140 



is neither saturated at room temperature nor perfectly dry. the change 

 in the degree of moistness of the air on becoming saturated is what the 

 air originally lacked of being saturated. The appropriate correction to 

 introduce is that fraction of the tension of aqueous vapor for the room 

 temperature which it lacked of saturation. Suppose the apparatus was 

 originally filled with the air of the room, and that it was forty per cent, 

 saturated at room temperature, sixty one-hundredths of the tension of 

 aqueous vapor is the number to be subtracted from the observed barometer 



reading; the corrected reading is B — — — — — w, in which B is the 



barometer reading, H is the hygrometer reading in per cent., and w is 

 the tension of aqueous vapor for the room temperature. 



Nearly all works accessible to the author give such directions for the 

 manipulation as involve the use of the air of the room in the inner tube, 

 yet give for the calculation the correction B-w. The error introduced in 

 this way would be greatest if the air were saturated with moisture, and 

 would then amount at a room temperature of 20 deg. C. to 17 in approxi- 

 mately 760, or 1 in about 45, and this condition is closely approached in 

 damp, warm weather. Omitting the correction altogether when the air 

 used is nearly dry gives an equal error in the opposite direction, approxi- 

 mated in very cold weather. 



A quite appreciable error, then, may be avoided and the calculation 

 made more nearly correct theoretically by using the correction given above. 



Of the works accessible to the author only H. Erdmann's Anorganische 

 Chemie discusses the correction, directing that if the apparatus is filled 

 with a dried gas the tension of aqueous vapor should be deducted ; if with 

 ordinary air, no correction should be made. All other works fail to con- 

 sider the point, some deducting the tension, others not, without specifying 

 the conditions. 



