g4 J- Lindhard. 



only whole litres are indicated an error of + 0*5 litres is introduced 

 as the maximum. 



The thermometer was placed outside the gas-meter, at the 

 side where the air passed in, in such a way that the receiver of the 

 thermometer was in close contact with the metal, about half-way 

 up the receiver. As it was impossible to maintain a constant tem- 

 perature in the room, the thermometer was read immediately before 

 and after the experiment, and the average of the two readings was 

 taken. So far as I could determine, however, the rise and fall of 

 the temperature proceeded quite evenly, and the error produced by 

 the uncertainty of the temperature readings in the calculated vol- 

 umes of air, where it occurs, is scarcely appreciable when whole 

 litres are considered. The thermometer was divided into whole 

 degrees, but tenths could certainly be read with some practice. 



The height of the barometer was read from a pocket- 

 aneroid, which was found to be trustworthy by comparison with 

 a normal mercury barometer. As I very soon detected a possible 

 connection between respirational frequency and air-pressure, the 

 barometer was always read after the experiment. The nearness of 

 the readings could not be placed closer than to 05 mm. 



The time or duration of the experiment was determined 

 as a rule by means of a stop-watch; when this was not working 

 rightly (it could not stand exposure to extreme cold), one of the 

 ship's chronometers or a reliable watch was used. 



The number of respirations was counted by myself. Every 

 time I came to 20, I moved 2 balls on a so-called "counting-machine", 

 a wooden frame with 2 rows of 10 wooden balls moving on iron 

 wares. On 100 being reached, I shoved this row back and moved 

 on one ball on the second wire. The moving of the balls proceeded 

 quite mechanically with small movements of the right arm. 



Samples ofthe air expired were taken continuously during 

 the experiment. In the side of the tube leading from the mixing- 

 vessel was inserted a short, narrow brass-tube which connected by. 

 means of a rubber-tube with a lead-tube 05 mm. in diameter. This 

 led to a sampling-receiver, which was arranged for my special pur- 

 pose by Dr. Krogh. The air-receiver was provided at both ends 

 with a tap with a double bore; the uppermost tap was connected 

 by rubber-tubes with two lead-tubes, one of which, as mentioned, 

 was connected with the mixing-vessel, the other served as connec- 

 tion with the apparatus for analysis. From the lower tap one rub- 

 ber-tube led to a mercury-tank, from which the air-receiver could 

 be filled, another was connected with the outlet-tube, which was 

 drawn out to a capillary-tube of such a diameter that a suitable 



