ATMOMETRY AND THE ATMOMETER 97 



be taken that air bubbles are not carried down so as to enter the 

 tube on rising. 



Where a burette reservoir is employed it will of course be 

 necessary to fill only when this is nearly emptied. The burette 

 is ordinarily used when observations are to be taken at frequent 

 intervals. Filling the bottle reservoir from a burette allows an 

 error of reading of possibly 0.2 cc, usually negligible where tem- 

 perature changes are not taken into account (thermometer action, 

 expansion and contraction of the water in the apparatus) and 

 where readings are not too small. With the burette the error of 

 reading is much smaller, depending upon its diameter and gradua- 

 tion, and the volume change due to temperature variations may 

 also be made much smaller here than in the case of the bottle. 



For very short periods of operation, as from minute to minute, 

 it is convenient to prepare a horizontal, graduated glass tube 

 of small bore and mount the cup upon one end of this, letting 

 the other end dip into a vessel of water (see fig. 6). An air 

 bubble is allowed to enter at the free end, to serve as an index 

 which may be followed along the horizontal tube. This whole 

 arrangement is practically the same as the potometer, employed 

 in physiology to study rates of absorption of cut plant parts. 

 Of course the tube may be calibrated so that units of its length 

 may be converted into volume units. . 



As has been mentioned, the error of volume change due to 

 temperature alteration in the water of the instrument is to be 

 completely avoided by weighing the whole apparatus at each 

 reading and obtaining the loss directly in terms of weight. 



The N on- Absorbing Mounting. A feature of the porous cup 

 atmometer mounted as just described, which it possesses in 

 common with the open pan of water, is that the results are ren- 

 dered more or less false by the occurrence of rain. 



It has been pointed out that the reason for the cup remain- 

 ing filled with water lies in the fact that the tiny water-air me- 

 nisci which close the outer ends of the pores of the wall are not 

 broken by air pressure of considerable magnitude. Miss Aleita 

 Hopping, working in the writer's laboratory, has shown, indeed, 

 that the ordinary type of atmometer cup, when its wall is wet 



