SHORT-TOED LARK. 375 



winter. It is a rare summer visitor to South Germany, but has been 

 obtained on Heligoland on migration. It is a resident in the Canaries 

 and North-west Africa ; but to North-east Africa it is only a winter visitor, 

 occurring as far south as Abyssinia. It breeds in Persia, Turkestan, and 

 South-west Siberia, at least as far east as Sergiopol (long. 80°), where it 

 was obtained by Finsch. It passes through Afghanistan on migration, and 

 winters throughout India *, but has not been recorded from Ceylon. It 

 is possible that it may occur in East Siberia; but travellers from that 

 country have not discriminated between this species and A. pispoletta f, 

 which differs in having the innermost secondaries half an inch shorter than 

 the primaries instead of the same length. 



Like all the other Larks, the Short-toed Lark varies in colour from 

 greyish brown to rufous brown. The palest examples I have seen are from 

 Samarkand, and the most rufous from Spain, Greece, and Palestine. The 

 only very near ally of the Short-toed Lark is the Tibet Short-toed Lark, 

 A. tibetana, which has been found in the North-west Himalayas at Sikkim, 

 and on the northern slopes of the Karakoram Pass. It differs in having 

 a slightly longer bill, white instead of smoke-brown outside tail-feathers, 

 and in having the general colour much less sandy. It is probably specifi- 

 cally distinct, because it has the first four primaries nearly equal in length, 

 whereas in A. brachydactyla the fourth primary is considerably shorter 

 than the three first. 



Although the Short-toed Lark is very common in South Europe, it 

 must be included in the list of Siberian birds which occasionally appear on 

 our shores in autumn among the regular visitors which form part of the 

 great stream of migration from the east. Fed by detachments from the 

 north up the valleys of the Yenesay, the Obb, and the Petchora, this 

 mighty wave of birds sweeps across Europe every autumn, losing itself 

 for the most part on the shores of the Mediterranean ; but its most 



* I have a specimen collected hj Oates, labelled " cJ, Pegu, 25 Sept. 1880." 

 t This species, like most of the other Larks, is subject to great climatic variation, which 

 has caused it to be subdivided into numerous imaginary species. The darkest of these 

 forms, A. pispoletta var. hatica, appears to be confined to Spain. The typical foriji is found 

 in South Russia and Persia. A slightly more sandy form, A. pispoletta var. minor, 

 occurs in North Africa, Palestine, and Asia Minor. A still paler form, A. pispoletta var. 

 leucophma, occurs in Turkestan, along with intermediate forms between it and the pre- 

 ceding race. I have not had an opportunity of examining examples from East Siberia ; 

 but a rufous form, A. pispoletta var. cheelensis, occurs in North China. These varieties 

 are doubtless climatic rather than geographical, and might be supplemented by many in- 

 termediate local forms from difierent localities ; but they all have short innermost secon- 

 daries, a character which makes them specifically distinct from the Short-toed Lark and its 

 allies. Dresser, who appears to have had access to a large series of A. brachydactyla, 

 recognizes but one species ; whereas of A. pispoletta his series appears to have been small, 

 and in the absence of intermediate forms, he has allowed specific rank to each of its,climatic 



varieties, 



T 3 



