: 
| 
on the Analysis of the Carbonates. 353 
Third.—A test bottle, containing a solution of nitrate of silver, 
placed between the decomposing vessel and the instrument for 
aspiration. ‘This appendage is introduced in the figure and refer- 
red to here, not because we deem it essential, when the operation 
is conducted with even common care, but as necessary to com- 
plete the picture of the apparatus of research, as used by us in 
our experiments. It will bgshown hereafter, contrary to the inti- 
mation of Erdman and Marchand, that hydrochloric acid does not 
pass through the drying tubes, either in company with the stream 
of evolved carbonic acid, or during the aspiration with the air. 
As however an extreme violence in the effervescence, accidental- 
ly occasioned, might cause some of it to escape, a fact not yet 
Witnessed in any of our experiments, we continue to use this ap- 
pendage as a sentinel to give us notice of the error. 
In adjusting the apparatus for use, great care should be taken 
to make all the connections, from the remote end of the large 
drying tube, to the short tube of the test bottle inclusive, perfect- 
ly air tight. 'To be sure of this, after putting the parts together, 
the end of the drying tube should be closed by a little fragment 
of soft cement, then setting the syphon in operation, if the junc- 
tions referred to are perfectly close, the stream of bubbles rising 
through the solution of nitrate, will soon entirely cease. The 
importance of this air-tight connection, will at once appear from 
considering that during the aspiration, the smallest opening in 
the corks or gum-elastic tubes of the decomposing bottle and its 
drying tubes, by giving admission to air from without, must in- 
crease the weight of the instrument, by the amount of moisture 
it brings with it, an increase which even ina seemingly tight 
condition of the apparatus, when the above precaution was not 
used, has in. some of our experiments amounted to two one-hun- 
dreths of a grain. | 
To facilitate the removal of the apparatus, proper for its con- 
hection,’ previous to the second weighing, the binding string 
Should be fastened by loop knots with long ends. ‘This caution, 
however insignificant it may appear, is necessary to prevent the 
ndling of the instrument, and to avoid any loosening of the 
corks. It may also be added, that in moving the instrument to 
or ftom the scale, it should be held by the horizontal tube, be- 
tween the folds of a piece of clean buckskin. For accurate re- 
Search, the experiments should be made in a dry atmosphere, of 
Vol. xnv1, No. 2.—Jan.-March, 1844. 45 
