316 Messrs. Haldane and Pembrey on an Improved Method 



procedure. The wet and dry bulbs were first read off; the 

 absorption-tubes, pairs 1 and 2, were then connected by two 

 long pieces of rubber tubing with their respective aspirators, 

 and the air was drawn through them at as equal a rate as 

 possible. The readings of the thermometers in the aspirators 

 were now taken. When one known volume (2900 cub. cent.) 

 had been drawn through, the aspirators were quickly reversed 

 and started again until 5800 cub. cent, of the air had been 

 taken for each determination. The thermometers in the 

 aspirators were now again read ; the readings of the wet and 

 dry bulbs entered 1 ; the apparatus disconnected; and the absorp- 

 tion pairs 1 and 2 stoppered. Everything was easily done 

 in twelve minutes, whilst the actual period of aspiration 

 was only about half this time, as shown by the Table of 

 experiments. t 



How closely the two determinations agree Table I. will 

 show, and there is no doubt that could they have been strictly 

 simultaneous the agreement would have been still closer. 

 This is well shown by a comparison of the first five or six 

 experiments with the last four, in which the determinations 

 were almost exactly simultaneous. 



To show that no practical error was introduced in the 

 weighings by carrying the tubes about, the variation of the 

 test pair 3 is given in Table I. 



In calculating the results for the psychrometer in Table I. 

 I have taken the mean of three readings — one at the beginning, 

 one at the middle, and one at the end of each period of 

 aspiration (i. e. one reading about every two minutes) . These 

 readings were obtained from a continuous photographic 

 record of the wet and dry bulb taken at the Radcliffe 

 Observatory. This record can be read off to two minutes and 

 one tenth of a degree Fahrenheit, and its accuracy has 

 been proved by years of use and comparison with eye- 

 readings. The volume of air drawn through the tubes has 

 always been corrected for temperature, aqueous vapour, and 

 barometric height. In calculating the tension of aqueous 

 vapour from the chemical determinations I have used the 

 table given by Shaw in his paper on Hygrometric Methods*. 



The mean of all these gravimetric determinations compared 

 with the mean of the amounts of moisture calculated from 

 the psychrometric readings is '0551 to *0549 grm., or less 

 than + ^ per cent. Thus although the psychrometric result 

 varied from the standard chemical method in the above series 

 by + 6 per cent, to — 5 per cent., the mean difference is 

 insensible. 



* Phil. Trans. 1888, A, pp. 78, 79. 



