594 Wedgwood’s additional Qbfervatiom on 
though there was full a fenfible difference, in confiftence, be- 
tween thofe which were firft and laft prefled out from the fame 
mafs of clay. The intermediate ones, within a certain dis- 
tance from the two extremes, correfponded very nearly with 
one another j fo that by rejecting a Sufficient number of the 
firft and laft, and ufing the intermediate ones only , the inequality 
may be confidered as almoft annihilated, 
I iieverthelefs Still found that, in Strong fire, the edges be- 
came a little prominent, though not fo much as before. I 
was aware that thefe pieces muft partake, in Some degree, of 
the imperfection of thofe made in the mould; having their 
Surfaces rendered, by their friction againft the fides of the 
aperture, more compad than the inner part. But I SufpeCted that 
Something might depend alio upon the form , and accordingly 
made many trials for ascertaining the form that might be leaSfc 
liable to this irregularity : the angles only were beviiled off, the 
fides were rounded, the pieces were rounded all over, made of 
ovals and other curves, and both the longeft and ffiortefi: di- 
mensions were ufed as the extent to be meafured : the general 
refult was, that the nearer they came to a circular figure, the 
lefs inequality they contracted in the fire, and by making them 
entirely circular, the imperfeCtion appeared to be obviated alto- 
gether ; cylindric pieces bearing the ftrongeft fires without the 
leaft appearance of prominence or inequality in any part of 
their Surface. I have therefore chofen this laft form, leaving 
only one narrow flat fide (db* fig. 2.) as a bottom for the pieces 
to reft upon, and to diftinguiffi the pofition in which they are 
to be meafured in the gage. 
I have endeavoured at the fame time to obviate whatever 
inaccuracy the inequality of compaCtnefs may be capable of 
producing, by fo adjufting the aperture through which the 
rods 
