45^ Ofi Mf, W<r4gzvood*s Pjyremeier, 



vvifli he had mentioned whetl>er all were taken from the 

 lame box,) whereas with his own, he fays., the variations 

 have only been from a half to a whole degree, and they exhi- 

 bited no figns of vitrification when cxpofed to a heat able to 

 fule iron and deflroy the bed Hcffian crucibles. 



From his experiments he infers, that, if a clay be employed 

 which contains 34 per cent, of argil, s^nd either rock cryftal, 

 triturated white fand, or pure filex be ^dded to it tp make iip, 

 the proportion before noted, pyrometer pieces may at any 

 time be formed equally refractory with thofe of Mr. Wedg- 

 wood, and whofe power of cojitra6lion will be exaxSlly the 

 fame as his. 



But why fliould natural clays b? made ufe of at all for 

 forming pyrometers, which muft fometimes vary in the pro- 

 portions of their component parts, even when taken from dif- 

 ferent parts of the fame bed ? 



If there be any difference between the Englidi pyrometers 

 now made and thofe firll prepared by Mr. Wedgwood, it 

 inoft probably arifes from fome fucb caufe; and the fame 

 difference may be expelled to be found in beds of clay ia 

 France. It does not feeni impoffible that chemifls every where 

 might agree to employ the fame proportions oi pure argil, 

 iilex, and water, in the formation of pyrometers. Even if the 

 argil were not abfplutely pure, if they employed argil obtained 

 alvvay? by the fame means (as from alum by ammonia, the 

 alum having previoufly undergone feveral folutions, filtra- 

 tions, a.nd cryftallifations to free it from foreign matters), 

 they would be fure to form pyrometers which would give 

 limilar refults. Lime and iron fhould be completely excluded, 

 as ingredients which muft render the pyrometers lefs refrac- 

 tory. 



If, in making pyrometers by fuch means as we have pro- 

 pofed, fuch proportions could be fallen uppn as would agree 

 in the refuks with Mr. \V^edgwo.o.d*s firfl-made pyrometers, 

 it would be a great convenience to. men of fcience, as many 

 fa6ls have already been eftabliflied by the indications they 

 aflbrdcd. Jf this pannot be done, one of the two following 

 ^lethods might he adopted : Either, 



\. To repeat the experiments on the degrees of heat nc- 



celTarv 



