Butterflies of Southern Kordofan. 271 fairly extensively, and gets a considerable crop of red millet. He also grows a little tobacco. . . . These khors sometimes fill with water and remain full for a long time, and even if they do not, water can generally be obtained by digging in the bed of them.

" The country is alternate ' cotton-soil ' (black stuff which cotton does not grow in, but in appearance like the cotton-soil of Eygpt) and a reddish soil which is probably the same black soil mixed with sand, etc."

The official guide-book cited above states that the cotton-soil plains contain too high a percentage of clay to be easily worked and are usually covered with forest, but that the slopes of the Dar Nuba hills, the banks of the khors, etc., are eminently suitable for cultivation, with a rich soil capable of bearing good crops of dura and vegetables (p. 25).

Again, to quote Mr. Willis' letter : " The hills stand straight up out of the plain, rising abruptly; sometimes there is a considerable extent of plateau high up where soil can be cultivated, but the appearance of the hills is extraordinarily rocky, though trees grow upon them."

This description of the hills vividly recalls to my mind what I saw at C4ebel En, and more particularly at Rejaf. The mountains there suggest Nasmyth's description of the mountains of the moon, which he compared to pellets of clay thrown by boys so as to stick upon a wall or fence. The official guide shows that these abrupt hills may be anything from 500 ft. to, in exceptional cases, nearly 2000 ft. above the plain, i. e. 4000 ft. above the sea. Mr. Willis goes on to make the sound suggestion : " It is most probable that the butterflies were captured close to water. It is extraordinary to see them collect on any little damp patch of ground which lies in the shade, and there are always cruantities by the watering-places, when the ground is clear of grass and the water has beeu spilled into troughs or the like for watering animals."

I have myself taken butterflies under like circumstances by the watering-places of baggage mules at Malakand (N.W. frontier of India), but I suspect that the Btaling of the animals has something to do with it. " Each side of a khor is heavily wooded and covered w it h thick vegetation, the depth, of course, varying, and it is in that little belt that all the beasts or butterflies lie at some period of the d