If deserts _have_ a fault (which their present biographer is far from admitting), that fault may doubtless be found in the fact that their scenery as a rule tends to be just a trifle monotonous. Though fine in themselves, they lack variety. To be sure, very few of the deserts of real life possess that absolute flatness, sandiness and sameness, which characterises the familiar desert of the poet and of the annual exhibitions--a desert all level yellow expanse, most bilious in its colouring, and relieved by but four allowable academy properties, a palm-tree, a camel, a sphinx, and a pyramid. For foreground, throw in a sheikh in appropriate drapery; for background, a sky-line and a bleaching skeleton; stir and mix, and your picture is finished. Most practical deserts one comes across in travelling, however, are a great deal less simple and theatrical than that; rock preponderates over sand in their composition, and inequalities of surface are often the rule rather than the exception. There is reason to believe, indeed, that the artistic conception of the common or Burlington House desert has been unduly influenced for evil by the accessibility and the poetic adjuncts of the Egyptian sand-waste, which, being situated in a great alluvial river valley is really flat, and, being the most familiar, has therefore distorted to its own shape the mental picture of all its kind elsewhere. But most deserts of actual nature are not all flat, nor all sandy; they present a considerable diversity and variety of surface, and their rocks are often unpleasantly obtrusive to the tender feet of the pedestrian traveller.

A desert, in fact, is only a place where the weather is always and uniformly fine. The sand is there merely as what the logicians call, in their cheerful way, 'a separable accident'; the essential of a desert, as such, is the absence of vegetation, due to drought. The barometer in those happy, too happy, regions, always stands at Set Fair. At least, it would, if barometers commonly grew in the desert, where, however, in the present condition of science, they are rarely found. It is this dryness of the air, and this alone, that makes a desert; all the rest, like the camels, the sphinx, the skeleton, and the pyramid, is only thrown in to complete the picture.

Now the first question that occurs to the inquiring mind--which is but a graceful periphrasis for the present writer--when it comes to examine in detail the peculiarities of deserts is just this: Why are there places on the earth's surface on which rain never falls? What makes it so uncommonly dry in Sahara when it's so unpleasantly wet and so unnecessarily foggy in this realm of England? And the obvious answer is, of course, that deserts exist only in those parts of the world where the run of mountain ranges, prevalent winds, and ocean currents conspire to render the average rainfall as small as possible. But, strangely enough, there is a large irregular belt of the great eastern continent where these peculiar conditions occur in an almost unbroken line for thousands of miles together, from the west coast of Africa to the borders of China: and it is in this belt that all the best known deserts of the world are actually situated. In one place it is the Atlas and the Kong mountains (now don't pretend, as David Copperfield's aunt would have said, you don't know the Kong mountains); at another place it is the Arabian coast range, Lebanon, and the Beluchi hills; at a third, it is the Himalayas and the Chinese heights that intercept and precipitate all the moisture from the clouds. But, from whatever variety of local causes it may arise, the fact still remains the same, that all the great deserts run in this long, almost unbroken series, beginning with the greater and the smaller Sahara, continuing in the Libyan and Egyptian desert, spreading on through the larger part of Arabia, reappearing to the north as the Syrian desert, and to the east as the desert of Rajputana (the Great Indian Desert of the Anglo-Indian mind), while further east again the long line terminates in the desert of Gobi on the Chinese frontier.

In other parts of the world, deserts are less frequent. The peculiar combination of circumstances which goes to produce them does not elsewhere occur over any vast area, on so large a scale. Still, there is one region in western America where the necessary conditions are found to perfection. The high snow-clad peaks of the Rocky Mountains on the one side check and condense all the moisture that comes from the Atlantic; the Sierra Nevada and the Wahsatch range on the other, running parallel with them to the west, check and condense all the moisture that comes from the Pacific coast. In between these two great lines lies the dry and almost rainless district known to the ambitious western mind as the Great American Desert, enclosing in its midst that slowly evaporating inland sea, the Great Salt Lake, a last relic of some extinct chain of mighty waters once comparable to Superior, Erie, and Ontario. In Mexico, again, where the twin ranges draw closer together, desert conditions once more supervene. But it is in central Australia that the causes which lead to the desert state are, perhaps on the whole, best exemplified. There, ranges of high mountains extend almost all round the coasts, and so completely intercept the rainfall which ought to fertilise the great central plain that the rivers are almost all short and local, and one thirsty waste spreads for miles and miles together over the whole unexplored interior of the continent.

