vegetable and animal substances. 463 
grains of the oxide. I then transfer it, by means of a plati- 
num-foil tray and small glass funnel, into the glass tube, 
clearing out the mortar with a metallic brush. Over that 
mixture I put 20 or 30 grains of the peroxide itself, and 
next, 50 or 60 grains of clean copper filings. The remaining 
part of the tube is loosely closed with 10 or 12 grains of ami- 
anthus, by whose capillary attraction the moisture evolved in 
the experiment is rapidly withdrawn from the hot part of the 
tube, and the risk of its fracture thus completely obviated. 
The amianthus serves moreover as a plug, to prevent the pro- 
jection of any minute particles of filings, or of oxide, when 
the filings are not present. The tube is now weighed in a 
very delicate balance, and its weight is written down. A 
little cork, channelled at its side, is next put into the tube, to 
prevent the chance of mercury being forced backwards into 
it, by any accidental cooling or condensation. The collar of 
caoutchouc is finally tied on, and the tube is placed, as is 
shown in fig. 2, but without the plate k, which is employed 
merely in the case of analyzing volatile liquids. A few frag- 
ments of ignited charcoal are now placed under the tube at 
the end of the furnace next to the cistern, and the remaining 
space in the semi-cylinder is filled up with bits of cold char- 
coal. The top, b, may then be put in its place, when the 
operation will proceed spontaneously, the progressive ad- 
vance of the ignition from one end to the other being pro- 
portioned to the expansion of glass, so that the tube very 
seldom cracks in the process. Indeed I have often used the 
same tube for a dozen experiments, in the course of Which 
it became converted into vitrite, or Reaumur’s porcelain. 
Since the evolved gas is saturated with moisture, I reduce 
