112 The Plant WorIvD. 



in the latter case for use in studies upon plant transpiration at 

 the Desert Laboratory. The scattered literature on the use of 

 this instrument comprises the following references: Livingston, 

 Publ. 50, Carnegie Inst., 1906; Plant World, 10: 110, 269, 1907; 

 ibid, 11: 1, 106. 1908; Science, N.S., 28: 319. 1908; Proc. 

 N. Y. Hortic. Soc, 1907; Transeau, Bot. Gaz., 45: 217. 1908; 

 Amer. Nat., 43: 487. 1909; Shaw, Amer. Nat., 43: 420. 1909; 

 Gager, Month. Weath. Rev., 36: 63. 1908; Yapp, Ann. Bot., 

 23: 276-319. 1909. While the instrument has proved satis- 

 factory in many ways and in the hands of many operators, yet 

 its use has entailed a number of difficulties, and it is with the 

 obviation of some of these latter that this paper has to deal. 



Correction of readings. — The main objection to unglazed 

 porcelain as an evaporating surface is that, even though good 

 distilled water be exclusively used (as should always be the case), 

 the salts of the clay gradually migrate to the outer and upper 

 portion and clog the pores to a greater or less extent, thus re- 

 ducing the effective surface from which evaporation may take 

 place. With high rates of evaporation, as in the desert, the effect 

 of this clogging may often be detected after a week or two. In 

 more humid regions the eft'ect is less marked, while in a nearly 

 saturated atmosphere it may practically vanish. With exces- 

 sive evaporation the reverse is found to be true with some cups, 

 and salts are, for several weeks at least, drawn from the interior 

 to the surface of the cup, where they crystallize in a frost-like 

 coating and are blown away by the wind. The result of this ap- 

 pears to be that the more soluble constituents of the clay, which 

 it seems practically impossible to wash out with water or acid, 

 are removed, leaving the walls somewhat more porous and thus, 

 relatively, increasing the rate of water loss. It is obvious that 

 an evaporating surface which thus capriciously varies in its 

 effective extent is highly unsuited to the measurement of the 

 evaporating power of che air, and this condition of affairs makes 

 it advisable, as has been mentioned in earlier papers, to re- 

 standardize the cups as frequently as possible. This should be 

 done at least once a month and is desirable at intervals of a week 

 or two. According to this plan, a standardized cup is placed 

 in position, allowed to operate for a period, and is then replaced 

 by another standardized cup. During the period of operation of 



