Processes for Potassium. 63 
my former paper, be attended to. ‘The heat should not be earried 
beyond full redness, otherwise it will carry over a portion of carbon, 
rendering the metal impure ; whenever this occurs, the process is at 
an end; for the tube is clogged, and all the metal, subsequently subli- 
mated, is condensed over the carbonaceous deposit, requiring a full 
red heat to separate it. Hence, when we redistil, care is necessary 
as to the proper temperature. If the tube should chance to be clog- 
ged, from too high heat, the only remedy is to remove the apparatus 
from the fire, unscrew the tube from the receiver, and, when cool 
enough, detach the contents, according to previous directions, and 
put them into the receiver, Clean the tube by washing and wiping 
ull perfectly dry ;—if any oxide of iron remain on the inside, remove 
it by friction with sand and soap, and when the iron is clean and 
bright, replace the tube, insert the apparatus into the furnace, and 
distil as directed. It was stated in my former paper, that the first 
Operation, or distillation, is known to be going on well when gas is 
uniformly evolved ; if this is suddenly diminished, the retort is cool- 
ed by throwing in fresh materials fused by the heat, or the tube may be 
clogged. But if the gas gradually diminishes, it was supposed that the 
process must be nearly ended, that is, that the potassa was exhausted. 
This however is not always the case; these appearances may arise 
from the fusion of the retorts, as occurred to me in two instances 5 
the hole at first being very small, but gradually enlarging so as to 
emit the vapor of the metal as fast as sublimed. There is, however, 
one sure indication, that the potassa in the retort is exhausted, that is 
by withdrawing a few of the card-teeth or iron turnings; if there be 
* supply of potassa, they will be bright and clean, but if it is exhaust- 
ed they will be blackened by oxidation. 
an excess of carbon have been used in the operation, though 
fat required is less, I have often thought the product of metal 
pis: greater 5 at any rate it is more impure, requiring a second redis- 
tllation before all the carbon is separated. That the carbon has not 
“£0 Temoved, is ascertained by the app ce of thenaphtha in which 
ae oe which after standing some days becomes of : inti 
effect id according to the proprotion of carbon eee 4 : 
re ae arise from the slight oxidation of the wa , by w ae 
ii theta in a state of minute division, 1s set free and suspende 
quid. It js stated, in a note to the third American edition of 
mistry, that potassium may be effectually preserved in 
lah ee oil of Copaiva, on which authority I mentioned that article 
€r paper. Since that, however, I have made experiments, 
the h 
