C 6o 7 ] 
mud neceflarily happen at laft, becaufe the thick- 
nefs of it will not exceed a certain * limit. 
67. Befides the giving way of this wall, the fire 
may undermine the lpace containing the water, and, 
by that means, open a communication between them. 
Let us fuppofe one of thefe come to pafs, and the 
time arrived when the partition begins to yield. If 
then the water had any way to efcape readily, the 
breach would be made, and the melted matter would 
burft forth immediately, and flow out in large quan- 
tities at once amongft it ; but as this is not the cafe, 
and it can only efcape by oozing flowly between the 
ffrata, and through the fiflures, the way that it came, 
the breach will be made gradually, from whence we 
may account for fome appearances that have preceded 
great earthquakes. 
68. We are told, that two or three days before 
an f earthquake in New England, the waters of fome 
wells were rendered muddy, and flank intolerably 
This limit will depend upon the thicknefs of the matter necef- 
fary to prevent fo quick a communication of the heat or cold 
through it, as that the water fhould be able to diminifh the heat 
of the lire confulerably. The thicknefs requifite to do this, is very 
different in different kinds of bodies. Metals of all kinds tranfmit 
heat and cold extremely readily ; but bricks and vitrified fubftances 
(with which laft we may clafs the matter under our prefent confi- 
deration) tranfmit them very flowly : the walls of the hotteft of our 
furnaces, when built of bricks, and eighteen inches thick, will not 
tranlmit more heat than a living animal can bear without injory, 
though the fires are continued in them for ever fo long a time • 
probably, therefore, if vve allow two feet for the thicknefs of the 
matter, cooled and rendered hard by the contact of the water, we 
fhal! not underdo it. 
t See Philo f. Tranf. N° 437. or Martyn’s Abridgm. vol. viii. 
p. 689. 
why 
