OF THE CRIMEAN PENINSULA.

7»

the principal divifions of which form fquares of different fizes,and frequently of an irregulär figure, the wall above alluded towas apparently compofed of dry ftone-work; fo that both fideswere faced with larger ftones, and the inner parts were filledup with fmaller fragments. Their thicknefs exceeds two arfhines(four feet, eight inches); and, had thefe walls been cementedwith clay, that tenacious earth could fcarcely have been wafliedaway, efpecially in thofe fevv places ftill remaining at theheight of feveral feet. Some of the inner fquares, but parti-cularly the exterior ones, all of which are in their divifionsraifed of fimilar walls, appear to have been empty, and withouthabitations : there is, at leaft, no trace of dwellings difcoverable.In other fquares, the only veftiges now left, confift of therubbifh of parallel walls eredted in ftraight lines; being alter-nately three and four arfhines diftant from each other, andheaped together of calcareous rock, fo as to be nearly two arfhinesthick at their bafe. From this arrangement, I am induced toconjedure, that the fpace between every two oppofite wallshad formerly been divided into dwelling rooms, by tranfverfeearthen walls j and that the larger interftices were defigned forftreets, or perhaps the reverfe. Such fuppofition is renderedftill more probable by the occurrence of an extenfive fquaretowards the South end of the city, where the parallel wallsterminate; a circumftance which fuggefts the idea, that regulärftreets had led thither in every dire&ion. Thefe walls, whichrun to a fhort diftance from the outer ones, and thofe formingthe fquare, generally terminate with a very large flat ftone,the breadth of which is equal to their thicknefs. The emptyfquares appear to have been common enclofures for cattle, or

to