( 14 ) 



water-holes in the sainl of Uio river-bed, ,i,'eiieriUly i)i>en, but Koraetimcs covered np 

 with ii i\-dt stone, or filled up with sand; when the saud is dug away for a foot 

 or two the water soon fills iij) the hole, and thus, by and by, some amount of liijnid 

 is gained, and it is qnite drinkable, being well filtered through the sand. During 

 the six days in the Oiied Mya region we tamped twice on a " til mas "' with water, 

 and we passed anotlier on the way ; but it would probably not be dilficnlt to sink- 

 wells in other jjlaees along the river-beds. 



Now, during the second week of April, migrants passed through in great 

 (luantities. ^Ve observed and partly collected : Ijixciiiio mc<iiirhiii)cha mi'(/urln/iiclia 

 (once), several Oenantlie oemutthe oeiiantlu; and one 0. o. Icacorrlioa, <>. /i/s/Mi/iicri, 

 Saxicola {P latincola auct.) rnhetra and riihkola, Pliylloxcoijus coll;/hita and 

 trorhiliis, l'hnciiicnra>i plioenicKrm, Si/lein currnca and cantilliinx, S. coiixiiirillulfi 

 (rare), MotariUn. JIam j/arn, Antlais tricialix and camjjr.i/r/s, Calandrclla bincln/- 

 davtijla (two flocks), LaniuH xencUor (common), i'kdiilon nustica, lliranilo urhiai, 

 h'/pana riparia, Vpiipa epopx (a few), a flock of Mcrops npiand'r, a single (TJnail. 

 We also found numerous jjcllets and the remains of a Scops-owl {Dins xcopx), 

 showing that these birds must have passed thmngli quite recently. 



Of mammals we caught a lew smaller rodents, shot a new " gnndi " 

 {Mmisoutiera Imrferti), iuid saw many traces of gazelles, Barbary-sheep, jackals, 

 and — a]i]iarently — foxes. Hares were common, and we shot several. We also 

 found the remains of a hedgehog-skin. Of resident birds only Amni'nnam'x 

 dexnrti mifu and OfJiinitkc {Sa.virnla) Iciirop'/t/o were more or less common and 

 widespread, and two or three times single sjiecimens or paiis of Er>/tlirospiza 

 githaginm were observed. Fulco hiarmicus erlangeri nested on the cliffs, and 

 eggs were taken ; lliilco fciox cirteiixix Was observed twice, a young Xeojiliron, 

 Havens, and a single MiUux Itnrxchun were noticed, and on the way back a single 

 Athene noctim xahitrae. was shot. 



Very interesting moths were collected at night, but. some evenings were cold 

 and stormy, and once it rained for five minutes — huge drops, not enough to refresh 

 even the plants. The amount of verdure and flowers in the river-beds was due to 

 heavy rains having lUllen in December or .lanuary, floixling the rivers to a great 

 height, and making them impassable for caravans and post-riders. This state, 

 however, lasted only a few days, and the water had not even reached the neigh- 

 bourhood of Hassi Inifel (north-east of El-(iolea), where Jjieut. l\Iaire-8ebille 

 found no trace of rain and not a green leaf or Idade in May, while in former times 

 the waters must sometimes have rolled on northwards at least ti' south of Ouargla, 

 where they were lost in the sand. Holland (cf. his admirable works ou the 

 '• H.vdrologie " and " (ieologie du .Sahara Algerien ") was probably the first to 

 regard the Oued Khir (extending from Touggourt to the L'hott Melrhir) as a 

 continuation of the Oued Mya, which would then have been united with the Chott 

 Melrhir by way of the Sebcha of Ouargla, the Sebcha Safroun, north of Ngoussa, 

 the Chott Bardad, and the (Uiotts of Bledet-Ahmar, and Temassiu. This theory 

 has generally been accei)ted ; but it is, to me, difficult to imaginj that enough rain 

 used to fall on the j)lateau of Tademait to send the waters n|) to Touggourt and 

 farther, and the Oued -Mya certainly drains only the Tademait. On the other hand, 

 the Oued Igharghar comes from the massive of the Hoggar Mountains, and its 

 waters were, 1 should think, much greater and more likely to extend to the (Jhott 

 Melrhir, so that the Oued Hhir appears to be more likely a continuation of the 

 Oued Igharghar than of the Oued Jlya, which only drains the plateau of Tademait. 