But why are deserts rocky and sandy? Why aren't they covered, like the rest of the world, with earth, soil, mould, or dust? One can see plainly enough why there should be little or no vegetation where no rain falls, but one can't see quite so easily why there should be only sand and rock instead of arid clay-field.

Well, the answer is that without vegetation there is no such thing as soil on earth anywhere. The top layer of the land in all ordinary and well-behaved countries is composed entirely of vegetable mould, the decaying remains of innumerable generations of weeds and grasses. Earth to earth is the rule of nature. Soil, in fact, consists entirely of dead leaves. And where there are no leaves to die and decay, there can be no mould or soil to speak of. Darwin showed, indeed, in his last great book, that we owe the whole earthy covering of our hills and plains almost entirely to the perennial exertions of that friend of the farmers, the harmless, necessary earthworm. Year after year the silent worker is busy every night pulling down leaves through his tunnelled burrow into his underground nest, and there converting them by means of his castings into the black mould which produces, in the end, for lordly man, all his cultivable fields and pasture-lands and meadows. Where there are no leaves and no earth-worms, therefore, there can be no soil; and under those circumstances we get what we familiarly know as a desert.

The normal course of events where new land rises above the sea is something like this, as oceanic isles have sufficiently demonstrated. The rock when it first emerges from the water rises bare and rugged like a sea-cliff; no living thing, animal or vegetable, is harboured anywhere on its naked surface. In time, however, as rain falls upon its jutting peaks and barren pinnacles, disintegration sets in, or, to speak plainer English, the rock crumbles; and soon streams wash down tiny deposits of sand and mud thus produced into the valleys and hollows of the upheaved area. At the same time lichens begin to spring in yellow patches upon the bare face of the rock, and feathery ferns, whose spores have been wafted by the wind, or carried by the waves, or borne on the feet of unconscious birds, sprout here and there from the clefts and crannies. These, as they die and decay, in turn form a thin layer of vegetable mould, the first beginning of a local soil, in which the trusty earthworm (imported in the egg on driftwood or floating weeds) straightway sets to work to burrow, and which he rapidly increases by his constant labour. On the soil thus deposited, flowering plants and trees can soon root themselves, as fast as seeds, nuts or fruits are wafted to the island by various accidents from surrounding countries. The new land thrown up by the great eruption of Krakatoa has in this way already clothed itself from head to foot with a luxuriant sheet of ferns, mosses, and other vegetation.

First soil, then plant and animal life, are thus in the last resort wholly dependent for their existence on the amount of rainfall. But in deserts, where rain seldom or never falls (except by accident) the first term in this series is altogether wanting. There can be no rivers, brooks or streams to wash down beds of alluvial deposit from the mountains to the valleys. Denudation (the term, though rather awful, is not an improper one) must therefore take a different turn. Practically speaking, there is no water action; the work is all done by sun and wind. Under these circumstances, the rocks crumble away very slowly by mere exposure into small fragments, which the wind knocks off and blows about the surface, forming sand or dust of them in all convenient hollows. The frequent currents, produced by the heated air that lies upon the basking layer of sand, continually keep the surface agitated, and so blow about the sand and grind one piece against the other till it becomes ever finer and finer. Thus for the most part the hollows or valleys of deserts are filled by plains of bare sand, while their higher portions consist rather of barren, rocky mountains or table-land.

The effect upon whatever animal or vegetable life can manage here and there to survive under such circumstances is very peculiar. Deserts are the most exacting of all known environments, and they compel their inhabitants with profound imperiousness to knuckle under to their prejudices and preconceptions in ten thousand particulars.

