[ 2 3 ° ] 
for fometimes a whole pillar appears as fplit all the 
way down ; of which there is a remarkable one at the 
Caufeway. 
In relation to the joints in the pillars, this work of 
nature feems to be different from any thing yet known : 
and it muff be very difficult to affign any fatisfadlory 
Caufes of it. 
I fubmit to the judgment of perfons more expe- 
rienced in thefe things what has occurred to me. 
I fuppofc, for reafons, which I ffiall give, that the 
feveral parts of thefe pillars were at firfi: formed either 
in the ffiape of a cylinder, with the upper end in a 
fpherical figure, if not both ends ; or that they were 
either fpherical or oblate fpheroids. 
For, being compofed of cryftal of fix fides, and 
fpar of three, and of a very fine black fand, it may 
be fuppofed, that, as the cryftals and fpars united, 
and formed an irregular body, the fine black fand 
fill’d up the interfaces, and formed fuch cylindrical or 
fpherical bodies, as yet foft ; but, in thin horizontal 
lamince or plates like talc, as they moflly appear to 
be j and, if great force is applied, the ftones will 
feparate in fuch plates between the joints j and thofe 
parts of the pillars, which have been expofed to 
the weather, and corroded by it, appear in fuch 
plates. Sometimes indeed there are perpendicular 
joints ; as in the fplit pillar, there feems to have 
been fuch a one all down the pillar. 
It is therefore probable, that, when this matter 
was in a fluid date, and when the ftratum of rock 
was formed, on which it was made, the fluid con- 
tiguous to the rock ftill continued in motion ; that, 
after a time, fome of the particles of matter, which 
compote 
