ivlth various Kinds of Glafs. 20 1 



tion, that they appear to have to each other a very flisrht 

 bond of union. It appears thereinie to me obvious, that 

 when a mafs fuch as is now dcicribcd is put under the 

 hammer, the parts will tend to repel each other, iinlefs that 

 precife degree of heal could be aCccrtaincd uhich would co- 

 alefee the cryftals without either imrting or dillipatmo- them. 



In thefe experiments four variovis fradures of call malle- 

 able iron have been obtained : renularly granulated, more or 

 iefs, on edge: laminated, parting iomciinics entire, at other 

 times in large flat facets: cubical, more or Iefs perfect: 

 mixed, compofed of the former three. 



The fird of thefe, except in one inftanee, was pecuhar to 

 the Swedifli iron ; the fecond were common to both ; the 

 third peculiar to the cold-fhort iron, one inftance exec pted ; 

 the fourth was common U) both. G^n ns. 



Exp. XIX, Britifh cold-fJiort iron - - 87c 



Kilkenny marble - 875 ciains. 



Stourbridrre clay (old) pot 875 grains. 

 A beautiful fmooth-?kinned button \vr<? obtained from 

 a fufion of this mixture, which weighed - gi^. 



Loft in fufion, equal to — j-th part the weicrht of iron, 6z 



This button fplit in circular laminas, and exhibited, in fome 

 refpects, an entirely new appearance. It forged with more 

 facility than any of the buttons obtained with glafs, and was 

 now uncommonly tough when cold. Gr.nns 



Exp. XX. Brilidi cold-fhort iron - , roo 



Kilkenny marble - >jc^o grains. 



Stourbridge clay (old) pot 750 grains. 

 A very perfect fufion was the refult of this cxpofure, 

 accompanied by a very fmooth button of metal, on 

 which were impreflcd fome faint traces of tlic ufual 

 cryftallization. It weighed _ _ _ ,^jj 



Loft in fufion, 32 

 *qual to --"s-ih part the original weight of iron. The frac- 

 ture of (his was pofTcflcd of a regularly granuIaKxl farm, the 

 cryfbis flat, and placed confiderHblv on edge. It did not in 

 the leall refcmble the former produi'l, hut \\\ tor^imr aflordcd 

 no mark fuflieicnt to diftinguifli any diflerence txlwixt their 

 qualities. 



In all tliefe experiments we invariably find that the chano-e 

 upon the quality of the iron is great. Julion invariably con?- 

 iiiunicates a gv ai <i<;\T^c ol itd-fhoilnds lu all qualities; hul 



with 



