340 TRAVELSTODISCOVER'
of this fifh great numbers in the Red Sea, and in the Indian Ocean ; how they came upon the buflies, or at the roots of them, appears more the bufmefs of the prefent narrative. To confine myfelf to the matter of fa6t, I fhall only fay, that throughout this defert are many fprings of fak- water ; great part of the defert is foflile fak, which, buried in fome places at different depths according to the degree of inclination of all minerals to the horizon, does at times in thefe foun- tains appear very near the furface. Here I fuppofe the feed is laid, and, by the addition of the rain-water that falls up- on the fak during the tropical rains, the quantity of falt- water is much increafed, and thefe fifhes fpread themfelves over the plain as in a temporary ocean. The rains decreafe, and the fun returns ; thofe that are near fprings retire to them, and provide for the propagation of future years. Thofe that have wandered too far oiFin the plains retire to the buflies as the only flicker from the fun. The in- tenfe heat at length deprives them of that fliade, and they perifli with the leaves to which they crept for flicker, and this is the reafon that we faw fuch a quantity of fliells un- der the buflies ; that we found them otherwife alive in the very heart of the fprings, we fliall further circumftantiate in our Appendix, when we fpeak of niuflels fo found in our hillory of the formation of pearls,.
Rashid was once full of villages, all of which are now ruined by the Arabs Daveina. There are feven or eight wells of good water here, and the place itfelf is beautiful beyond defcription. It is a fairy land, in the middle of an inhofpitable, uninhabited defert ; full of large wide fpread- ing trees, loaded with flowers and fruit, and crowded with- an immenfe number of the deer kind. Among thefe,-
we
