351 
tubes, the one horizontal, the other vertical, both projecting a 
short distance into the bottle. It is also penetrated centrally by 
a stout platinum wire about four inches long, hooked at the low- 
er and bent twice at the upperend. A thin glass bucket for con- 
taining the solid carbonate, is represented in the figure as hang- 
ing within the bottle. This is perforated at bottom, and furnish- 
ed with a handle of platinum wire, to allow of its suspension 
from the upper and from the lower hook, in the successive stages 
of manipulation. Such is the arrangement we employ in exper- 
iments with carbonates of lime, baryta and soda, and the other 
¢arbonates which admit of accurate weighing while exposed. 
ted; the carbonate enclosed in a thin, sealed tube, is placed at the 
bottom of the bottle, and the hydrochloric acid retained until need- 
ed in the globular pipetteabove. ‘The latter appendage, drawn 
Out to a delicate and even tube below, is inserted through the 
Centre of the cork, and projects into the bottle about three fourths 
ofan inch. As in this arrangement, which is free from the errors 
incident to the use of the gum-elastic bag, the column of acid is 
€xposed to an undiminished atmospheric pressure at top, some 
care is necessary in forming the tapering stem of the pipette, oth- 
erwise the liquid will escape in drops during the first weighing. 
This is obviated by a very gentle convergence of the tube, and 
by drawing in a column of air, so as to fill the lower half, or two 
