inherited aptitude for the work, which they do with a will, and with pride in the growth of individual trees. One casuarina of a year old, having shot up to the unusual height of 2 metres, the men have named "Kitchener Pacha." The others of this nursery of the same age are rather less than 1 metre; all were grown from seed in practically pure sand. The lines of trees are as regular and straight as in any English nursery. Near by I saw a casuarina which had been for six months standing in a bucket of water, yet appeared perfectly healthy, which

shows how well suited the tree is to a district subject to periodic inundation. Ground-nuts planted at foot of the casuarina serve to fix nitrogen, to shade the ground, and as an index of proper watering. They are also worth something for oil. Near by, the casuarina were being planted away from the canal, entering on their proper work in the desert itself. The method of planting is as follows. An iron cylinder, 14 cm. diameter and 60 cm. long, is sunk in the sand. The workman scoops out the sand (they can burrow in the material with surprising facility) and fills the cylinder with a good light mould.

The cylinder is then removed, and the tree is planted in the mould, so that it has nourishment provided until the roots have gone down 60 cm., when the damp sand is reached. A man can plant sixty casuarina per diem in this way. The wage is one franc a day. Fig. 17 shows a fine plant of Agave Rigida Sisalana further out in the desert. Being more distinguished than his fellows, the men have called him "Sheikh-as-Saadat." The figure in the photograph is that of Mr. Floyer. The "Sheikh has been two years in the desert without attention, and has nine well-grown suckers. Other plants near by have eight, nine, twelve, and thirteen suckers. They have flourished exceedingly without watering, and Mr. Floyer expects that they will rapidly cover the neighbouring desert. The fibre, I am told, is worth £34 a

ton, and the manufacture is extremely simple. The casuarina also has considerable value in a country where even firewood is imported to the value, I am told, of £30,000 per annum.

On May 3 Mr. Floyer and I left the rest-house (north of Menashi) at 6 a.m., and boarded the train for Terieh. On the right hand I could see islands of blown sand among the cultivated lands of the delta. After Katatbeh we passed through a belt of land which is still cultivated, showing what can be done about here. An old ruin near by supplies a nitrogenous manure. There is a bank of sand on the west of the line at the railway-station, Terieh, produced by a reed fence placed to check the sand-drift. Fig. 18 is a microphotograph of a portion of sample 21, taken on the lee side, and near the summit of this artificial dune. The want of uniformity of size of grains should be

noticed.

The artificial dune was round-topped; there was no very

active eddy here to perfect the work of sorting.

there is coarse sand.

At Terieh I noticed that black earth is brought up in well-making from below the sand. Below the black earth, Mr. Floyer informs me, These facts point to former cultivation. Fig. 19 is a photograph (looking north) of a cemetery wall 6 feet high, at Terieh, banked up by drifting sand on both sides. The top of the wall keeps free of sand, which is usual, but worthy of remark. Sample 22 is from the eastern side of the wall.

Sample 24 is from the sandy plain, taken between the cemetery and the village.

After returning by train to the rest-house near Menashi, I took sample 25 from the top of a small dune, where grass had prevented rippling. I returned the same day to Helwan.

On May 4, half a mile north of the town of Helwan, between the railway and the quarries, I collected sample 26 by scraping from the surface of the ground within the space of a few square inches. I noted that there appeared to be round about a sufficiency of hard sand-sized grains to form dunes, if only they had formed a sufficiently large proportion of the material in which they occur. Some scrapings from the top layer of the crest of ripples appeared to contain quartz grains, with much fine, soft stuff. Taking a handful of the pulverulent ground,

I threw it into the air: stones dropped down straight to the ground, and a copious cloud of dust, best seen by its shadow, floated away in the northerly breeze. The range of size of the material is too great, and the proportion of sand-sized particles too small, for the ready formation of dunes. Their absence is not due to lack of wind-action, as is well illustrated by the wind-cut tables hard by, one of which (height about 4 feet 6 inches) is shown in Fig. 20. The action of the wind upon the rocks here produces stones and dust; sand-sized particles can also be found. Is the almost complete absence of sand-dunes connected with the too-rapid formation of dust from stones, and the necessary consequence that sand-sized particles are kept in a hopeless minority?

On May 7 I crossed the Nile to Mariette's House, in order to see the country near the pyramids of Sakhara. Here (west of the Nile) are rolling hills of hard gravel, similar to those which border the plain west of the country from Menashi to Terieh. Lying in the more sheltered spots is a scattering of golden sand, which gives the country a tint very different from that of the desert on the east of the Nile, near Helwan. In this connection I quote the following passage from a letter by Mr. Floyer (March 9, 1899), in which he says: "[Riding down the east bank, between Halfa and Assuan], it was easy to see the hills on the west bank covered with golden sand pouring into the Nile. There is on the east none of this golden sand at all."

In crossing the Nile this day I found on the west side a sandy

foreland exposed by the subsiding waters of the river, which were then near their lowest level, on which, within the space of a few hundred acres, I saw a greater variety of forms of dunes than I met with during the rest of my tour. They were small-the highest not more than 4 feet— but perfect dunes; by no possibility to be confused with ripples. Their daily observation throughout a fall and rise of the Nile would have been an ideal exercise in the study of sand-dunes. Even the three visits I was able to pay to the sandy foreland yielded results which I believe to be of considerable interest.

*

Fig. 21 (Plate I.) is looking up-wind (N.N.E.). The steep lee cliffs stand out strongly. It will be noticed that the dunes are in transverse ridges, but that these ridges undulate. The view looking down-wind, the lee cliffs not being visible, showed no such striking contrast; indeed, the photograph is so faint that I have not had it reproduced. The sand is Nile sand (sample 27), fine, glittering, micaceous, splintery; quite unlike the rounded quartz sand of the samples I collected in the desert. During most of the time I spent on the foreland the sand was flying briskly. Fig. 22 (Plate I.) shows longitudinal ridges piled up against lee cliffs. The structure is readily understood when one watches the sand sweeping round from both ends of the cliff. I suppose it to be similar to a structure described by Dr. Blandford: "From the north-east corner of most of the high hills near Bálmir a long ridge of sand runs out, evidently developed by the wind under the lee of the hill." It is not difficult to see that if the wind should continue sweeping out the sand from the depressions between the higher portions of the transverse ridges shown in the photograph, and depositing a part of it thus behind the highest part of the cliff, the longitudinal ridge thus formed may play an important part in a conversion of transverse into longitudinal dunes. I first saw these dunes in the morning; on my return in the afternoon the wind had increased, and there was a haze of flying sand 20 or 30 feet high over the foreland. In many places, where in the morning there had been loose sand, the wind had now cut down to the damp, compact floor, and the ripples appeared distinctly larger. Fig. 23 (Plate I.) shows a small dune, about 2 feet high, produced by a reed fence. The crest of the cliff is parallel to the fence, although, as shown by the surface-markings on the sand, the fence is not at right angles to the wind. Fig. 24 (Plate II.) is a perfect example of one of those singular structures known as fuljes in Arabia.t This was only 30 yards from the river, off which the wind was blowing. The photograph was taken looking down-wind at 5.15 p.m. The circular floor of compact material is about 12 feet in diameter. On the left there is a double ridge. I was able to watch the process by which this was being formed, sand coming to it from the left

* "On the Physical Geography of the Great Indian Desert," Journ. Asiatic Soc. of Bengal, vol. 45 (1876), p. 98.

+ Geographical Journal, March, 1897: "Sand-dunes."

!