To begin with, all the smaller denizens of the desert--whether butterflies, beetles, birds, or lizards--must be quite uniformly isabelline or sand-coloured. This universal determination of the desert-haunting creatures to fall in with the fashion and to harmonise with their surroundings adds considerably to the painfully monotonous effect of desert scenery. A green plant, a blue butterfly, a red and yellow bird, a black or bronze-coloured beetle or lizard would improve the artistic aspect of the desert not a little. But no; the animals will hear nothing of such gaudy hues; with Quaker uniformity they will clothe themselves in dove-colour; they will all wear a sandy pepper-and-salt with as great unanimity as the ladies of the Court (on receipt of orders) wear Court mourning for the late lamented King of the Tongataboo Islands.

In reality, this universal sombre tint of desert animals is a beautiful example of the imperious working of our modern _Deus ex machina_, natural selection. The more uniform in hue is the environment of any particular region, the more uniform in hue must be all its inhabitants. In the arctic snows, for example, we find this principle pushed to its furthest logical conclusion. There, everything is and must be white--hares, foxes, and ptarmigans alike; and the reason is obvious--there can be no exception. Any brown or black or reddish animal who ventured north would at once render himself unpleasantly conspicuous in the midst of the uniform arctic whiteness. If he were a brown hare, for example, the foxes and bears and birds of prey of the district would spot him at once on the white fields, and pounce down upon him forthwith on his first appearance. That hare would leave no similar descendants to continue the race of brown hares in arctic regions after him. Or, suppose, on the other hand, it were a brown fox who invaded the domain of eternal snow. All the hares and ptarmigans of his new district would behold him coming from afar and keep well out of his way, while he, poor creature, would never be able to spot them at all among the white snow-fields. He would starve for want of prey, at the very time when the white fox, his neighbour, was stealing unperceived with stealthy tread upon the hares and ptarmigans. In this way, from generation to generation of arctic animals, the blacker or browner have been constantly weeded out, and the greyer and whiter have been constantly encouraged, till now all arctic animals alike are as spotlessly snowy as the snow around them.

In the desert much the same causes operate, in a slightly different way, in favour of a general greyness or brownness as against pronounced shades of black, white, red, green, or yellow. Desert animals, like intense South Kensington, go in only for neutral tints. In proportion as each individual approaches in hue to the sand about it will it succeed in life in avoiding its enemies or in creeping upon its prey, according to circumstances. In proportion as it presents a strikingly vivid or distinct appearance among the surrounding sand will it make itself a sure mark for its watchful foes, if it happen to be an unprotected skulker, or will it be seen beforehand and avoided by its prey, if it happen to be a predatory hunting or insect-eating beast. Hence on the sandy desert all species alike are uniformly sand-coloured. Spotty lizards bask on spotty sands, keeping a sharp look-out for spotty butterflies and spotty beetles, only to be themselves spotted and devoured in turn by equally spotty birds, or snakes, or tortoises. All nature seems to have gone into half-mourning together, or, converted by a passing Puritan missionary, to have clad itself incontinently in grey and fawn-colour.

Even the larger beasts that haunt the desert take their tone not a little from their sandy surroundings. You have only to compare the desert-haunting lion with the other great cats to see at once the reason for his peculiar uniform. The tigers and other tropical jungle-cats have their coats arranged in vertical stripes of black and yellow, which, though you would hardly believe it unless you saw them in their native nullahs (good word 'nullah,' gives a convincing Indian tone to a narrative of adventure), harmonise marvellously with the lights and shades of the bamboos and cane-brakes through whose depths the tiger moves so noiselessly.

