Transpiration asja Factor in Crop Production. 31 
transpiration from the crop plus the evaporation from the soil 
surface in a pot represented a similar combined loss under field 
conditions and have called it transpiration plus evaporation. 
The amount of evaporation varies naturally in addition to the 
before-mentioned factors, with the amount of surface exposed, 
which has seldom been comparable with the area occupied by a 
similar amount of crop under field conditions. 
ENTRANCE OF RAINFALL. 
Most investigators of water requirements of crops have avoided 
the entrance of rain into the potometers. Part of the experiments 
have been conducted in greenhouses; part have been arranged to 
be moved under shelter in time of rain; while still others, which 
were fully exposed in the open, have been corrected for rain by 
keeping a record of the rainfall and adding the weight of the 
known depth and area of water to the amount lost in growing the 
crop. Very few have sealed the potometers so as to exclude the 
rain and yet permit continual standing in the open. 
The greatest error from rain would result from sealing the 
entire surface except a small opening for the plant and assuming 
that this would keep out the rain. Our tests have shown that the 
plant acts as a funnel, and for each inch of rainfall 6J pounds 
of water were caught by the leaves and conducted down the out- 
side of a full-grown, well-developed cornstalk into the potometer. 
This would make an error of 65 pounds per corn plant, if 10 inches 
of rain should fall after it had made much growth. 
EXPOSURE OF POTOMETER TO ATMOSPHERE. 
In most transpiration experiments the potometer has been left 
exposed to the atmosphere, which results in soil temperatures 
somewhat different from normal field soil. The effect of such 
difference in soil temperature due to exposure is slight, at least 
within reasonable limits as indicated in Table 9 of this bulletin, 
and data obtained from the use of exposed potometers should be 
comparable and not far from normal. 
UNINTENTIONAL LACK OF UNIFORMITY IN THE SOIL. 
An unintentional lack of uniformity in the soil is not likely 
to arise frequently. It is perhaps best illustrated by Widtsoe's 
(1909, p. 5) fertilizer experiment in 1908 with four different types 
of soil. Pots filled with soil were used which had been left un- 
disturbed since 1905, previous to which time the same soil contents 
had been subjected to varied treatments. Widtsoe assumed that 
the effects of different treatments in earlier years had been oblit- 
erated by the lapse of time, but such an assumption is not war- 
ranted. 
