16 



BULLETIN 700, U. S. DEPARTMENT OF AGRICULTURE. 



Another source of error is the use of indices of leaf area instead 

 of actual leaf area. Thus when the total length of the leaves of a 

 wheat plant is used for comparative purposes, it is assumed that the 

 total leaf length is proportional to the total area of the plant. The 

 leaves of grasses are approximately triangular in shape, and their 

 actual area may therefore be determined by multiplying the length 

 of the leaf by the width and taking half of the product thus ob- 

 tained. In order to ascertain whether the leaf length was a reliable 

 index of leaf area, the leaf area was obtained in the manner de- 

 scribed above for a number of cases and the results decided by the 

 corresponding leaf lengths. Table 5 gives the leaf length, calculated 

 area, the ratio of these two, the average error in millimeters, and 

 the per cent determined by means of the formula given in the foot- 

 note to Table 4. 



TABLE 5.- 



-Relation between leaf area and le,af length in wheat and mountain 

 brome grass. 



The fact that the ratio of leaf area to leaf length is nearly con- 

 stant shows that the length furnishes a reliable index of area. The 

 average error of the ratio for wheat, using 5 plants with about 25 

 leaves in all, is 2|- per cent ; while in the brome grass, using 5 plants 

 with about" the same number of leaves as the wheat, the error is 

 about 3J per cent. 



MEASUREMENT OF PHYSICAL FACTORS. 



Each of the three type stations was equipped, in the main, with 

 automatic instruments. Air temperature, precipitation, evaporation, 

 relative humidity, sunshine, and barometric pressure were recorded 

 continuously at each station. In addition, a continuous record of the 



