[ 4-0 ] 
thefe groups are ailb irregularly difperfed, and of un- 
equal lize, like rliofe of pumice Hones and other com- 
mon pori ignei. Thofe of the columns of San Biafio are 
moreover commonly invefted with a fort of crocus mar- 
tiSj w’liich I have alfo frequently obferved in the pores 
of other vulcanic concretions. Thefe properties are 
furely further marks in favour of the igneous origin 
of fuch columnar cryftallizations ; efpecially, hnce they 
feem contrary to the principle by which the common 
aqueous cryftals are formed, fuccellively, et per juxta- 
pofitione7n partiwn ad partes. In fa61, thefe cryftals ma- 
nifeft no fuch poroftty. I alfo obferved, that the columns 
of Achon, though of a homogeneous fubftance, yet dif- 
fer from the common bajaltes by their immenfe ftze as 
well as colour, which is rather brown than black. The 
columns of San Biafto are likewife very large, meafuring 
often two feet in diameter. They are alfo of the ftmple 
fpecies, that is not jointed, and moftly quadrangular, 
which figure feems rather a principal charadteriftic of 
this group, being rarely obferved in others. So 
true it is, as I formerly remarked, that fome particular 
charadleriftic ever diftinguifties the different groups of 
bajaltes ; which, therefore, cannot be too* narrowly ob- 
ferved, before we pretend to form any opinion about 
their origin. Some few, but very few, chiefly of the 
:frnaller columns of San Biafio, are of a pentagonal form, 
like the fpecimen which I propofe to tranfmit to the So- 
ciety. But there are no hexagonal columns, which, in 
other 
