2396 On the Affaying of Iron Ores 
if {corched, at the end of from fifteen to twenty days, amd 
the plants themfelves die at the end of four or five weeks. 
No feed germinates in oil. A fingle gram of common falt ‘ 
in two hundred grains of water is fufficient to retard the ve- 
, getation of plants, and may even kill them: .if they are wa- 
tered with that faline liquor *. 
Shavings of horn, next to infufion animals, are’ the mot 
favourable to- vegetation : charcoal holds the third rank. 
V. On the Affaying of Tron Ores and Iron- ae by Fufon, 
By Mr. Davip Musuer, of the Clyde Iron Works +. 
I T will eafily be conceived, from the mode of operation 
which I have adopted, that, in order to procure accurate re- 
fults, the proportion: of flux muft be varied according to the 
mixtures in the iron-f{tones or ores; and that no univerfal 
folyent can be ufed as capable of affaying al/ ores. 
As the gradation of mixtures in. the ores is almoft imper- 
ceptible, there are, in fact, no fixed limits by which Nature- 
has diftinguifhed the various claffes: we find all the varieties 
diminifhing their predominant earth, and affuming, im equal 
proportions, thofe of each other, thus conftituting the clafs 
of equalifed mixtures; yet, here, the variety of combination 
ceafes not, the predominating earth gradually becomes the 
minor part of the mixture, and that which only held a fe- 
eond rank, as to quantity, is now the chief component earth ; 
the permutation goes round, til? the earth, which exifted im 
the moft {paring quantity, now predominates to excefs, 
* C, Sylveftore obtained a refult abfolutely fimilar, by employing ma- 
rine falt as manure. 
+ The prefent is a part of the communication from Mr. Mufhet which 
appeared im our Number for July laft, (Vol. IV.p. 178,) but by an over 
fight of the Editor was oniitted insits place., It contains the table of pro- 
portions alludsd: to in our Number for September laft (Vol. IV. p. 380.) 
requifite for the obtaining from all the various iron-ftones an accurate aflay, 
and fhould have maka es: followed Mr. Mufhet’s article given in our 
July number, 
