136 EFFECTS of HEAT 



the fharp end of the periwinkle, retained its original fhape in 

 a great meafure, fo as to be quite difcernible ; but the whole 

 was glazed over with a truly vitreous luftre. This glaze co- 

 vered, at one place, a fragment of the (hell which had been ori- 

 ginally loofe, and had welded the two together. All the angles 

 are rounded by this vitrifaction 5 the fpace between the en- 

 tire fhell and the fragment being filled, and the angles 

 of their meeting rounded, with this mining fubflance. The 

 colour is a pale blue, contrafled, in the fame little glafs, 

 with a natural piece of periwinkle, which is of a reddifh-yel- 

 low. One of the fragments had adhered to the lid, and had 

 been converted into a complete drop, of the fize of a muftard- 

 feed. It is fixed on the wax (at b), along with the other fpe- 

 cimens of the experiment (fig. 32.). This refult mews, as yet, 

 no fign of decay, notwithftanding fo great a lofs of weight. 



The laft experiment was repeated on the fame day, and pre- 

 pared in the fame manner, with large fragments of fhell, and the 

 point of the periwinkle {landing up in the cup. A heat of 34 

 was applied ; a lofs took place of 13 per cent. All the original 

 form had difappeared, the carbonate lying in the cup as a com- 

 plete liquid, with a concave furface, which did not mine, but 

 was fludded all over with the white fphericles or tufts, like 

 thofe {een in the former refults, without any fpace between 

 them. When detached from the cup, the furface moulded on 

 the platina, was white and pearly, with a flight glofs. The 

 mafs was quite folid ; no veflige whatever appearing, of the 

 original form of the fragments, (fig. 33.). A fmall piece, bro- 

 ken off near the apex of the cone, fhewed the internal ftruc- 

 ture to be quite faline. In the ad of arranging the fpecimen on 

 its ftand, another piece came off in a new direction, which pre- 

 fented to view the molt perfect cryflalline arrangement : the 

 #iinmg plane extended acrofs the whole fpecimen, and was 

 more than the tenth of an inch in all directions. This fracture, 



likewife, 



