June, 1919] Variations in Transpiration 495 



Experimental Methods. 



The experiments reported in this paper were performed 

 under the direction of Dr. E.N. Tanseau in the Plant Physiology 

 Laboratory of the Ohio State University. The apparatus used 

 in determining the water loss from the plants, which has 

 been described by Transeau^, automatically records the loss of 

 weight from sealed potted plants. In order to equalize light 

 conditions in the greenhouse, the apparatus was mounted and 

 operated on a rotating table. In the darkroom the apparatus 

 was arranged on a laboratory table. 



Records of the environmental factors influencing water loss 

 from plants were obtained from instruments placed on the 

 table with the other apparatus. The temperature and humidity 

 of the air were recorded by a hygrothermograph, which was 

 checked from time to time with a standard sling psychro- 

 meter. The evaporation rates were obtained from standardized 

 porous cup atmometers. The record for the duration of sunshine 

 was obtained from the U. S. Weather Bureau Station, 

 Columbus, Ohio. 



The water loss from the plants was measured as described 

 by Transeau^ from potted, irrigated plants. Leaf areas of 

 the plants were measured at the beginning and at the close 

 of each experiment and all rates of water loss are given as 

 the rate from one hundred square centimeters of leaf area, con- 

 sidering one surface of the leaf. Leaf water content was deter- 

 mined by weighing and drying leaves from plants treated 

 exactly like those used in the determination of the water loss. 

 The sizes of the pore openings of the stomata were measured 

 from strips of epidermis fixed in absolute alcohol as suggested 

 by Lloyd- except that the calculation of the dimensions was 

 different. 



The results of the experiments given in this paper are 

 expressed by curves because they give a better representation 

 of what took place in the experiment, besides occupying less 

 space and allowing more results to be placed together for 

 comparison. A standard method for plotting the curves was 

 chosen and all curves, as nearly as possible, made in that way. 

 The times in hours, beginning with 1 for 1 A. M., to 24 for 12 

 o'clock midnight, are used as abscissae, five millimeters equal- 

 ling one hour. The rates of transpiration per hour, or varia- 

 tions per hour are used as ordinates. 



