[ 2 35 1

nulated texture, not ill resembling that , of the commonporcelains.

Those which were longer and longer exposed to the sire,received more and more alterations. The grains, at firstfine and of some degree of glossiness, grew larger andduller ; and at length, through sundry gradations, less re-markable and less uniform, what had once been glass,and afterwards a compact hard porcelain, became a porousfriable substance, like a mass of white sand slightlycohering.

During the change os the glass into a fibrous porcelain,it generally preserved the smoothness of its surface, and thesand freely parted from it: in the subsequent changes,part of the sand baked together upon the surface, andstrongly adhered, not to be got off, and not greatly differ-ing from it in appearance; I have sometimes been at aloss to distinguish the matter which had been glass fromthe sand that surrounded it. In some pieces the sandy coatwas parted from the internal matter by a number of finecavities resembling a dotted line : in others, they wereclosely applied together.

Such were the general effects of continued cementationin many repetitions of the experiment, though not with-out variations in some particulars. Sundry pieces becamethroughout white, and almost opake, and some blue,before they contracted any fibrous coat, which afterwardsproceeded in the lame manner as in the others. Somepieces, being broken in different parts after they had beenchanged to a fibrous state, instead of the uniform trans-verse disposition os the fibres, had several prominences onthe surfaces of the fractures, from which the fibres iffuedas rays in all directions. After the fine granulated state,which succeeded to the fibrous on a continuance of the ce-mentation, some pieces became porous, while in others the

H h 2 grains