I 



METHOD FOR DETERMINING TRANSPIRING POWER 307 



continues until the color of the cobalt-chloride paper is altered 

 beyond the point where it matches the color of the less intense 

 standard. The rate of color-change thus brought about must 

 be great enough so that lateral leakage of moisture into the 

 hygrometric slip (from the air outside or from other parts of 

 the surface not covered by the slip but covered by the glass plate) 

 shall be practically negligible. At the same time, surfaces 

 that give off water too rapidly cannot be studied in this way, 

 for in such cases the rate of color-change becomes too rapid to 

 be measured. 



The water reaching the slip must be vapor and not liquid; 

 leaves whose surfaces are wet with liquid water (as rain or dew) 

 camiot be dealt with by this method. 



The difficulty arising from the fact that certain plant surfaces 

 give off moisture at a rate too slow to be studied by this means 

 may be a far more serious one, being encountered with at least 

 one of the two surfaces of certain extremely xerophilous leaves. 

 Thus, Livingston (1913) was unable to determine the trans- 

 piring power of the upper surfaces of leaves of Rhus acromatica 

 because the paper slips he employed upon these surfaces failed 

 to complete their color changes.^ ^ Such difficulties may some- 

 times be overcome by adopting a more intense blue as the weaker 

 color-standard. In other cases a thinner paper may be employed, 

 but in any event such cases require special attention. 



As has been mentioned, the hygrometric paper employed 

 in ou:' later studies, the one for which the two permanent color 

 standards have been developed, is much slower in its color re- 

 action than that used earfier, by Bakke and by ourselves. At 

 20°C. it requires about 45 seconds (41-48) for response on the 

 standard evaporating surface, with a gas layer a millimeter 

 thick between paper and porous plate. Plant surfaces to be 

 studied by means of this paper should show periods between 

 about 15 or 20 seconds and perhaps an equal number of minutes. 



'* It may be mentioned here that Livingston's statement regarding the com- 

 parative transpiring powers of Sphaeralcea plants growing at Oracle and on the 

 Desert Laboratory grounds (page 31) is transposed; it was of course those at 

 the Laboratory (lower station) that exhibited the lower indices, as is shown by 

 his table 6. 