Looking into the gloom of a tangled jungle, it is almost impossible to pick out the beast from the yellow stems and dark shadows in which it hides, save by the baleful gleam of those wicked eyes, catching the light for one second as they turn wistfully and bloodthirstily towards the approaching stranger. The jaguar, oncelot, leopard, and other tree-cats, on the other hand, are dappled or spotted--a type of coloration which exactly harmonises with the light and shade of the round sun-spots seen through the foliage of a tropical forest. They, too, are almost indistinguishable from the trees overhead as they creep along cautiously on the trunks and branches. But spots or stripes would at once betray the crouching lion among the bare rocks or desert sands; and therefore the lion is approximately sand-coloured. Seen in a cage at the Zoo, the British lion is a very conspicuous animal indeed; but spread at full length on a sandy patch or among bare yellow rocks under the Saharan sun, you may walk into his mouth before you are even aware of his august existence.

The three other great desert beasts of Asia or Africa--the ostrich, the giraffe, and the camel--are less protectively coloured, for various reasons. Giraffes and ostriches go in herds; they trust for safety mainly to their swiftness of foot, and, when driven to bay, like most gregarious animals, they make common cause against the ill-advised intruder. In such cases it is often well, for the sake of stragglers, that the herd should be readily distinguished at a distance; and it is to insure this advantage, I believe, that giraffes have acquired their strongly marked spots, as zebras have acquired their distinctive stripes, and hyaenas their similarly banded or dappled coats. One must always remember that disguise may be carried a trifle too far, and that recognisability in the parents often gives the young and giddy a point in their favour. For example, it seems certain that the general grey-brown tint of European rabbits serves to render them indistinguishable in a field of bracken, stubble, or dry grass. How hard it is, either for man or hawk, to pick out rabbits so long as they sit still, in an English meadow! But as soon as they begin to run towards their burrows the white patch by their tails inevitably betrays them; and this betrayal seems at first sight like a failure of adaptation. Certainly many a rabbit must be spotted and shot, or killed by birds of prey, solely on account of that tell-tale white patch as he makes for his shelter. Nevertheless, when we come to look closer, we can see, as Mr. Wallace acutely suggests, that the tell-tale patch has its function also. On the first alarm the parent rabbits take to their heels at once, and run at any untoward sight or sound toward the safety of the burrow. The white patch and the hoisted tail act as a danger-signal to the little bunnies, and direct them which way to escape the threatened misfortune. The young ones take the hint at once and follow their leader. Thus what may be sometimes a disadvantage to the individual animal becomes in the long run of incalculable benefit to the entire community.

It is interesting to note, too, how much alike in build and gait are these three thoroughbred desert roamers, the giraffe, the ostrich, and the camel or dromedary. In their long legs, their stalking march, their tall necks, and their ungainly appearance they all betoken their common adaptation to the needs and demands of a special environment. Since food is scarce and shelter rare, they have to run about much over large spaces in search of a livelihood or to escape their enemies. Then the burning nature of the sand as well as the need for speed compels them to have long legs which in turn necessitate equally long necks, if they are to reach the ground or the trees overhead for food and drink. Their feet have to be soft and padded to enable them to run over the sand with ease; and hard horny patches must protect their knees and all other portions of the body liable to touch the sweltering surface when they lie down to rest themselves. Finally, they can all endure thirst for long periods together; and the camel, the most inveterate desert-haunter of the trio, is even provided with a special stomach to take in water for several days at a stretch, besides having a peculiarly tough skin in which perspiration is reduced to a minimum. He carries his own water-supply internally, and wastes as little of it by the way as possible.

What the camel is among animals that is the cactus among plants--the most confirmed and specialised of desert-haunting organisms. It has been wholly developed in, by, and for the desert. I don't mean merely to say that cactuses resemble camels because they are clumsy, ungainly, awkward, and paradoxical; that would be a point of view almost as far beneath the dignity of science (which in spite of occasional lapses into the sin of levity I endeavour as a rule piously to uphold) as the old and fallacious reason 'because there's a B in both.' But cactuses, like camels, take in their water supply whenever they can get it, and never waste any of it on the way by needless evaporation. As they form the perfect central type of desert vegetation, and are also familiar plants to everyone, they may be taken as a good illustrative example of the effect that desert conditions inevitably produce upon vegetable evolution.

