(488) 



The distribution is very interesting. While the species is absent from North 

 Aig:eria and the Atlas, and even tVoiu the neighlionrluiod of Biskra, it appears 

 suddenly on the stony desert-plain south of P.ordj Saada, between 30 and 35 kilo- 

 metres sonth of Biskra. Tliis stony plain is partly sandy, and it is on that kind of 

 soil only where we found it — stony stretches intermixed with sandy tracts ; but it is 

 absent from the regions of the sand-dunes proper. There is quite au amount of 

 variation in these birds, some being more greyish, others more isabelline, others 

 again more reddish ; but none approach A. phooiicura clnctiii-a from the Cape Verde 

 Islands or any of the Asiatic representatives. 



Besides our series collected between Bordj Saada (3U km. sonth of Biskra) and 

 Touggonrt, we have skins from the Natron Valley near Cairo, from Kerma and 

 Shendi in Nubia (N. ('. Rothschild), and a pair collected by ('omte Dalmas near 

 Cape Blanco (" Bale dn Levrier") on May lu, 1895. As far as one can judge from 

 their very worn plumage, the latter do not differ from A. p. arenicolor. Iris dark 

 brown ; feet pale l)rownish flesh, toes greyish or whitish ; bill pale horn-colour, 

 tiji darker. 



In 1911 we found this little Lark by no means rare from a few miles soutii of 

 Laghonat to the region of the Dayats, in places where the desert was not too densely 

 covered with stones, but where a few bushes and partly sandy or clayey soil existed. 

 We also came across it on the stony elevated plateau east of Gharduia, where 

 Ammomuiies desert i ulgerienais was also common, though the latter was more partial 

 to the rocky hills. 



The Crested Larks. 



('rested Larks are the horror of '"lumpers,"' who do not care to trouble about 

 closely allied forms, and the delight of those who study the geographical rejire- 

 sentatives of species! Moreover the Crested Larks occur in many j)arts of Sonth 

 Europe and North Africa in two separate species, each of which shows a similar 

 variation in close connection with the soil and surroundings of the areas inhabited 

 by them. 



Of the long-billed species, Galerkla critstatu, we did not find a representative 

 in North Algeria, but only found forms of G. cristata sonth of El Kantara, from 

 Laghouat southwards to Ghardaia, and throughout the country traversed from 

 Biskra to Temacin, south of Touggourt; while of the other species, (ialerida 

 theklae, we collected four forms — a very dark one in North Algeria, a paler one in 

 the southern parts of the Atlas range down to Biskra, a very pale desert form in 

 the desert, and a red form from Laghouat to Ghardaia. 



But, although we failed to find them, a darker long-billed race apjiears to exi.st 

 after all in Algeria. One of us united with G. c. macrorhyncha (1859) Loche's 

 Galerula raticlonii (described 1800); the exact locality of which was not known. There 

 is, however, in the Museum at Milan, which contains the greater part of the Loche 

 collection, a ('rested Lark named G. ramlonii, from Ain Oussera, between Djelfa 

 and Laghouat. This is the region of the " Hants Plateaux," with their sea of halfa- 

 grass. In this region we saw only a form of G. theklae, though specially looking 

 out for long-billed forms, but we passed through on a motor-car and did not proj)erly 

 collect in tliese districts. As the t/w/dae-iovm of this country is Itilyerti and not 

 carolinae, it is quite possible — though not necessary — that the cristata-iorm also 

 differs, and the Milan specimen is indeed much darker than specimens of macro- 

 rhyncha in the same museum, and we have a specimen from the Riocour-collection, 

 which is also darker. Evidently the Milan specimen is the type of randonii, 



