-an interest in his meals-failed me. From what I remember of the bill of fare, tinned salmon and currant buns were the standing dishes at every repast. I was so far fortunate in having fellow-sufferers of an agreeable disposition, and in mutual sympathy we derived the means of alleviating our discomforts and awaited patiently our release at Aden.

After a few days at this unjustly vilified spot, where, perhaps, the kind hospitality of General and Mrs. Blair caused me to view everything too much—not couleur de rose, that would be out of place in Aden, where a hot red tint preponderates-but couleur de verdure, I set out for Suez, and had an exceedingly rapid passage, in the British India S.S. Dacca, of under five days. Owing to the quarantine, however, this advantage was neutralized, as I just missed the overland mail, and had to resign myself to a week in Egypt, which was not unpleasantly spent in looking up old friends and revisiting old haunts.

Then travelling via Brindisi and Calais, I reached London on the 31st of December, 1884; in spite of all delays, not much more than six weeks after I last saw the snow-peaks of Kilima-njaro from the reedy shores. of Lake Jipé.

Thus ends the history of my expedition to the snow mountain of Eastern Equatorial Africa.

A sketch of the results of my researches and observations in Natural History, Anthropology, Languages, &c., will be given in the following chapters.

Y

CHAPTER XVII.

CLIMATE, GEOLOGY, BOTANY, ETC.

ELSEWHERE in this work I have briefly pointed out the main features in the physical geography of the Kilimanjaro region, the dry, almost uninhabited plains, covered with coarse yellow grass and stunted trees, save where perennial streams induce rich vegetation; the misty snow-capped mountains, where the damp climate causes a verdure and a freshness reminding us of the fairest portions of our own island; the majestic forests that clothe certain districts like Taveita, or, not far distant, the broad flats white with natron-salt, the beds of dried-up lakes. With a country offering such very diverse features there are naturally wide ranges of temperature and an infinite variety of climate. Two days' climb on Kilima-njaro will transfer you from equatorial heat to Arctic cold. You may change your surroundings in this short period from the tropical vegetation of Taveita to a lifeless wilderness of ice, rocks, and snow. Midway up the mountain there are lovely regions, mild, equable, and moist, with a climate resembling a Devonshire summer. Here the intense verdure and the luxuriance of fern life testify to the constant showers of gentle rain. A few thousand feet below, in the salt plains, you may be parched and scorched by the hot desert winds, recalling-though

in a much modified degree-the sirocco or khamsin of Northern Africa. In the forests of Taveita the (to me, not disagreeable) unvarying, moist, greenhouse warmth of the tropical lowlands awaits you. Here, perhaps, the utmost range of the thermometer during the twenty-four hours will be ten or twelve degrees. In the outer plains it may vary from 91° in the early afternoon to 58° before dawn.

During the four months (June-September) of my stay in Moši, at an altitude of 5000 feet, the average readings in the shade of the thermometer were as follows: At noon, 71°; at 8 p.m., 60°; at 6 a.m., 58°. The lowest reading at noon was 68° (August 12th); the highest 80° (July 3rd and June 12th). The lowest reading at night (from 10 to 12 p.m.) was 54°; the highest 62. In my more elevated collecting stations, at 10,000 and 11,000 feet, the lowest night temperature (recorded) was 29°; the highest at 3 p.m., 65°. At Taveita the highest registered was 90° at 3 p.m.; the lowest at 4 a.m., 60°. The highest temperature ever recorded in the plains between Kilima-njaro and the coast was 91°.

From these observations the reader will perceive that, for at least six months of the year, the temperature of this portion of Eastern Africa is quite bearable, if not in some places delightful. Nowhere is there extreme heat, and the nights are always so far cool, that bedclothes are supportable and sleep is refreshing. What it is for the other half of the year I cannot personally say, as I have had no experience, but travellers and missionaries who know the adjacent countries assure me that the temperature is but little higher during the other months when the wet season prevails; it is rather that the climate becomes every

where more equable, and the thermometer does not descend so relatively low at night.

The seasons that prevail over this district are also influenced by local surroundings. Close to the coast the rains commence in October, intermit from December to March, and return in all their force during April and May. The true "dry season" is from June till October, during which not a drop of rain falls, though the sky is often clouded. Between the coast country and the elevated plateau about 100 miles inland there is a district which is much stinted in its rainfall, except where unusually high mountains arrest the moisture-laden winds from the Indian Ocean. There we arrive at the most general type of African scenery-wide, rolling savannahs, covered with grass and scattered clumps of trees. These extend, here and there broken by mountain ranges, to the shores of the Victoria Nyanza, and over all this district the rainfall is generally limited to the months of November, December, March, April, and May. On the western shores of the Nyanza Lake rain falls during ten months of the year, a contrast to the regions lying in the same latitudes between the Victoria Nyanza and the Indian Ocean. In all lofty plateaux or high mountain ranges the rainfall is very different to what it is in the plains below. Thus in Kilima-njaro it rains more or less throughout the year. My residence there was during the dry season, yet, nevertheless, in the month of June I had to record six days of rain; in July, 8; in August, 9; in September, 7; in October, 8; and in the first half of November, 5. The real rainy season is from November to May, but the Wa-čaga inform me that there is seldom at any season a continuous heavy down

pour, but rather a greater or less frequency of showers.

In the plains, however, near the mountain, the fall of rain is torrential, and is equal in violence to what one usually meets with in the tropics.

While snow is never absent from either of the twin summits of Kilima-njaro, and, indeed, at all times covers the upper part of the dome of Kibô with a mantle of unvarying white, yet the quantity and downward extent of the snows vary almost daily, even in the dry season. After a rainy night in the lowlands, the snow on the following morning may be seen on Kibô down to a level of 14,000 feet, and even a little lower on the western slope, while the whole of craggy Kimawenzi is a pinnacle of scintillating whiteness, like to use a very hackneyed simile-a sugarloaf. Yet if the succeeding day be warm and sunny, the snow on Kimawenzi may shrink in twenty-four hours to a tiny patch and a streak in between the jagged walls of black rock, while on Kibô it will withdraw its inroads 1000 feet above the level of the day before. On the whole, the least snow observable is during the months of July and August. In October there is a great deal. I should think the most snow fell, from what I hear by native report, during February and March, but at this time, nevertheless, the natives maintain that an ascent of the mountain is easiest, as the mists are not so frequent. Neither is the cold so great, curiously enough.

Snow is reported occasionally to fall on the summit of Mount Meru, the pyramidal peak lying to the southwest of Kilima-njaro, and on its western slope to lie for some months. The western slopes of Kibo and Kimawenzi are much more snow-covered