Quaint, shapeless, succulent, jointed, the cactuses look at first sight as if they were all leaves, and had no stem or trunk worth mentioning. Of course, therefore, the exact opposite is really the case; for, as a late lamented poet has assured us in mournful numbers, things (generally speaking) are not what they seem. The true truth about the cactuses runs just the other way; they are all stem and no leaves; what look like leaves being really joints of the trunk or branches, and the foliage being all dwarfed and stunted into the prickly hairs that dot and encumber the surface. All plants of very arid soils--for example, our common English stonecrops--tend to be thick, jointed, and succulent; the distinction between stem and leaves tends to disappear; and the whole weed, accustomed at times to long drought, acquires the habit of drinking in water greedily at its rootlets after every rain, and storing it away for future use in its thick, sponge-like, and water-tight tissues. To prevent undue evaporation, the surface also is covered with a thick, shiny skin--a sort of vegetable macintosh, which effectually checks all unnecessary transpiration. Of this desert type, then, the cactus is the furthest possible term. It has no flat leaves with expanded blades, to wither and die in the scorching desert air; but in their stead the thick and jointed stems do the same work--absorb carbon from the carbonic acid of the air, and store up water in the driest of seasons. Then, to repel the attacks of herbivores, who would gladly get at the juicy morsel if they could, the foliage has been turned into sharp defensive spines and prickles. The cactus is tenacious of life to a wonderful degree; and for reproduction it trusts not merely to its brilliant flowers, fertilised for the most part by desert moths or butterflies, and to its juicy fruit, of which the common prickly pear is a familiar instance, but it has the special property of springing afresh from any stray bit or fragment of the stem that happens to fall upon the dry ground anywhere.

True cactuses (in the native state) are confined to America; but the unhappy naturalist who ventures to say so in mixed society is sure to get sat upon (without due cause) by numberless people who have seen 'the cactus' wild all the world over. For one thing, the prickly pear and a few other common American species, have been naturalised and run wild throughout North Africa, the Mediterranean shores, and a great part of India, Arabia, and Persia. But what is more interesting and more confusing still, other desert plants which are _not_ cactuses, living in South Africa, Sind, Rajputana, and elsewhere unspecified, have been driven by the nature of their circumstances and the dryness of the soil to adopt precisely the same tactics, and therefore unconsciously to mimic or imitate the cactus tribe in the minutest details of their personal appearance. Most of these fallacious pseudo-cactuses are really spurges or euphorbias by family. They resemble the true Mexican type in externals only; that is to say, their stems are thick, jointed, and leaf-like, and they grow with clumsy and awkward angularity; but in the flower, fruit, seed, and in short in all structural peculiarities whatsoever, they differ utterly from the genuine cactus, and closely resemble all their spurge relations. Adaptive likenesses of this sort, due to mere stress of local conditions, have no more weight as indications of real relationship than the wings of the bat or the nippers of the seal, which don't make the one into a skylark, or the other into a mackerel.

In Sahara, on the other hand, the prevailing type of vegetation (wherever there is any) belongs to the kind playfully described by Sir Lambert Playfair as 'salsolaceous,' that is to say, in plainer English, it consists of plants like the glass-wort and the kali-weed, which are commonly burnt to make soda. These fleshy weeds resemble the cactuses in being succulent and thick-skinned but they differ from them in their curious ability to live upon very salt and soda-laden water. All through the great African desert region, in fact, most of the water is more or less brackish; 'bitter lakes' are common, and gypsum often covers the ground over immense areas. These districts occupy the beds of vast ancient lakes, now almost dry, of which the existing _chotts_, or very salt pools, are the last shrunken and evanescent relics.

