If we scrutinise in the same manner the endemic forms of the higher region of the Great Atlas, we find that out of the thirty-five enumerated eight, or less than one-fourth, are to be ranked as sub-species. Of these, three are nearly allied to wide-spread Mediterranean species, one to a plant common to Spain and Algeria, two to endemic Spanish species, one to an Algerian endemic form, and one is related to a species indigenous in the Alps and other high mountains of Central Europe.
While recognising the fact that the relations between the vegetable population of the Great Atlas and that of the south of Spain are less close than might have been expected on theoretical grounds, we must yet admit that, on the whole, the Great Atlas is more nearly connected in a botanical sense with this than with any other mountain region that is known to us; and it becomes a matter of some interest to compare closely the list of species obtained by us in the Atlas, with the comparatively well known Flora of Southern Spain. The results of this comparison are given for the Great Atlas generally, and for the superior zone separately, in the following table, in which the Atlas species are distinguished under five heads: 1, those found in the higher region of the Sierra Nevada; 2, in the mountain region of Andalusia; 3, in the lower warm region below the level of about 2,000 feet; 4, absent from Southern Spain, but found in the central or northern provinces; and 5, those not included in the Spanish Flora.
Table III.
| Superior region, Sierra Nevada | Mountain region of Andalusia | Lower region of Southern Spain | Central, or Northern Spain, exclusively | Absent from Spain | |
|---|---|---|---|---|---|
| Great Atlas Valleys. 455 sp. | 103 | 82 | 100 | 44 | 126 |
| Superior region of the Great Atlas. 176 sp. | 61 | 19 | 20 | 21 | 55 |
The figures given in this table are of much interest, proving, as they do, the wide differences that exist between the Floras of two mountain regions not widely separated from each other, and exposed to climatal conditions not altogether dissimilar. We see that three-sevenths of the plants found in the higher region of the Great Atlas are absent from the South of Spain, and that the same remark applies to considerably more than one-third of all the plants found in the portion of the Great Atlas visited by us, although a notable proportion (in both cases) is to be found in Central and Northern Spain. Especially noteworthy is the fact that many of the species thus absent in Southern Spain are plants of Central Europe, most of which extend to the northern part of the Spanish Peninsula, although some of them are altogether wanting in the Floras of Spain and Portugal.
A simple inspection of our list suffices to show that it discloses no trace of affinity between the Great Atlas Flora and that of the Canary Islands, or, to use a term of wider geographical import, that of Macaronesia. The few species belonging exclusively to the latter region and to Marocco are nearly all confined to the coast region.[5] Almost all the species common to the Atlas and to Macaronesia are widely spread Mediterranean plants that ascend from the low country into the valleys. The solitary mountain plant belonging to this category is Arabis albida, the southern form of A. alpina, common in the East, and in the Apennines of Central and Southern Italy, but which, strange to say, has not been found in Spain. In Teneriffe, as in the Atlas, it ascends to about the level of 2,700 metres above the sea. The only fact suggesting a remote affinity between the Great Atlas and Macaronesian Floras is the presence in the former of a species of Monanthes, a generic group hitherto found only in the Canary and Cape de Verde Islands. But the absence of any closer connection clearly shows that the separation between the Macaronesian group and the main land of Africa must date from a period, even geologically speaking, remote.
When we come to sum up the results of the foregoing discussion, bearing always in mind the fact that we possess a mere fragment of the Flora of the Great Atlas, and that future exploration may largely modify our conclusions, we find as its most striking characteristic the presence of a large proportion of plants of Central and Northern Europe, along with a considerable number of peculiar species not hitherto known elsewhere; and we observe that these two constituents, which together form about one-half of the Flora of the region here discussed, amount to very nearly two-thirds of the species found in the higher zone. We remark that of these northern plants none are of Alpine or Arctic type, that nearly all belong to what has been called the Germanic Flora, and all are plants of the plain, not in Europe characteristic of mountain vegetation.[6]
Of the species belonging to the Mediterranean region, which constitute more than one-half of the vegetation of the middle zone, and about one-third of that of the higher zone of the Atlas, the large majority are widely diffused species. The remaining number, for the most part mountain plants, may be divided into three nearly equal sections, some being common both to Southern Spain and Algeria, others to the Atlas and Southern Spain exclusively, and others to the Great Atlas and the Lesser Atlas of Algeria. Nothing indicates any special connection with the Floras of either of those regions.
The absence of any distinct generic types from the Great Atlas Flora has already been remarked. It is not less important to note the absence of any of the southern types, characteristic of the sub-tropical zone, some representatives of which are found in the same or even in higher latitudes, in Arabia, Syria, Persia, and Northern India, and which also appear in the Canary Islands. We finally are led to regard the mountain Flora of Marocco as a southern extension of the European temperate Flora, with little or no admixture of extraneous elements, but so long isolated from the neighbouring regions, that a considerable number of new specific types have here been developed. The physical causes which have operated to bring about these conditions are doubtless numerous and complicated, but the most important of them are easily indicated. The influence of the Atlantic climate, and the prevailing direction of the aërial and oceanic currents, have fitted this region for the habitation of such northern species as do not require a long period of winter repose. In the present condition of the African continent, the Great Desert, extending for a distance of 700 or 800 miles between the Atlas and the river region of tropical Africa, effectually prevents the northward extension of most forms of animal and vegetable life; while in a period geologically recent, it is most probable that the same area was occupied by a wide gulf, which served the same purpose of barring the migration of southern forms.
It may be premature to attempt to trace in further detail the origin of the Great Atlas Flora; but the facts already ascertained certainly authorise some negative inferences. The absence of plants of Arctic type proves that if some mountains of Southern Europe received contributions to their vegetation during the glacial period by means of floating ice-rafts, that mode of diffusion did not extend to the Great Atlas. If we suppose that during the glacial period the temperature of the region north of the Atlas had fallen so low as to permit the migration of northern species across the intervening low country, we find it difficult to understand why so many species which, according to this theory, must have retreated to the Atlas on the subsequent rise of temperature, should have failed also to find a refuge in the mountains of Southern Spain.