Xo. 510] VARIATION IN NUMBER OF SEEDS 



37; 



The conditions of humidity, which appear to be so im- 

 portant to plant life, can best be measured directly by 

 means of the dew-point apparatus, but the instrument 

 is not as satisfactory in the operation as could be desired. 

 The whirled psychrometer and the wet and dry bulb 

 thermometer are more easily operated and are quite 

 satisfactory as regards results, especially where the 

 humidities dealt with are not too low. Attention may be 

 called here, however, to the inadequacy of the wet and 

 dry bulb thermometer without a strong current of air. 

 The current should be so strong that any increase in its 

 velocity would produce no further lowering of the tem- 

 perature of the wet bulb. The hair hygrometer is unre- 

 liable unless often standardized by some other instru- 

 ment. Especially is this so in regions where the humidity 

 is usually low or where its fluctuations are very great. 

 Much improvement is surely possible in connection with 

 this factor. 



The evaporating power of the air, an integration of 

 the effects of temperature, wind velocity, relative humid- 

 ity, and, to some extent, of light intensity, is at least not 

 as difficult of measurement as formerly. The porous cup 

 atmometer can be made to give data for a curve as well 

 as a final integration for a long period. It seems that 

 we may expect much from this or some similar instru- 

 ment. 



Precipitation data are easy to obtain with amply suffi- 

 cient accuracy, but the factors of superficial and subter- 

 ranean run-off as well as that of evaporation from the 

 soil surface (all of which are almost hopelessly difficult 

 to measure) make these data very hard to interpret, ex- 

 cept for particular localities. Their final interpretation 

 will probably go hand in hand with that of evaporation 

 and soil moisture. 



For the measurement of light intensity— a factor which 

 has been shown conclusively to be of prime importance 

 in many ecological problems— we have at present no re- 

 liable and practical instrument. The delicate bolometer 



