NOV. 22. 191$ 
Improved Respiration Calorimeter 
325 
degree. With the regular leads to each thermometer from one branch 
of the bridge circuit was a compensating loop from the opposite branch 
of the bridge, to balance the resistance of the leads in both branches of 
the circuit, and to eliminate the effects of changes in the resistance of the 
leads due to changes in their temperature and of thermal electromotive 
forces. All connections in the bridge circuit were soldered—that is, there 
were no contact connections; hence, no possibility of error due to contact 
resistance in any part of the circuit. 
The special Kohlrausch bridge was designed to measure any difference 
as large as 10 degrees or as small as 0.01 degree in the temperature of the 
water as it entered and as it left the heat absorber. The slide wire of the 
bridge, which was about 4.5 meters long, consisted of 10 turns of man- 
ganin wire wound spirally on a cylinder of marble about 15 cm. in diameter. 
The battery-circuit contact, which balanced the bridge by the adjustment 
of its position on the slide wire, was mounted on the inside of a hood 
surrounding the cylinder, which, when rotated, moved up or down on a 
threaded center post. Since the contact was in the battery circuit, 
whatever contact resistance there might be had no effect on the balancing 
point of the bridge. The resistance of the total calibrated portion of the 
slide wire was sufficient to balance the bridge when the resistance of the 
two thermometer coils differed by as much as 1 ohm, which would occur 
with a difference of 10 degrees between the temperature of the ingoing and 
that of the outgoing water. With one rotation of the hood the contact 
was moved over sufficient of the slide wire to balance a difference of 0.1 
ohm or 1 degree in the thermometers. On the edge of the hood was a 
scale with 200 divisions, each corresponding to a little over 2 mm. of the 
slide wire. A movement of the contact on the wire the space of two divi¬ 
sions would be sufficient to balance a difference of 0.001 ohm or 0.01 
degree in the thermometers. 
The sensitivity of the galvanometer was sufficient to indicate a change 
of even one division in the bridge setting, equivalent to 0.005 degree in 
the temperature of the thermometer. With the usual current of 0.03 
ampere in each half of the bridge, a change of 0.001 ohm would be indi¬ 
cated by a deflection of several millimeters on the galvanometer scale. A 
current of 0.03 ampere flowing in each resistance thermometer would 
not cause an increase of 0.005 degree in the temperature of either, when 
immersed in water flowing at the rate of 200 c. c. per minute, which 
would be not over half the common rate in the experiments. 
Provision was made for checking the results obtained with the electric- 
resistance thermometers. The second type of resistance bulb mentioned 
above was constructed so that the bulb of the mercury thermometer 
formerly used could be inserted into the bulb of the resistance ther¬ 
mometer, and the temperature differences determined by both sorts of 
thermometers at the same time. The results obtained by the two 
