( 487) 



and on this da)' as well as five days Inter near Biskra three and four were killed 

 with one shot. 



Early in Ajiril tliey were seen, evidently in pairs, near Laghonat, ami iilioiit the 

 middle of the niontli near one of the Dayats not far from Tilrhempt, where it was 

 apparently hrei-dint;, thonj,'h they diil not yet seem to have egirs. 



:^3. Ammomanes deserti algeriensis iSliarpe. 



Tlie distrilmtion of the varions forms of Ammomanes deserli. is not yet r|nite 

 certain, especially that of .1. deserti deserti and A. deserti isahelliii'i. With regard 

 to the Algerian form, however, it is quite certain that it is the only one inhabiting 

 Algeria and Tunisia as far as they are explored, and that it is very constant in these 

 countries. It appears to extend throughout the Sahara as far east as the Libyan 

 desert, where it reaches the Nile. 



The Algerian commoji Desert-lark inhabits stony desert and rocky plateaux 

 and slopes in the sonthi-rn Atlas mountains, where it occurs frennently just north 

 and south of El Kantara. Near Biskra it is everywhere to be met with where the 

 gronnd is stony, and on the slopes of the rocky ridges in the desert, bnt not on the 

 top of them. Rather to our surjirise we also found it a good way south of Biskra, 

 near Bord ('hegga and Kef-el-Dor, in jilaces where the ground is very stony, and 

 even patches of bare rock appear. It is absent from the regions of the sand-dunes 

 and from the sebcha-depressions. 



We found dutches near Biskra, April 27, 1909, and on April 3U and May 3 near 

 El Kanfara. The nests have cnrinns po.sitions. One was placed under a flat stone, 

 so that it was just protected from above ; the nest was a thick soft padded structure, 

 and consisted of the whitish bnff flowers of a desert plant (? Gn'iplialium), and short 

 j)ieces of plant-stems. The eggs were hard set, April 27, 1909. A second nest was 

 quite similar to the last one. A third was nnder a small bush, and by its side was 

 a flat stone. A fourth was also nnder a small bush, but instead of a large stone it 

 was on the two outward sides surrounded by a four-inch- wide wall of little stones, 

 evidently collected by the birds. It seems that the large stone is some sort of pro- 

 tection, and the birds not finding one heap np a wall of little stones themselves. 

 The clntches number three or four. The eggs are milk-white, almost without 

 gloss, with brownish grey spots, and a few underlying pale grey markings. In 

 another clutch the markings are more frequent and much smaller, mere dots, and 

 the shell is more glossy. Another has the rather glossy ground-colour cream, with 

 a faint pink tinge, and the spots pale reddish brown and mauve. Measurements 

 22-0 X 15-8, 22 x 16, 225 x loS, and 229 x 15-5 ; 22-7 x 10-9, 23-4 x l<;-3, and 

 23 X 10-5 ; 211 x 16,20-4 x 151, aud 20(5 x 15-7 mm. 



In 1911 we found these Larks near Laghonat, on stony hills, and very common 

 in the M'zab valley, near the town of Ghardaia. Eggs were frequently found during 

 the second half of Ajiril. 



34. Ammomanes phoenicurus arenicolor (Sund.) 



At first sight the Bar-tailed Desert-lark appears to be very similar to A. deserti, 

 bnt a closer examination shows, in addition to the black tips to the rectrice^ that 

 the wing is of a very dift'erent sha[)e ; the inner secondaries are considerably long<'r 

 than tlie outer ones, and the secoml (first long) jirimary is about as long as the 

 third, or only a, millimetre or so shorter, while in A. ili'serti it is considerably 

 shorter. 



