After the completion of the system of Rift Valleys, the study of East Africa becomes the work of the archæologist and historian, instead of the geologist. The stone implements of the old terraces of Lake Baringo (see pp. 323, 324) show that man entered the region while yet the lakes were larger than they are to-day. Volcanic eruptions still took place and earth-movements continued, for some of the fault-scarps are so bare and sharp that they must be of very recent date. This continuation of earth-movements into the human period is one of the most striking features of the district. Whereas, according to the old view, British East Africa was supposed to have acquired a condition of stable equilibrium at a very early age, it has, on the contrary, been in a continual state of change since the time of the formation of our Oolitic limestones. During the eras between those of the Archean and of the Chalk, the country may have enjoyed comparative rest; but in the age of the latter, there began one of the two greatest of the series of volcanic outbursts known in the world's history. This and the resultant series of earth movements have kept the region ever since in a condition of disorder and unrest. One region has been raised and another depressed; in one place a fiord has been opened from the sea, and then separated from it; elsewhere a line of movement has reversed the direction of rivers, and transferred lakes from one river system to another; while differences in elevation have caused variations in climate and rainfall. The evidence of these changes is apparent on every hand. Scars of great earth-movements, extinct volcanic craters, dried lake basins and old river beds, show the extent and recent date of these events, and the structural instability of the region of the Great Rift Valley.

In later chapters we shall see that this structural instability has had a most important influence on all branches of the natural history of British East Africa, for its results have affected the development of both animals and plants, and helped to mould the character of the people.

CHAPTER XIII

PROBLEMS OF THE DISTRIBUTION OF THE EAST AFRICAN

FLORA AND FAUNA

"Kisauni kutamea mvinde?
"Mvita kutamea mgomba?"

(Will Kisauni grow the she-oak?

Will Mombasa grow the banana tree?)

Zanzibari Proverbs.

THE problems of the distribution of animals and plants in a country are riddles, the difficulty of which varies with the complexity of its history. In regions of great stability they are simpler than where important geographical and climatic changes have taken place in the past. Thus, on the old view of the geological uniformity of the continent of Africa, these problems might have been expected to be comparatively simple, whereas they have always proved exceptionally confused and intricate. After making a preliminary collection in East Africa, I compared notes with those of the residents, such as Mr. Ainsworth of Machakos, the late Mr. Bell Smith of Melindi, and the late Dr. Charters of Kibwezi, who had had experience on the west coast. The result sorely puzzled me, by bringing out apparently glaring contradictions in the facts of distribution. Thus certain groups run across Africa from east to west, while others extend from north to south. The commoner beetles, butterflies, and birds seemed to belong to a fauna that spread across the continent from the Atlantic to the Indian Ocean. On the other hand, some less important groups of animals, and some of the more striking of the plants, have their nearest affinities with those of

Abyssinia and the Cape. That different groups of living creatures have different geographical distributions is a wellknown fact. As shown by Dr. Blanford, in his remarkable presidential address to the Geological Society in 1890, it can be easily explained by the assumption, that the distribution of land and water has varied greatly at different periods in geological history. A group of animals, therefore, that made its appearance on the earth at one period, was able to spread along very different lines from those followed by a later group, when old land-masses had been broken up, and seas once connected had become separate. Thus, if we compare the distribution of different groups of animals, we find a gradually increasing specialisation as we pass from the oldest to the youngest groups. This is shown by the accompanying four sketch-maps. The oldest of the five classes of vertebrate animals is that of the fish, of which only the fresh-water forms are of any value in this connection; when they were introduced, they were apparently able to spread in any direction, and thus their present distribution appears to be determined mainly by temperature, for the faunas range round the world in three bands. By the time the tortoises appeared in the period of the Trias, or New Red Sandstone, the land of the southern hemisphere had apparently been broken up, and it was only to the north of the equator that the animals were able to range round the globe. According to the tortoises, therefore, North America, Europe, and Asia are all part of the same province; but before the introduction of the lizards, remains of which first occur in the Purbeck limestones, great geographical changes had occurred. These reptiles could not spread westward into North America, but they made up for this restriction by extending southward throughout Africa. After another great lapse of time, snakes appeared upon the scene in the age of the London Clay (a part of the Lower Eocene); the European species were now cut off from Africa, and in Asia were limited to the western half of the continent. The passerine birds, on which Dr. Sclater's classification is mainly based, soon followed the snakes; they were cut off from America and Tropical Africa, but their powers of flight enabled them to spread over the whole of Europe and Asia, though they did not succeed in entering India, Siam, and the Malay Peninsula.

The application of this simple explanation to the anomalies in the internal distribution of plants and animals in Africa seemed prohibited by the general assumption that this continent had always maintained its present form. Moreover, this explanation does not fully account for the distribution of some groups in small local patches; and as soon as the scientific exploration of Equatorial Africa began, numerous small outliers belonging to one province were found to occur in the middle of others.

Thus, when Baron von der Decken returned from his memorable expedition to Kilima Njaro in 1862, he brought back with him a collection of plants, many of which were determined by Ascherson to be species previously known from the mountains of Abyssinia, such as Helichrysum abyssinicum, Sch. Bip., Spilanthes abyssinica, Sch. Bip., and Achyrocline Hochstetteri, Sch. Bip. Others, such as the tree lobelias (Tupa), are allied to those of the same region, and others belonging to genera such as the Wormwood (Artemisia) are typical of the north temperate zone. During New's daring visit to Kilima Njaro in 1871 he obtained specimens of twenty plants, which enabled Sir Joseph Hooker and Professor Oliver to add the Bartsia to the list of northern genera growing on the mountain. The later collections of Teleki, Johnston, and Volkens have fully confirmed the fact of a flora existing on the higher part of this mountain, unlike that of the surrounding lowlands, and allied with those of the mountains of Abyssinia and the Cameroons, and to a less extent with those of the Cape and the Mediterranean basin.

The most striking of these plants was a giant groundsel, which was first discovered on Kilima Njaro by Mr. H. H. Johnston, after whom it was named Senecio Johnstoni. This plant, though belonging to the same genus as our English groundsels and ragworts, grows as a tree from 20 to 30 feet in height, resembling members of this genus previously known from the mountains of Abyssinia and the Cameroons.

In some photographs taken by Gedge at a corresponding elevation on Mount Elgon, similar arborescent groundsels form a conspicuous feature in the scenery. Count Teleki observed another species on Kenya, and a view published by Stuhlmann demonstrated the existence of a similar form on Ruwenzori.