322 Dc5criptio7i of an improved Jpparatus^ <^c. 



sium in, and which purpose it answers exceedingly well, 

 obviating an inconvenience and waste which sometitnes 

 hapj)en From its being dispersed through the barrel. 



In the first apparatus which I had ci)nsiructed on thii 

 plan, I had the pans made to screw together : tliis answered 

 exrreftiely well, but it po.'^sesscs no advantage over a ground 

 joini, and is more expensive. 



llie first attempt to use this apparatus, I obtained less 

 than 20 grains of the potassium : this was, however, encou- 

 ragement, my apparatus being in every respect perfect as 

 when I began the process; it served me twelve times very 

 completely, and in the last experiment I collected HO 

 grains of the metalloid from eleven drachms troy of the al- 

 kali. 



Having succeeded in preserving my apparatus (it at 

 length was melted in a small place, the lute having fallen off 

 in the process), my object was to substitute the conmioii 

 caustic potass instead of the pure potass which had been 

 generally used. I tried it; and to my satisfaction obtained 

 the result just mentioned. At this I was much pleased, 

 the kali purum being sold at one eighth of the price of 

 the pure potass. In my subsequent experiments twelve 

 drachms yielded the extraordinary produce of 170 grains, 

 the apparatus being taken out of the fire entirely free from 

 being acted on. 



Having attempted the caustic soda alone, and not suc- 

 ceeding with it, J used the proportions of two drachms of 

 soda to six drachms of potass, and obtained CO grains of a 

 beautiful compound nearly fluid, of considerable lustre, 

 and which lustre was inconsiderably diminished some days 

 after ; it nearly floated in the naphtha, being apparently of 

 the same specific gravity. 



From one p4rt of soda to seven of potass, the proportiori 

 |jsed by Mr. Davy in a similar experiment, I oblaint^d from 

 ten drachms 150 grains of metalloids, in appearance resem- 

 bling quicksilver, equally fluid at a low temperature, though 

 with this striking difTerence, of less specific gravity than 

 some very pure naphtha, in which it floated. 



The furnace 1 use i? the comtrion black-lead crucible, 

 about eight inches diameter; I generally leave off the upper 

 section as well as the flue. The furnace u^ed at the li*- 

 stitution is, if I do noL mistake, above 20 inches. The 

 double bellows are small in proportion, about 36 inches 

 {ly 14 inches. 



In performing the experiment, it may be observed in 

 addition to what has bcea said, 1 leave the stopper out of 

 ' try€ 



I 



