MODIFIED by COMPRESSION. 145 



making the whole ten feet in length. In this manner, a pref- 

 fure upwards was applied to the barrel, equal to the weight of 

 10 cwt. 



In the former method, in which the barrel flood with its 

 muzzle upwards, the weight was applied while the metal was 

 liquid. In this cafe, it was neceflary to let it previoufly con- 

 geal, otherwife the contents would have run out in placing the 

 barrel in the muffle, and to allow the liquefaction ellential to 

 thefe trials, to be produced by the propagation of heat from 

 the muffle downwards. This method required, therefore, in 

 every cafe, the ufe of an air-tube ; for without it, the heat 

 acting upon the breech, while the metal at the muzzle was 

 ftill cold, would infallibly have deftroyed the barrel. A great 

 number of thefe experiments failed, with very confiderable 

 wafte of the fufible metal, which, on thefe occafions was nearly 

 all loft. But a few of them fucceeded, and afforded very fatis- 

 faclory refults, which I fhall now mention. 



In November 1803, fome good experiments were made in 

 this way, all with a bore of 0.75, and a prefliire of 10 cwt. 



No. 3. — On the 19th, a good limeftone was obtained in an 

 experiment made in a temperature of 21 , with a lofs of only 

 1.1 per cent. 



No. 4. — On the 2 2d, in a fimilar experiment, there was 

 little exudation by the muzzle. The pyrometer gave 31°. 

 The carbonate was in a porous, and almoft frothy ftate. 



No. 5. — In a fecond experiment, made the fame day, the 

 heat rofe to 37 or 41 °. The fubftance bore ftrong marks of 

 fufion, the upper part having fpread on the little tube : the 

 whole was very much fhrunk, and run againft one fide. The 

 mafs fparkling and white, and in a very good ftate. 



No. 6. — On the 2 5th, an experiment was made with chalk,and 

 fome fragments of mail fhell, with about half a grain of water. 

 The heat had rifen to near 51 or 49 . The barrel had been 



Vol. VI.— P. I. T held 