And this point about the water brings me at last to a cardinal fact in the constitution of deserts which is almost always utterly misconceived in Europe. Most people at home picture the desert to themselves as wholly dead, flat, and sandy. To talk about the fauna and flora of Sahara sounds in their ears like self-contradictory nonsense. But, as a matter of fact, that uniform and lifeless desert of the popular fancy exists only in those sister arts that George II.--good, practical man--so heartily despised, 'boetry and bainting.' The desert of real life, though less impressive, is far more varied. It has its ups and downs, its hills and valleys. It has its sandy plains and its rocky ridges. It has its lakes and ponds, and even its rivers. It has its plants and animals, its oases and palm-groves. In short, like everything else on earth, it's a good deal more complex than people imagine.

One may take Sahara as a very good example of the actual desert of physical geography, in contradistinction to the level and lifeless desert that stretches like the sea over illimitable spaces in verse or canvas. And here, I fear, I am going to dispel another common and cherished illusion. It is my fate to be an iconoclast, and perhaps long practice has made me rather like the trade than otherwise. A popular belief exists all over Europe that the late M. Roudaire--that De Lesseps who never quite 'came off'--proposed to cut a canal from the Mediterranean into the heart of Africa, which was intended, in the stereotyped phrase of journalism, to 'flood Sahara,' and convert the desert into an inland sea. He might almost as well have talked of cutting a canal from Brighton to the Devil's Dyke and 'submerging England,' as the devil wished to do in the old legend. As a matter of fact, good, practical M. Roudaire, sound engineer that he was, never even dreamt of anything so chimerical. What he did really propose was something far milder and simpler in its way, but, as his scheme has given rise to the absurd notion that Sahara as a whole lies below sea-level, it may be worth while briefly to explain what it was he really thought of doing.

Some sixty miles south of Biskra, the most fashionable resort in the Algerian Sahara, there is a deep depression two hundred and fifty miles long, partly occupied by three salt lakes of the kind so common over the whole dried-up Saharan area. These three lakes, shrunken remnants of much larger sheets, lie below the level of the Mediterranean, but they are separated from it, and from one another, by upland ranges which rise considerably above the sea line. What M. Roudaire proposed to do was to cut canals through these three barriers, and flood the basins of the salt lakes. The result would have been, not as is commonly said to submerge Sahara, nor even to form anything worth seriously describing as 'an inland sea,' but to substitute three larger salt lakes for the existing three smaller ones. The area so flooded, however, would bear to the whole area of Sahara something like the same proportion that Windsor Park bears to the entire surface of England. This is the true truth about that stupendous undertaking, which is to create a new Mediterranean in the midst of the Dark Continent, and to modify the climate of Northern Europe to something like the condition of the Glacial Epoch. A new Dead Sea would be much nearer the mark, and the only way Northern Europe would feel the change, if it felt it at all, would be in a slight fall in the price of dates in the wholesale market.

No, Sahara as a whole is _not_ below sea-level; it is _not_ the dry bed of a recent ocean; and it is _not_ as flat as the proverbial pancake all over. Part of it, indeed, is very mountainous, and all of it is more or less varied in level. The Upper Sahara consists of a rocky plateau, rising at times into considerable peaks; the Lower, to which it descends by a steep slope, is 'a vast depression of clay and sand,' but still for the most part standing high above sea-level. No portion of the Upper Sahara is less than 1,300 feet high--a good deal higher than Dartmoor or Derbyshire. Most of the Lower reaches from two to three hundred feet--quite as elevated as Essex or Leicester. The few spots below sea-level consist of the beds of ancient lakes, now much shrunk by evaporation, owing to the present rainless condition of the country; the soil around these is deep in gypsum, and the water itself is considerably salter than the sea. That, however, is always the case with fresh-water lakes in their last dotage, as American geologists have amply proved in the case of the Great Salt Lake of Utah. Moving sand undoubtedly covers a large space in both divisions of the desert, but according to Sir Lambert Playfair, our best modern authority on the subject, it occupies not more than one-third part of the entire Algerian Sahara. Elsewhere rock, clay, and muddy lake are the prevailing features, interspersed with not infrequent date-groves and villages, the product of artesian wells, or excavated spaces, or river oases. Even Sahara, in short, to give it its due, is not by any means so black as it's painted.