[ »3i ] 
compofe thefe pillars, being difengag’d from the par- 
tides of water, ceas’d to move, and form’d the parts 
of thefe pillars, which are next to the rock, in the 
above-mention’d figures 3 fo much being formed only 
at once, or in a very fhort time, as extends to the firth 
joint : that then, either by change of feafon, or fome 
other accident, fo much more water mixed with thefe 
particles, as prevented their continuing to form them- 
felves into fuch a jfhape, and gave the former motion : 
that, afterwards, the decreafe of the water might 
again be the caufe of the former effedj and fo on, 
till the intire pillars were formed 3 and the top of the 
laft formed being convex, that, which was formed 
upon it, would probably be concave, and fit to it, 
either by its gravity, or by being fofter. 
All being as yet in fome one or other of thefe fi- 
gures, we fuppofe the gravitation of the fecond ftra- 
tum above the firft joint to operate in fuch a manner 
on that which was firft formed, and ftill foft, as to 
prefs it down 3 and fo eight ftones being round one 
ftone, would naturally prefs the middle ftone into an 
odtagon. 
The reafons for concluding, that they were at firft ; 
in fome of thefe figures, are thefe : 
That the concavity or convexity are either in an 
intire circle, or part of a circle : 
That fometimes the ends of the ftones appear to be 
of a fpherical form for fome fpace down, all round 
the ftone 3 fill’d up only by a matter, that feparates 
from it, as ftiall be further explain’d. 
For it is to be obferved, that the pillar is not al- 
ways fo prefs’d out, as in each ftone to form a regu- 
lar multangular figure 3. but fometimes there is a 
narrow 
