Spiculums of Rejlccling Tdefcopes. 493 



of the cf ucible, and the copper at the top of it, which I have frequently done, the copper 

 will melt with a very little heat j whereas when copper is put into the crucible by itfelf, it 

 requires a pretty ftrong heat to caufe it to melt. When I firft made ufe of this method, I 

 imagined I had difcovered avery eafy one to melt the copper, and confequently I thought 

 I had greatly improved the common method ; and as Mr. Mudge afcribed the pores in the 

 metal to the tin being calcined by the great heat of the fluid copper at its firft melting, 1 

 naturally expected to find the metal made by the above procefs totally free from pores, 

 efpecially in the fecond melting, as the heat was confiderably lefs than if the copper had 

 been melted firft by itfelf. However, I always found it full of pores, much more porous 

 than I had ever feen it before. For fome time I could not difcover the true caufe, having 

 no idea that the pores were owing to the tin remaining fo long in the fire in a fluid ftate, 

 and therefore in a ftate of perpetual calcination. I attributed the porofity of the metals 

 which I made of this compoGtion to a multitude of caufes, till, thoroughly tired of expe- 

 riments and conje£lures, to afcertain the true reafon, I was determined to melt the copper 

 firft, and the tin afterwards, as I had always done, before I dropped upon this improved 

 method as I imagined. The refult was, the metal was infinitely more compaft, and much 

 lefs porous. By melting the copper firft, and then adding the tin to it, I foon difcovered 

 that the longer the tin remained in the fire, the more porous the metal turned out ; and 

 vice verfa, the fooner I poured it off, after I had added the tin, the more compaA and free 

 from pores it proved. From thefe obfervations, I determined to try the efFcdl of adding 

 the tin in a fluid ftate to the melted copper, and to cool the whole immediately, to prevent, as 

 far as I was able, calcination from taking place. Experience confirmed my conje£lures } 

 and I foon found, that by pouring the whole melted mafs the inftant they were mixed and 

 ftirred together into cold water, the metal always proved ia the fecond melting foiid, and 

 much more compaft, beautiful, and white, than I had ever feen it by any other procefs. 

 One thing I cannot pafs over, as it affords a clear proof of the ufe of arfen'ic in rendering 

 the metal much more folid and compafl, and confequently more free from pores, than if 

 no arfenic had been ufed in the compofition. Whenever I made the compofition by melt- 

 ing the copper and tin together, by putting them into the crucible at the fame time, and 

 melting them down together, the metal was always porous, as I obferved before : however, 

 frequently I melted it afterwards, and though I gave it no more heat than was barely ne- 

 ceffary to melt it, yet if 1 added to this very porous metal, after it was melted, a fmall 

 quantity of arfenic, viz. one ounce to one pound of the metal, it was really aftoniftiing 

 to fee how much better the metal turned out, being confiderably harder than before, and 

 incomparably lefs porous. I mention this circumftance, which any one may eafily try, to 

 fl^ew the very great advantage of ufing a fmall portion of arfenic to render it more com- 

 pact ; and as Sir Ifaac Newton juftly obferves *, more ivh'tte than before. The ufe of the 

 fmall portion of brafs in this compofition, is to render it more tough, and not fo exceffive 



* See Appendix to Gregory's Optics, p. 419, or the Philofophicarl Tianfafllons, Nc. LXXXI. 

 Vol. III. — February iSooi jS brittle 



