42 GASEOUS METABOLISM OF INFANTS. 



constructed of britannia, it may without injury be completely immersed 

 in water in a large aquarium vessel and so leveled as to be easily read. 

 The corrections for temperature changes are minimized by this immer- 

 sion in water. It is not possible, of course, to control the barometric 

 fluctuations, and the meter readings should therefore be corrected not 

 only for the average of the temperature fluctuations obtaining through- 

 out the experimental period, but also for the average changes in the 

 barometer. For relatively short periods this can best be done by 

 using the temperature readings taken at the beginning and end of the 

 period, and the barometer readings taken at the same time. 



The meter is calibrated by the method of weighing the gas delivered 

 from an oxygen cylinder. 1 Many tests of this type of meter show that, 

 when properly installed, it gives admirable results and when a long 

 series of experiments is contemplated, its use is strongly to be recom- 

 mended. A small, weighable cylinder of oxygen is required in either 

 method, since such a cylinder is necessary for the calibration of the 

 gas meter. 



Temperature measurements. — In the dog apparatus, the volume of air 

 inside the respiration chamber is about 250 liters; in the infant appa- 

 ratus it is about 75 liters. It is clear, therefore, that correct tempera- 

 ture measurements of this air are necessary in order to determine the 

 actual volume of the air in the chamber at the end of every experimental 

 period. We have thus far employed two carefully calibrated mercury 

 thermometers to measure the average temperature of the air in the 

 chamber, one in the cover of the chamber (Ti), the other the dry-bulb 

 thermometer of the psychrometer (T 2 ). While the two thermometers 

 rarely read alike, their fluctuations in temperature are usually parallel ; 

 consequently, for lack of better measurement, the average of the readings 

 of the two thermometers is taken as representing the average tempera- 

 ture of the air in the chamber. Experiments are now in progress seek- 

 ing a better record of the average temperature of the air by means of 

 electrical-resistance thermometers. 



Temperature control of the respiration chamher. — The importance of 

 temperature measurement has just been outlined, but it is likewise 

 important to conduct the experiments so that the respiration chamber 

 shall not be subjected to sharp and sudden fluctuations of temperature 

 during the experimental period. It has therefore been found necessary 

 to construct the water-jacket entirely around the chamber, except on 

 the top. The space between the two metal walls is filled with water. 

 During cold weather, with a mercury thermo-regulator and a small 

 burner beneath, temperature control can be very readily secured. In 

 the excessively warm days of summer, when the temperature of the 

 laboratory is considerably higher than that of the chamber, it is neces- 

 sary to place ice in the water-tank. The ice floating on the water melts 



Benedict, Physical Review, 1906, 22, p. 294. 



