268 : Correspondence of J. Nicklés. 
occasional explosions, which throw out or set fire to the oil, in conse- 
quence of the projection through the oi] of gases produced by the ex- 
‘plosion. M. Berthelot avoids this result by surrounding the iron tube 
by a case (“‘manchon”) of sheet iron, closed below: the gas in this 
case, does not pass into the oil, but escapes above. 
:, 
Ane 
3 
2. A A 
i\ | == ‘ 
Oo OO) 4 ==, , 4 
whens By a 
a Deion 
= power Ty BE fies 
fe = 
— = 
os = 
-— eo sa 
Sate ee 
Parca arconge of the Figures —1. A, brick structure, reduced to one-twentieth ; bB, ae 
of the furnace; c, ash-pit; , chimney,—2. Horizontal section of the pei: posi 
: bela: 3 orizo) 
mometer enclosed in an iron tube; F, grating ; G, cast iron plate; P, ore on pile ; 
In recapitulation, the glass tube (c) is placed in a tube of iron (s)s 
surrounded b se; the whole-is heated in the oil bath, 
enclosed within a brick oven or vat. ‘ 
The closed vessels employed are either balls, matrasses, or tubes, of 
glass. The first hardly resist a pressure of more than 3 or 4 atmos- 
pheres. They are closed by means of a lamp after the substance S 
introduced. _ Glass tubes are generally used; they should be bard, little 
- fusible, unattacked by the reagents and by the heat of a lamp- The 
glass tubes used for organic analysis, are excellent for the purpose: 
They will sustain a pressure of 50 to 60 tmospheres; they — 
