ELIMINATION 177 



constant, and this calculation the student should make for all 

 plants that he studies. 



Measurement op Leap Akeas. This may most rapidly be accom- 

 plished by following the outlines of the leaves themselves, or else tracings 

 of them, by a planimeter, an engineer's instrument which reads off the areas 

 directly. Or the leaves may be traced upon cross-section paper and the 

 enclosed areas counted. Better, however, especially if many are concerned, 

 is the plan of tracing the outlines upon one sheet of paper of uniform thick- 

 ness, cutting the tracings out, and weighing them in comparison with the 

 weight of a definite measured area of the same paper. Instead of tracing 

 the outlines they may, for greater accuracy, be printed under a sheet of glass 

 upon sensitized paper, these prints being, later cut out and weighed. There 

 is, of course, some transpiration from stems and petioles, but since this, is, 

 relatively to that from the leaves, very small, and, further, since the areas 

 of those parts are in general proportional to those of the leaves, no great 

 error absolutely, and none at all relatively, is made by omitting them from 

 the calculation, though for very exact work they would be included. It is 

 much better to take the area of the leaf as a whole for the standard of com- 

 parison, and not the area of its two surfaces, for thus it is possible to keep 

 to the gram-meter-hour system already used for photosynthesis and respira- 

 tion. 



The records of transpiration in the foregoing experiment 

 show notable fluctuations from day to day and from hour to 

 hour, and even a cursory observation thereof will show that the 

 fluctuation is in some measure connected with the immediately 

 surrounding conditions. Hence we are now faced by this prob- 

 lem: 



In what measure is transpiration affected by variations in 

 atmospheric conditions, viz., temperature, light, humidity, baro- 

 metric pressure? 



This may be determined most simply by accurate simultaneous 

 observations of the transpiration in comparison with , the records of 

 instruments exhibiting the surrounding conditions, a method which 

 is wholly satisfactory only when all the instruments are autographic 

 or self-registering. 



Experiment. Place a plant, freshly prepared as for the preced- 

 ing experiment, under the ordinary fluctuations of the experiment 

 greenhouse. Then, either by very frequent (at least hourly) weigh- 

 ings, with simultaneous readings of the meteorological instruments, 

 or else, and much better, by use of autographic instruments, deter- 

 mine the transpiration of the plant and the contemporary meteoro- 

 logical conditions for a considerable period of time, preferably a week. 



