5 g- Dr. Beddoes’s Obfervations on the 
huge mafles, muft have cooled more {lowly, and perhaps they 
have undergone different degrees of fufion. Befides toadftone, 
bafaltes inclofing feldfpath, zeolite, &c., various lavas clearly 
demonftrate that heterogeneous earthy cryftals do feparate from 
a fufed parte, once undoubtedly as uniform as a metallic calx and 
its reducing flux before the fubfidence of the metallic particles. 
We fhall, 1 imagine, be much deceived by a narrow analogy 
if, becaufe in our proceffes for glafs-making an homogeneous 
produft is obtained from heterogeneous materials, we conclude, 
that an heterogeneous produft may not, under other circum- 
ftances, refult from fufion; and that fire keeps infeparably 
blended whatever it has once reduced to an uniform liquid parte. 
It muft alfo be carefully remembered, that this difficulty 
does not prefs the igneous more than the opportte hypothefis. 
Since the conftituent parts of granite are cryftals, the whole 
mafs muft once have exifted in that ftate of entire difunion of 
its particles which is neceffary to cryftallization. Now, whe- 
ther fuch a folution have been effeded by the repulfive power 
of fire, or the intervention of water, it is juft as eafy to con- 
ceive heterogeneous earthy cryftals {hooting from different 
points of an uniform liquid, according to the former fuppofition, 
as the latter. 
In the natural hiftory of granite and bafaltes, another 
{hiking circumftance occurs *. they lie fo contiguous, and arefo in- 
volved in one another, that we cannot but fuppofe both to have 
undergone the fame operations of nature at the fame time. This 
is feen with the utmoft frequency upon every poffible fcale, 
and under a vaft variety of modifications. The fads already 
quoted afford inftances in point. I have before me a fpecimen 
from the park at Stockholm, confifting partly of trappand partly 
of granite. The adjacent parts are as firmly united as the other 
