198 BULLETIN OF THE BUSSEY INSTITUTION. 
produced by ammonia seemed to diminish gradually in quantity up to the 
sixth fraction, but became tolerably constant when 300. cc. of percolate 
had passed through the earth, and persisted for a long time afterward. 
Up to 300 cc., the fractions were tested with lime water directly, both hot 
and cold, and they always gave reactions. So, too, on boiling the fractions 
of percolate and ¢ausing the steam to pass into lime water, reactions 
were always obtained, even when no acid had been added to the liquid. 
When acid was used, a still stronger reaction occurred. Several fractions 
of the percolate were tested in this way at irregular intervals between 
300 ce. and 1,000 ce., and the reaction of carbonic acid was obtained in 
every instance. 
In order to ascertain whether any of the carbonic acid could be removed 
from the dry earth by displacing it by means of another gas, a small column 
of the earth was arranged as if for percolation, and four litres of hydro- 
gen gas that had been washed with caustic soda were passed through the 
earth. The gas being received in a small volume of lime water gave a 
very slight reaction for carbonic acid. ‘The earth was then percolated 
with pure water in presence of the hydrogen, and the percolate tested 
with ammonia. A slight precipitate of carbonate of lime was obtained. 
The experiment was repeated upon a larger scale, eight or ten litres of 
pure hydrogen being passed through 1,000 cc. of the earth and received 
in 20 cc. of lime water. No reaction for carbonic acid was perceived 
until the lime water was boiled; but then it was well marked. Two more 
litres of hydrogen being then passed through the earth, no reaction for 
carbonic acid could be detected on receiving the gas in hot lime water. 
The earth was finally percolated with pure water in presence of hydrogen, 
and the filtrate tested with ammonia. A very decided precipitate of car- 
bonate of lime was obtained. 
The proportion of carbonic acid contained in successive fractions of the 
percolate obtained when the original dry earth was treated with water was 
carefully determined in several instances, with the following results: 
Measured quantities of the percolates were boiled with acid in an appro- 
priate flask, and the dry carbonic acid was collected in a soda-lime tube 
and weighed as such. 800 grms. of the earth were percolated with pure 
water, and there was found in the first 500 ce. fraction of percolate 
0.0065 grm. of CO,; in the next 250 cc., 0.0075 grm. CO,; in the next, 
0.013; in the next 0.014; and in the next 0.0095 grm. The accuracy 
of the process was tested as follows: Between the third and the fourth, 
and the fourth and the fifth, of these trials, the soda-lime tube was 
reattached to the apparatus after weighing, and the contents of the 
flask were treated as if a new fraction of percolate had been added; but 
no increase of weight could be detected in the soda-lime tube in either 
instance. In two other trials, each with 800 grms. of the dry earth, the 
first 250 cc. of percolate gave 0.002 grm. of carbonie acid, in each in- 
stance. A single attempt was made to estimate the amount of lime in 
these percolates by direct titration with a highly dilute standard acid; 
