34 
the mosque swallows also may well have come from that direction. 
I have retained the name striolata for our Calayan birds as that 
name has been used for the Philippine mosque swallow and as our 
birds appear to be typical of neither striolata nor japonica. In 
dealing with this group of swallows it is well to note the remarks of 
Dr. R. B. Sharpe on “Hirundo daurica and its allies.”? In part 
he says: “The breadth of the rump band or the amount of striation 
on the rump, seems to me to be equally as variable as the length 
of the wing and the coarseness or fineness of the shaft stripes on 
the under surface. Some stress has also been laid by Mr. Hume on 
the dusky ear coverts; and at one time I thought myself that this 
would prove a reliable point of difference between the races; but 
I find that in a large series the darker and coarser the breast stripes 
the darker the ear coverts, and just as every gradation in length of 
wing can be found, so every intermediate form between the nar- 
rowly striped examples, with finely streaked ear coverts, and the 
coarsely streaked specimens, with nearly uniform blackish ear 
coverts, can be met with in a series * * *. It also appears to 
me impossible to settle the question of some of these races until a 
larger series is at our disposal from various parts of China, and 
especially from the headquarters of these swallows in winter.” 
*Cat. Bds., X, pp. 158-159. 
O 
