Almost all the so-called dolmens in the Channel Islands are of this class. One has already been given (woodcut No. 11), and it may safely be asserted that all chambers which were wainscoted with slabs, so as to form nearly perfect walls, and all that had complicated quasi-vaulted roofs were, or were intended to be, covered with mounds--more especially those that had covered passages leading to them. There is, however, a very wide distinction between these sepulchral chambers and such a monument as this at Pentre Ifan, in Pembrokeshire.[195] The top stone is so large that it is said five persons on horseback have found shelter under it from a shower of rain. Even allowing that the horses were only Welsh ponies, men do not raise such masses and poise them on their points for the sake of hiding them again. Besides that, the supports do not and could not form a chamber. The earth would have fallen in on all sides, and the connexion between the roof and the floor been cut off entirely, even before the whole was completed. Or, to take another example, that at Plas Newydd, on the shore of the Menai Strait. Here the cap stone is an enormous block, squared by art, supported on four stone legs, but with no pretence of forming a chamber. If the cap stone were merely intended as a roofing stone, one a third or fourth of its weight would have been equally serviceable and equally effective in an architectural point of view, if buried. The mode of architectural expression which these Stone men best understood was the power of mass. At Stonehenge, at Avebury, and everywhere, as here, they sought to give dignity and expression by using the largest blocks they could transport or raise--and they were right; for, in spite of their rudeness, they impress us now; but had they buried them in mounds, they neither would have impressed us nor their contemporaries. [Illustration: 50. Dolmen at Plas Newydd. From 'Archæologia Cambrensis.'] As before mentioned, however, the great argument against the theory of their having been always covered up is the impossibility of accounting for the disappearance of the tumuli. If they had been situated on fertile plains where the land was valuable for agricultural purposes, it might be assumed that a civilized people with highly cultivated antiquarian tastes might have been at the trouble and expense of removing the tumuli for the sake of the land, and of preserving the dolmens for their historical value. But that the rude peasantry of Cornwall and Wales should have done this is inconceivable, more especially as by far the greater number of these monuments are situated on bleak moorlands of no agricultural value whatever. Still more inconceivable is it that they should have done it so neatly and so carefully that no trace of the mound can now be found either around the stones or in the neighbourhood.