PUKIFICATION OF SULPHURIC ACID. 
29 
It has been proposed to introduce into the retort, pieces of 
glass, strips of platinum, &c. Some have recommended to 
apply the heat only to the upper part of the liquid, but 
these methods I find by experience do not accomplish the 
object desired. I have endeavoured to discover some plan 
capable of facilitating the formation and the disengagement 
of the vapour, and the following I find to succeed the 
best : 
1. I introduce into the retort some pieces of Quarzite, a 
variety of granulated quartz, resulting from the conglomera- 
tion of particles of that substance, probably under the in- 
fluence of a high temperature. It is important to choose 
this variety, the others giving far less satisfactory results. 
The shape of the pieces is by no means a matter of in- 
difference, with reference to the success of the process. I 
have remarked in a great number of distillations, that the 
fragments obtained by striking on the edge of a former 
cleavage, facihtates the distillation better than very thick 
pieces. The number as well as the size of the pieces, must 
also be observed. If they are too small, and in too small a 
quantity, the bumping may occur, because the frag,ments 
are easily lifted by the vapour which is disengaged, con- 
sequently, the bottom of the retort being no longer covered 
with the siliceous substance, assumes the condition of a 
smooth vessel, and the vapour is produced with difficulty. 
If they be small, and in too great quantity, the ebullition is 
irregular, bumping occurs without ebullition, bursts of 
vapour take place at intervals, the liquid rises in the re- 
tort, and if the latter be rather full, it may pass over into the 
neck. The fragments should be at least a centimetre (one- 
third of an inch) in length, for the operation to succeed well. 
I use them generally as large as the neck of the retort will 
allow. Ten or twelve scales of quarzite of this size are 
sufficient to distil readily several pounds of acid. It is well 
to cover the retort, that is to say, to distil in a reverberatory 
3* 
