Weaver: effects of rusts upon transpiration 385 



No. Grams 



Solio per sq. dm. 1 1.860 



II 1.861 



III 1.857 



IV 1.839 



V 1.841 



VI 1.849 



Average weight of 1 sq. dm., 1.851 ; greatest variation from the 

 average, 0.012; greatest variation from the average in per 

 cent of average, 0.6 per cent. 

 From these figures and other similar ones it appears that the soHo 

 paper was unform within one per cent of error. 



Finally, it may be stated that the shrinkage of a turgid leaf placed 

 between the solio paper and the glass of the printing frame is prac- 

 tically nil during the short period necessary to get a leaf print of 

 sufficient clearness to be plainly seen. 



The method for accurately determining the area of the rust 

 pustules was not so easily found. Finally, the following plan was 

 used. It combines ease of manipulation with a high degree of ac- 

 curacy. The parts of the leaves which bore pustules were cut off, 

 and solio leaf prints made as above described. The leaves were 

 then again placed in a moist chamber and the surface spores re- 

 moved by means of a camel's hair brush. Then they were quickly 

 placed in a row on a clean smooth piece of glass and at once covered 

 with a similar piece, the two fitting closely together. This held them 

 flat and at the same time allowed little water loss. They were then 

 photographed, first the lower surface and then the upper. The latter 

 was accomplished by simply turning the glass plates over, and with- 

 out readjusting the camera (Plate XLII). The longest leaf photo- 

 graphed was measured, and from the prints an enlargement of ten 

 times the original size was made. This increased the size of the 

 ■pustules to dimensions that could easily be worked with (Plate 

 XLI). Each leaf of the enlargement was carefully compared with 

 the original leaf and any spots or dots (of which there were few) 

 not corresponding to a rust pustule,_ were eliminated. After weigh- 

 ing the enlarged leaf prints (the upper and lower svirfaces of which 

 fit exactly back-to-back) the pustular area of each was cut out and 

 weighed. From this data the per cent of rusted area could easily b-^ 

 calculated. By weighing the solio prints of the parts photographed 

 and determining their proportional relation to the total transpiring 

 surface, it was possible to determine the per cent of the pustular 

 area based on the total transpiring surface. 



