Bk. VI. Ch. V. ROUND TOWERS. 233 seems singularly plausible when coupled with the knowledge that the custom did prevail on the Continent in the Middle Ages. It is, how- ever, far from being a complete explanation, since many of these towers have only one or two very small openings in their upper story ; and there is also the staggering fact that this use is not mentioned in any legendary or written account of them which has come down to our time. On the other hand, they are frequently described as bell- towers, and also as treasuries and places of refuge, and seem even better adapted to these purposes than to that of displaying lights. That they may have been applied to all these purposes seems clear, but a knowledge of their use does not explain their origin ; it only removes the difficulty a step farther back. No attempt has been made to show whence the Irish obtained this very remarkable form of tower, or why they persevered so long in its use, with peculiarities not found either in the contemporary churches or in any other of their buildings. No one imagines it to have been invented by the rude builders of the early churches, and no theory yet proposed accounts for the perseverance of the Irish in its employment, at a time when the practice of all the other nations of Europe was so widely different. It must have been a sacred and time-honored form somewhere, and with some people, previous to its current adoption in Ireland ; but the place and the time at which it was so still remain to be determined. ^ Although, therefore, Mr. Petrie's Avritings and recent investiga- tions have considerably narrowed the grounds of the inquiry, they cannot be said to have set the question at rest, and any one who has 1 One of the towers in the East that bears most directly on the history of these Irish towers is that discovered by Dr. Tristram near Um Rasas. It is de- scribed and figured at page 145 in bis work on the " Land of Moab; " but un- fortunately the woodcut is taken from the side that does not i-epresent the door- way with the cross over it so Uke that at Antrim (Woodcut No. 671), and else- where. Like most of the Irish exam- ples, it is situated at about 10 ft. from the groimd. There is no other opening to the tower, except one on each face at the top. It has also the peculiarity tiiat it stands free but close to a small cell or chapel, as is the case with almost all the Irish towers. The one point in which it differs from the Irish examples is that its plan is square instead of being circular. This does not seem so im- portant as it at first sight may appear, seeing how many circular minarets were afterwards erected in the East, which must have had a model somewhei-e. Prac- tically, therefore, this Moabite tower may be described. Hibemice, as a square Irish lound tower. t)67. Doorway in Tower at Uni Basas. (From a Photograph.)
