PHYSICAL PEOPERTIES OF WATER, ETC. 73 



" I propose to employ the instrument in this peculiarly sensitive state for the 

 purpose of estimating the amount of moisture in the air, when there is considerable 

 humidity ; but in its old form when the air is very dry. For this purpose the end of 

 the tube of the atmometer is to be connected, by a flexible tube, with a cylindrical 

 glass vessel, both containing mercury. When a determination is to be made in moist 

 air, the cylindrical vessel is to be lowered till the difference of levels of the mercury 

 amounts to (say) 25 inches, and the diminution of this difference in a definite time is 

 to be carefully measured, the atmospheric temperature being observed. On the other 

 hand, if the air be dry, the difference of levels is to be made nil, or even negative, at 

 starting, in order to promote evaporation. From these data, along with the constant 

 of the instrument (which must be determined for each clay ball by special experiments), 

 the amount of vapour in the air is readily calculated. Other modes of observation with 

 this instrument readily suggest themselves, and trials, such as it is proposed to make 

 at the Ben Nevis Observatory during summer, can alone decide which should be 

 preferred." 



(PHYS. OHBM. CHALL. EXP. — PART IV. — 1888.) 10 



