori in 
rejecting the oldest Linnean name leucorhynchus. Brisson’s figure is correct, 
his description leaves no doubt, and only this form inhabits the Philippines, 
so that there is no reason for rejecting this name ; in fact, there are many 
less certain names in use in ornithology than A. leucorhynchus. The upper- 
side may, in contradiction to the white underside, have been termed black 
without going far wrong.” 
Two years later, in a paper dealing with Timorlaut Island birds, Hartert 
retracted in a peculiar manner, writing : “ I am not prepared to unite the 
Timorlaut Artamus straightway with A. leucorhynchus. It is true that the 
alleged white tips to the rectrices are not a distinguishing feature . . . 
The rather large, high bill, however, seems a good distinguishing character. 
. . . There would be three races, namely : 
“ Artamus leucorhynchus leucorhynchus. Philippines and most of the 
eastern islands. (Bill median. ) 
musschenbrocki. Tenember Islands, and perhaps 
some other islands. (Bill 
large, higher.) 
leucopygialis. North Australia, and perhaps 
parts of Southern New 
Guinea. (Bill smaller.) 
This is the form which I named A. leuc. parvirostris, but I think Gould’s 
name leucopygialis should be used for it.” 
It will be noted that Hartert did not mention his blunder of citing 
authorities for lumping, whereas Gould and Sharpe, the only two ornitho- 
logists who really mattered in this connection, had both recorded the 
opposite. Apparently Hartert had never referred to either of these authorities, 
and when he was righting his error he wrote : “I think,” as if there was any 
doubt. Gould had named the bird correctly and on the same grounds as 
223 
