i9o8] CLAPP— STUDY OF TRANSPIRATION 259 



tions I have given this average loss for the day and night separately, 

 and have added the extremes observed in order to show how widely 

 the transpiration may vary in the same plant. In the plant under 

 standard conditions similar results are given, though for day only. 



A comparison of the M^HG figures for the ordinary and standard 

 conditions of the same plant will reveal some remarkable discrepancies. 

 Thus, Helianthus under the ordinary conditions stands far ahead of 

 any other plants studied, but under the standard conditions it is sur- 

 passed by several. The explanation of such differences is evident. 

 The week in which the experiment was tried may have been one of 

 exceptional dryness and brightness, or one in which air currents from 

 ventilation, or perhaps greater heat, increased the transpiration of 

 the plant under ordinary conditions but could not materially affect 

 that under standard conditions. It is evident therefore that for com- 

 parison of the transpiration of one plant with another the columns 

 under ordinary conditions have little value, and for this the column 



E 



under standard conditions should be used. It is in order to exhibit 

 clearly the effects of such external conditions upon transpiration that 

 I have plotted for the same plants the transpiration in graphs, along 

 with the graphs of external conditions, in the series of diagrams which 

 follow below. Another peculiarity of the tables is found in the great 

 range in the same plant under standard conditions. This comes in 

 part from the difficulty of keeping the conditions constant, in part from 

 the great sensitiveness of the process to even slight alterations of 

 external conditions, but chiefly from the variations in light. Abso- 

 lute proof of this could only be obtained from a series of careful experi- 

 ments where all factors would be under control and only one varied 

 at a time. It can reasonably be inferred, however, on this ground: 

 on bright sunny days there was a gradual increase and decrease pro- 

 portional to the light intensity. On cloudy days or when the roof 

 was covered with snow, with the curtain aside, the loss per hour varied 

 within much narrower limits. Comparison of the losses at the same 

 hours on different days also justifies this conclusion. 



It will occur to the reader at this point that not enough plants of 

 each species were used to give results which are fair tests of the trans- 

 piration of a given species. This criticism is only in part just. It is 

 obviously impracticable to provide either the excessive time or the 



