On the Component Parts of Iron-floncs. 2 39 



loo parts of atmofphcric air. Air drawn from the water of 

 the Seine, tried by the fame eudiometer, gave the fame di- 

 minution. Air drawn from rain water was, on the other 

 hand, diminifhed by phofphorus from o^'a to 0-40. As the 

 mean of a great number of experiments was 0-35, it thence 

 follows that the proportion of oxygen in the air obtained 

 from rain water, newly fallen, is greater and more confider- 

 able than that contained in atmofpheric air, and in the air 

 of other water. 



Since rain water thus differs from common water by the 

 oxygen it holds in folution, and flnce oxygen has an influ- 

 ence on germination, and even on vegetation, as is proved, 

 by the experiments of Ingenhoufs and Senebier, it is natu- 

 ral to afcribe to it a part of the peculiar aclion which rain 

 has on the vegetation of plants different from that of other 

 water with which thev are watered. 



III. On the Comp07ient Parts of Iron-Jlones, and bozv thefe 

 in the manufacturing afecl the Quality of Crude Iron. 

 By Mr . D A v 1 D M u s h e r t of the Clyde Iron I Forks. Com- 

 mumeated by the Author. 



[Continued from p^ge zio.] 



H 



AVING in the former part of this paper ftatecf the ave- 

 rage lofs which the various natures of iron-ftones fuftaiii 

 when expofed to torrefactiorj in external air, I fliall now 

 fimply ftate the quantity of oxygen which the various clafles 

 are apt to imbibe whencxpofed to a high temperature, after 

 thofe volatile mixtures capable of aff inning the gafeous flate 

 by the combination of caloric have been expelled. 



The facility with which iron-ftones become oxydated is 

 entirely dependent upon the nature of the mixture conftitut- 

 ing fufibility or otherwife : fo that were argillaceous, cal- 

 careous, and fiiiceous iron-ftones, previoufly de-oxygenated, 

 expofed to the fame degree of heat — a degree capable of oxy- 

 dating their iron — the refult would be, that the quantity of 

 oxygen combined would be in a relative proportion to the 



lulibility 



