METHOD FOR DETERMINING TRANSPIRING POWER 293 



ard.) The solution, 100 cc. in a glass pan somewhat larger in 

 diameter than the diameter of the filter-paper circle, is agitated 

 dm*ing the entire treatment and the sheet is reversed once. 

 Next follows double rolUng on the ferrotype plate, with the 

 bare roller and with blotting-paper interposed, the latter being 

 again renewed between rollings. There once more results a 

 uniformly moist sheet of filter paper, which is now placed in 

 the proper solution of potassium ferrocyanide. Here it re- 

 mams for the same time as in the case of the ferric chloride 

 solution, for the more intense color, and from fifteen to twenty 

 seconds, for the less intense. Upon being removed from this 

 solution the sheet is once more subjected to double rolhng on 

 the plate, and to another double rolling with blotting-paper 

 interposed (the latter once more renewed between the two roll- 

 ings). Finally, the circle is placed between two fresh sheets 

 of dry blotting-paper and thoroughly dried by means of the 

 hot electric flatiron. 



Each finished sheet is compared (after cooHng) with a stand- 

 ard sheet of the proper intensity of color, and is discarded if 

 any difference in color can be detected. The final intensity 

 of color may be altered by altering the time during which the 

 circle remains in the ferrocyanide solution, by altering the tem- 

 perature, or by altering the concentrations. A chemical ex- 

 cess of ferrocj^anide is always to be used. After a number of 

 circles have been treated the solutions are renewed. 



The colored papers thus prepared are found to be permanent; 

 at least they do not alter their color when exposed to air and 

 sunshine for several days, and they do not alter their color with 

 changes in hydroscopic moistm-e, such as are obtained when 

 they are placed over the standard evaporating surface or upon a 

 leaf. The rather great labor and expense pf preparing them 

 are offset by the fact that but little paper is required to make a 

 very large number of sHps of the proper size for leaf tests. 







\Ci 



\ 



T^-- 



