One question still remains to be discussed. From what direction did megalithic architecture come, and what was its original home? This is clearly a point which is not altogether dependent on the means by which this architecture was diffused. Montelius speaks in favour of an Asiatic origin. He considers that caves, and tombs accessible from above, i.e. simple pits dug in the earth, were native in Europe, while tombs reached from the side, such as dolmens and corridor-tombs, were introduced into Europe from the east. Salomon Reinach, arguing mainly from the early appearance of the objects found in the tombs of Scandinavia and the rarity of the simpler types of monument, such as the dolmen, in Germany and South Europe, suggests that megalithic monuments first appeared in North Europe and spread southwards. Mackenzie is more inclined to believe in an African origin. If he is right it may be that some climatic change, possibly the decrease of rainfall in what is now the Sahara desert, caused a migration from Africa to Europe very similar to that which many believe to have given to Europe its early neolithic population. The megalithic people may even have been a branch of the same vast race as the neolithic: this would explain the fact that both inhumed their dead in the contracted position.
It is probable that the problem will never be solved. The only way to attempt a solution would be to show that in some part of the megalithic area the structures were definitely earlier than in any other, and that as we move away from that part in any direction they become later and later. Such a means of solution is not hopeful, for the earliest form of structure, the dolmen, occurs in all parts of the area, and if we attempt to date by objects we are met by the difficulty that a dolmen in one place which contained copper might be earlier than one in another place which contained none, copper having been known in the former place earlier than in the latter.
It still remains to consider the question of the origin of the rock-hewn sepulchre and its relation to the megalithic monument. The rock-tomb occurs in Egypt, Phoenicia, Rhodes, Cyprus, Crete, South Italy, Sicily, Sardinia, Malta, Pianosa, the Iberian Peninsula, the Balearic Isles, and France. In all these places there are examples which are certainly early, i.e. belong to the neolithic or early metal age, with the exception of Malta and perhaps Rhodes and Phoenicia. Two types are common, the chamber cut in the vertical face of rock and thus entered from the side, sometimes by a horizontal passage, and the chamber cut underground and entered from a vertical or sloping shaft placed not directly over the chamber, but immediately to one side of it. It is unlikely that these two types have a separate origin, for they are clearly determined by geological reasons. A piece of country where vertical cliffs or faces of rock abounded was suited to the first type, while the other alone was possible when the ground consisted of a flat horizontal surface of rock. We frequently find the two side by side and containing identically the same type of remains. In South-East Sicily we have the horizontal entrance in the tombs of the rocky gorge of Pantalica, while the vertical shaft is the rule in the tombs of the Plemmirio, only a few miles distant.
Two curious facts are noticeable with regard to the distribution of the rock-hewn tombs. In the first place they are all in the vicinity of the Mediterranean, and in the second some occur in the megalithic area, while others do not. The examples of Egypt, Cyprus, and Crete show that this type of tomb flourished in the Eastern Mediterranean. Was it from here that the type was introduced into the megalithic area, or did the megalithic people bring with them a tradition of building rock-tombs totally distinct from that which is represented by the tombs of Egypt, Cyprus, and Crete?
The question is difficult to answer. One thing alone is clear, that in certain places, such as Malta and Sardinia, the megalithic people were not averse to reproducing in the solid rock the forms which they more usually erected with large stones above ground. The finest instance of this is the Halsaflieni hypogeum in Malta, where the solid rock is hewn out with infinite care to imitate the form and even the details of surface building.
Similarly we have seen that both in Sardinia and in France the same forms of tomb were rendered in great stones or in solid rock almost indifferently.
There can therefore be no doubt that the hewing out of rock was practised by the megalithic people, and that they were no mean exponents of the art. We have no proof that they brought this art along with them from their original centre of dispersion, though if they did it is curious that they did not carry it into other countries where they penetrated besides those of the Mediterranean. It may be that early rock-tombs will yet be found in North Africa, but it seems improbable that, had they existed in the British Isles, in North Germany, or in Scandinavia, not a single example should have been found.
On the other hand, if the megalithic people did not bring the idea of the rock-tomb with them we must suppose either that it evolved among them after their migration, or that they adopted it from the Eastern Mediterranean. The last supposition is particularly unlikely, as it would involve the modification of a burial custom by foreign influence.
We have, in fact, no evidence on which to judge the question. Perhaps it is least unreasonable to suppose that the idea of the rock-tomb was brought into the megalithic area by the same people who introduced the megalithic monuments, and did not result from contact with the Eastern Mediterranean. Similarly we ought perhaps to disclaim any direct connection between the corridor-tombs of the megalithic area and the great tholoi of Crete and the Greek mainland. At first sight there is a considerable similarity between them. The Treasury of Atreus at Mycenæ with its corbelled circular chamber and long rectangular corridor seems very little removed, except in size and finish, from the tombs of Gavr' Inis and Lough Crew. Yet there are vital points of difference. The two last are tombs built partly with upright slabs on the surface of the ground, entered by horizontal corridors, and covered with mounds. The Treasury of Atreus is simply an elaborated rock-tomb cut underground with a sloping shaft; as the ground consisted only of loose soil a coating of stone was a necessity, and hence the resemblance to a megalithic monument.