The Regulfu, upou tfee third meking, weighmg 3 pe^ 
fr^/^^f, refined upon the CopcJ, produced of Eut Gold 
2 peny weighs and J gram. This holds in rhe proportion 
of four fifths: but fomcwhat fliort of the uext be- 
fore. 
Upon thefe eomparifons, in this EKperimeD t of repeat- 
ing the melting of the Regul^ with ihthmt Antimony iXM 
Kegttlm gaineth weight each time 5 but is in proportion lefs 
rich in Gold: both which arc contrary, in repeating the 
melting of thtRegultfs with frefli Antimony^ as in the former 
Experiments. 
The remaining Antimony being reduced to a Regulmhy 
firing with Mtre znd Tamr ^ of each equal weight to ii: 
felf, and that Regulus exhaled upon the Cope! , there re- 
mained of Gold I ^ grains. This w'as lefs fine than that 
fetched out of the firfl: Antimony^ in the former Experi- 
ment of pairing Gold through feveral parcels of Antim<h 
«y ; though lofing little fenfiblc in weight , upon melting 
with a ftroBg heat and blafi upon it. So that the impurity 
was not from any remaimog ^^/i?=?^i?^/Vf/ fubftance in it | 
but from the Stiver and Copper mixed with it in the firft 
Alloy, And thefe were efteemed to about a third parr^ 
by the judgment of the eye upon the Totich.ftone* And fo 
proved upon refining ; firft with L^^i upon the Gopel, for 
ferchiug out the C<j//^r ; upon which it weighed ly grains 
and half, i\e, ons grain and half fnort of what it was be- 
fore : and then with Aq.f ortU^ after the melting down with 
more than the double weight of Silver ; upon which Ope- 
ration there remained i^ grMns^ and that not perfe(9; fine^ 
but retaining fomewhat of Silvery but finer thanCrown- 
Gi^^ allay *d with Silvery upon the Touch, about twenty 
ihYGC keratts. 
For computing the lofs of Gold upon this refining from 
the firft quantity^ videlivft , 5 peny might 21 grains and 
half,a Twelfth part , (which is r 2 grains, fave, about one 
fixih of a grain3beingdedudled for Alloy,the remainder is 
5 fenytveight 91 grains^ and And the feveral parcels of 
fine G(?W produced of the Regulm^ according to the ac- 
count 
