( 490 ) 



Makinir a loni,' story (of shootiog, skinning, carefnlly comparing and 

 measuring all onr material) short, we will state that in fart tho two forms 

 scpariiteil by Tristram in lf>oO are very closely allied, and only dilfer in having 



in average slightly different dimensions, as well as a very slightly different 



shade of c.olonr, the hirds from Laghonat to (ihardai'a, which chiellv inhabit the 

 neighbourhood of the dayats and cornfields, being generallv a little larger and 

 sometimes darker, those found from the plain of El Ontaya to Tonggonrt, which 

 iidnxbit principally nncnitivated sandy soil and often real sand-desert (thongh not 

 as a rnle the bare sand-dunes of the Souf), being as a rnle a little smaller and 

 a shade jialer, when series for series are eom]iared. The differences in the size 

 of the bill are very slight and not constant, thongh G. c. macrorhjncha have often 

 larger bills. 



When one of us wrote aboat the Oested Larks l)efore, he thonght the differences 

 between these two forms were much greater, or he would perhaps not have 

 separated them from the scanty material then available. The reason for this is 

 that the $ type of Tristram's G. macrorhyncha is an unusually large specimen 

 with an excej)tionally large bill, and that his specimens have become darker, 

 through being more or less soiled and kept in insutlieiently tiglit-fitting cabinets 

 for more than half a century. Had Canon Tristram collected twenty of eacli of 

 these forms instead of two of one and three of the other, he would probably not 

 have separated them. 



The i type of G. macrorln/nclia has a wing of ll>io, a culuien of 2(3, and a 

 tail of 75 mm.; the ? wing IlO, culmen 23, tail ')S mm. Our lo males have tlie 

 wings 110 5 to 116-5, and culmina up to 25, our 8 females wings of l<fj to lOo'o mm. 

 Our series of 70 males and females of G. crisfata arenicola measure : J wings 

 li):5-112, once 113 (a male from south of Tonggonrt, shot together with a fairly 

 small female), ? 100-104 mm. The cnlmina reach a ma.Kimum of 24, but are 

 generally shorter. 



While the Algerian and Tunisian forms of Galerkla criatata are comparatively 

 easy to understand (their range of variation being much smaller), those of the 

 species G. tlirhlae are extremely difficult, because their variation is astounding. 

 Having examined (Mr. Rothschild, Ernst Hartert and Carl Hilgert, during a 

 visit of the latter to Tring) 132 Algerian and Tunisian specimens, belonging to 

 the Collections of Tring and Ingelheim, of the pale races {i.e. leaving out the 

 dark northern form called Iiftrtt'iti), we have made the following observations : 



The series from Central Tunisia — i.e. from south of the Atlas range, about 

 Gafsa, from the Djebel Batoum, Djebel 8idi Aich, Djebel Tfel, Djebel Sidi ben 

 Aoun, the Seggi, Bou Hamran, Bir Mrabont, and thereabouts— vary from an 

 ujiperside of a pale sandy rufous tinge with cinnamon-rufous spots, 'to a pale buii 

 sand-colour with blackish lirown spots, while the majority are of a jiale greyish 

 sandy coloration between the two extremes. Although the extremes are strikingly 

 different, they are connected by intermediate ones, and there can be no doubt that 

 all these birds, 30 in number, belong to one form. The soil here is more or 

 less rocky or stony, interspersed with oases and fertile ground, and with more 

 or less of the scanty vegetation peculiar to such districts. 



The birds from these districts were named Galerida pallida by Mr. Whitaker, 

 and I introduced the name super/Inn for them, liecause />"///'/" had been anticii)ated 

 by Brehm. 



If we compare with these birds the ten skins collected by the late Baron 



