ATHENS. 339 
consequent friction against the sides of the CHAP. 
well, the interior of those massive marble cylin- w - > 
ders has been actually grooved all round, to 
the depth of two or three inches: in some 
instances, transverse channels appear crossing 
the others obliquely, and to an equal depth. 
An effect so remarkable, caused in solid marble 
by its attrition with one of the softest sub- 
stances, affords convincing proof that a great 
length of time must have elapsed before any 
one of those furrows in the stone could have 
been so produced; and that many ages would be 
requisite to form such channels in any number. 
Having selected a dry well for our experi- 
ment, whose mouth was covered by a cylinder 
remarkably distinguished by this appearance, 
we removed a quantity of stones and rubbish, 
and found at the bottom a substratum of moist 
marie. In this humid substance (the original 
deposit of the water when the well was used), 
the number of terra-cotta vessels, lamps, 
pitchers, bottles, some entire, others broken, 
was very great. We removed thirty-seven in 
an entire state, of various sizes and forms. 
They were chiefly of a coarse manufacture, 
without glazing or ornament of any kind; but 
the workmen brought up also the feet, handles, 
z 2 
