( 334 ) 



except that, during the laft four hours^ 

 the fire was violent. The glafs on which 

 I made this experiment was the fifth fpe- 

 cies of the glafles of Lipari ; which is the 

 xnoft perfed: and the pureft glafs afford- 

 ed by the ifland. The refults were as fol- 

 low : 



The fmall quantity of air which rofe 

 above the mercury during the ignition was 

 found to be purely atmofpheric. After 

 half an hour's violent fire, I perceived, 

 through the glafs tube inferted into the raa- 

 trafs, the vitreous mafs beginning to rife in 

 tumours, which prefently rofe fo Itrongly 

 that I could only fee a tumultuary inteftine 

 motion of the liquid fiery mafs. In thefe 

 experiments I took the precaution to open, 

 from time to time, the fmall door of the 

 furnace, to examine the appearance of the 

 outfides of the matraffes. After a conti- 

 nuance of a very ftrong fire for two hours 

 and three quarters, I perceived that the 

 matrafs was broken, and that a long thread 

 of glafs had begun to iffue through the 



fradure. 



