HIRUNDININ^. 143 



are four skins in the Pryer collection from Yokohama. It builds a 

 long bottle-shaped nest under the eaves of buildings^ and lays six 

 white eggs (Blakiston and Pryer, Ibis, 1878, p. 231, no. 173). The 

 Mosque- Swallow of Asia Minor, Hirundo rufula, builds a similar 

 nest in caves, and lays similar eggs. The Mosque-Swallow of South 

 Africa, Hirundo cucullata, builds a similar nest, often under the 

 verandahs of houses, and also lays white eggs, 



The arrival of the Hume Collection in the British Museum has 

 been of invaluable service to ornithology, and has made many groups, 

 which were formerly in hopeless confusion, comparatively easy to 

 arrange in a satisfactory manner. Of these the Asiatic Mosque- 

 Swallows are a conspicuous example, and I find myself obliged to 

 modify the opinions formerly published (Seebohm, Ibis, 1883, p. 167). 

 The Japanese Mosque-Swallow belongs to the group in which the 

 colour of tBe rump is uniform (not gradated) . This group appears 

 to contain four species, though it is very probable that some of them 

 may hereafter be found to intergrade. Two of these species have 

 narrow streaks on the underparts, whilst those on the rump are 

 almost or quite obsolete. One of them is large (wing 5*2 to 4*9 inches), 

 somewhat rufous on the underparts, and may be called Hirundo 

 alpestris, breeding in South Siberia, and wintering in Mongolia and 

 Thibet. The other is smaller (wing 4'5 to 4'2 inches), not quite so 

 rufous on the underparts, and may be called Hirundo erythropygia, 

 breeding in the Himalayas and wintering in the plains of India. 

 The other two species have hroad streaks on the underparts, whilst 

 those on the rump, though narrow, are very conspicuous. One of the 

 second pair is large (wing 5'4 to 4'9 inches), with little or no rufous 

 on the underparts, and may be called Hirundo striolata, breeding 

 from Assam across Southern China to Formosa, and in winter ranging 

 as far south as Java. The other is smaller (wing 4*8 to 4'4 inches), 

 more rufous on the underparts, and may be called Hirundo nipa- 

 lensis, breeding in the Himalayas across North China to Japan, and 

 wintering in Burma, Flores, and doubtless other islands of the 

 Malay Archipelago. I am, however, of Mr. Sharpe's opinion 

 (Sharpe, Cat. Birds Brit. Mus. x. p. 159), that these four forms 

 intergrade, and can only be regarded as subspecies of Hirundo alpes- 

 tris, in which case the Japan examples must be called Hirundo alpes- 

 tris nipalensis. The extremes appear to be very distinct, but 

 H. alpestris probably intergrades with H. erythropygia, which appears 

 completely to intergrade with H. nipalensis, which again appears to 

 intergrade with H. striolata. 



