Affinity between Bafaltes and Granite . 
tirely in allowing the black glafs time to cryftallize by a flow 
refrigeration ; and the very fame mafs, according as the heat is 
condu&ed, may, without any alteration of its chemical con- 
ftitution, be fucceffively exhibited any number of times as 
glafs, or as a ftony matter with a broken grain. In the flag 
of the iron furnaces, the fame piece generally exhibits both 
thefe appearances ; the upper furface cools fa ft, and is glafs ; 
what lies deeper, lofes its heat more gradually, and is allowed 
time to take on the cryftalline arrangement peculiar to its 
nature, in as far as a number of cryftals, ftarting from various 
points at once, and crowding each other, will admit of it. 
Here indeed the cryftals are uniform, and not of a different 
form and compofition, as in granite ; fo that this analogy 
applies clofely only to bafaltes ; and it perfectly explains why 
this body in congealing has affumed an earthy and not a 
vitreous grain. But it is eafy to conceive how, under certain 
variations of heat and mixture, a melted mafs may coagulate 
into quartz, feldfpath and fhoerl, or mica*. The moft 
permanent difference between bafaltes and granite, as to mix- 
ture, confifts in the quantity of iron $ for the earths In the 
innumerable varieties of each vary indefinitely in their propor- 
tions ; and as to heat, that the latter having been perhaps in 
general raifed from a greater depth, and confifting of more 
. 
* 66 11 elt certain qne dans les porphyres les criftaux de feld-fpath n’exiftoient 
pas avant l’epoque de la precipitation de leur bafe.** He fuppofes them to be of 
aqueous origin; but the fa6t is much to my prefen t purpofe. “ On y fait les 
progres fucceffifs de leur formation : on voit que peu a peu les fubflances qui leur 
font propres fe rapprochent, s’epurent et prennent les forces qui conviennent a 
leurs molecules ; ils etoient comme en diffolution dans leurs matrices et ils ont 
d’autant plus de facilite a fe joindre que la fluidite a ete plus parfaite et que le 
deiTechement a ete plus long.” Ifles Ponces, 247, 248* 
Vol. lxxxi. 1 
h «ge 
