Restored in this manner, the external appearance of these monuments would have been very similar to that of the Kubber Roumeia near Algiers and the Madracen near Blidah. The former was 200 feet in diameter, with a cone rising in steps to the height of 130 feet, which was lower in proportion than suggested above, but its interior was nearly solid, and admitted therefore of any angle that might appear most beautiful. The Madracen looks even lower, but no correct section of it has been published. The Kubber Roumeia has now been ascertained to have been the tomb of the Mauritanian kings down to the time of Juba II., or about the Christian era.[493] Judging from its style, the Madracen may be a century earlier. Be this as it may, it hardly seems to me doubtful but that these tombs are late Roman translations of a type to which the Maltese examples belonged; but the intermediate links in the long chain which connects them have yet to be recovered.
184. View of Madracen. From a plate in Blakesley's 'Four Months in Algeria.'
Internally, these Maltese monuments are rude, and exhibit very little attempt at decoration. The inner apartments, being dark, are quite plain, but the outer, admitting a certain quantity of light by the door, have a proportionate amount of ornament. At Gozo, in the outer apartment, there are, as mentioned above, scrolls and spirals of a style very much more refined than is found in Ireland or in rude monuments generally, but more resembling that of those found at Mycenæ and other parts of Greece. At Hagiar Khem and Mnaidra the favourite ornament are pit markings. Whether these have any affinity with those which Sir J. Simpson so copiously illustrated,[494] is by no means clear. In Malta they are spread evenly over the stone, and are such a decoration as might be used at the present day ([woodcut No. 181]). An altar was found in one of the outer chambers at Hagiar Khem, and in both the Maltese monuments, stone tables from 4 to 5 feet high (one is shown in the [woodcut No. 181]), the use of which is not clearly made out. They are too tall for altars, and, unless in the Balearic Islands, nothing like them is known elsewhere.
After what has been said above, it is hardly worth while to enter into the argument whether these buildings are temples or tombs. Their situation alone, in this instance, is sufficient to prove that they do not belong to the former class. Men do not drop three or four temples irregularly, as at Gozo, within a stone's throw of one another, on a bare piece of ground, far away from any centres of population. The same is the case at Hagiar Khem, where certainly three, probably four, sets of chambers exist; and Mnaidra may almost be considered a part of the same group or cemetery.
Malta, it is said, was colonised by the Phœnicians, at least was so in Diodorus' time,[495] though how much earlier they occupied it, we are not told, nor to what extent they superseded the original inhabitants. We also learn incidentally that they possessed temples dedicated to Melkart and Astarte. This is very probable, and if so, their remains will be found near their harbours, and where they established themselves; and Colonel Collinson informs me that remains of columnar buildings have been found both at Marsa Sirocco and near the dockyard creek at Valetta. These, most probably, are the remains of the temples in question, though possibly rebuilt in Roman times. The little images found in the apartments at Hagiar Khem may be representations of the Cabeiri, though I doubt it; but little headless deformities, 20 inches high, some of stone and some of clay, are not the divinities that would be worshipped in such temples, though they might be offerings at a tomb.
If these buildings were tombs, they were the burying-places of a people who burnt their dead and carefully preserved their ashes, and who paid the utmost respect to their buried dead long after their decease. The inner apartments have shelves and cupboards in stone, and numerous little arrangements which it seems impossible to understand except on the supposition that they were places for the deposit of these sacred remains. Some of the recesses have doors cut out of a single slab 2 and 3 feet square at the opening, some are so small that a man could hardly squeeze himself through, and some are holes into which only an arm could be thrust,[496] but from the rebate outside of all, the intention seems to have been for them all to be closed.
Although from all these arrangements it may broadly be asserted that they are not temples in the ordinary sense of the term; the outer apartments may be considered as halls in which religious ceremonies were performed in honour of the dead, and, so far, as places of worship; but essentially they were sepulchres, and their uses sepulchral.
We know so little of the ancient history of Malta that it is extremely difficult even to guess who the people were who erected and used these sepulchres. Most people would at once answer, the Phœnicians; but, in order to establish their claim, one of two things is necessary—either we must have some direct testimony that they erected these monuments, or we must be able to show that they erected similar tombs either near their own homes or elsewhere. Neither kind of proof is forthcoming. No such tombs are found near Tyre or Sidon, or near Carthage, and classical authorities are absolutely silent on the subject. The monuments most like them are the tombs at Mycenæ, but the differences are so great that I would hesitate to lay much stress on any slight similarities that exist. The Greek monuments were always intended to be buried in tumuli. Those at Malta have so strongly marked and so ornamental a podium outside that it is evident they never were so covered up. It may be difficult to prove it, but I fancy if we are ever to find their originals, it is to Africa we must look for them. They are too unlike anything else in Europe.