4 PROCEEDINGS OF THE AMERICAN ACADEMY 



slowly and uniformly displaced by the sulphuric acid returning from 

 K, passed into and through H aud I, arranged as already described. 

 The glass bottles /and if held each about 10 litres, and one was filled 

 with concentrated sulphuric acid. K was placed about a meter above 

 J. The two were connected by two glass tubes, a and h, each contain- 

 ing a stop-cock, and had other tubes arranged as shown in the sketch. 

 To withdraw the sulphuric acid from J into K, and to replace this by 

 the CO2, it was necessary simply to connect P to the gas-holder or gen- 

 erator, to close e and y, to open c, to exhaust the air from K by an 

 aspirator, and, opening the cock in h, to allow the H.^SO^ to be trans- 

 ferred to any desired extent. This being accomplished, d and c were 

 closed and the generator disconnected. When a measurement is in 

 progress, the CO.^ passing at a uniform rate from J through Q must be 

 replaced by sulphuric acid from K through a. This flow is regulated 

 by a tangent screw motion on the cocky, until the liquid drops or runs 

 into J at such a rate as to displace not only the necessary amount of 

 CO,,, but slightly more than this, the excess escaping slowly through c 

 and a fine orifice at P. This arrangement maintains an almost con- 

 stant pressure in J, and, as the pressure is thus always outward, there 

 can be no inward leakage of air either during the measurements or at 

 other times. The gas, after passing, always with this slight excess of 

 pressure within, through HI, enters the glass bulbs y^r, and the pressure 

 7>i at entrance is measured by the barometric height at the time plus 

 the pressure indicated by the gauge at R, read by the cathetometer. 

 The bulbs g g were sphei'ical enlargements of about three centimeters 

 in diameter of a glass tube of about half a centimeter in diameter. 

 The contents of these bulbs would be transpired by the capillaries once 

 in about fifteen minutes. The tube g g and the capillary were con- 

 nected into one piece at A by melting them together, and a similar 

 solid connection was made at B to an exit tube. The tubes CD and 

 h h were similar in all respects to A B and g g, the capillaries being 

 of very nearly the same length, and cut from the same piece of tubing, 

 which was one of those used in the experiments of 1876. The 

 connections of these tubes to the other tubing of the apparatus, as shown 

 in the drawing, were all made as described in I. 3, p. 29. The tube 

 connecting B to h h was of about 3 mm. internal diameter and 50 cm. 

 length, and at its middle point contained a branch, which extended, 

 with one joint of the kind described, to the gauge E, which in connec- 

 tion with the barometer served to measure p^. The gas at exit from 

 passed about 50 cm. of 3 mm. tubing to the point G, where a side 

 branch led to the gauge F, which gave the pressure p„. Passing by G, 



