344 
Journal of Agricultural Research 
Vol. V, No. 8 
TESTS OF THE ACCURACY OF THE RESPIRATION CALORIMETER 
At frequent intervals the accuracy with which the respiration calorim¬ 
eter will function is tested by burning ethyl alcohol in the chamber in 
such manner as to insure complete combustion and measuring with the 
apparatus the amounts of oxygen consumed and of carbon dioxid, water 
vapor, and heat produced. Commercial alcohol, pure in quality and 
containing about 90 per cent of ethyl hydroxid, is satisfactory for the 
purpose. The actual percentage is ascertained from determination of 
the specific gravity of the alcohol. The amount of oxygen that would 
be required to burn 1 gm. of the commercial alcohol, and the amounts 
of water vapor and carbon dioxid that would result from the combustion, 
are computed from the chemical equation for the reactions occurring 
in the combustion of ethyl hydroxid, with allowance for the water present 
in the sample burned. The heat of combustion of ethyl hydroxid at 
constant pressure is taken as 7.08 Calories per gram. 
The burner used in these experiments was made of two concentric 
tubes of thin brass 5 cm. in length, the outer tube being 18 mm. in ex¬ 
ternal diameter. At the lower end each tube is soldered to a brass ring, 
which provides an annular space between them 3 mm. wide, in which is a 
wick of asbestos; and as the inner tube is open at both ends, there is a 
center draft for the flame. No products of incomplete combustion have 
been found in the air of the chamber during a test in which alcohol was 
burned with this burner. 
The burner is soldered to one end of a piece of 4-mm. copper tubing, 
the other end of which passes through the wall of the chamber to the 
alcohol supply outside, from which the burner is fed in such manner that 
the rate of consumption is uniform. Attached to the outer end of the 
copper tube is an elbow of 4-mm. glass tubing, with a side outlet in the 
vertical arm to provide for an overflow. The height at which this outlet 
is set with relation to the top of the burner governs the rate of consump¬ 
tion of the alcohol. The level having been fixed, alcohol is fed into the 
vertical tube so that some of it will drop regularly from the outlet. The 
rate of combustion in the chamber is then very constant. The overflow 
alcohol is caught in a small bottle, which is weighed with the supply flask 
at the end of each period, the loss in weight of both containers showing 
the quantity of alcohol burned. The connection between the overflow 
bottle and the outlet and also that between the supply flask and the feed 
tube are such that the loss of alcohol by evaporation is negligible. 
The results of alcohol check tests of the respiration calorimeter indi¬ 
cate that the determinations made with it are at least sufficiently 
accurate for investigations of the character in which it is used. This 
is shown by the data in Table II, which summarizes the results of two 
tests selected from many equally good. 
In November, 1913, in a 3-hour period, 56.3 gm. of commercial alcohol 
containing 88.32 per cent of ethyl hydroxid were burned to test the 
