260 



Methods. 

 1. Measurement of Transpiration. 



It was impossible to measure the loss of water by transpiration by weighing 

 owing, first, to the difficulty of access to Dilkera with the necessary instruments 

 of precision, and secondly, to the size of some of the plants and the great length 

 of time which must elapse before potted plants are ready for experimentation. 

 The means adopted, therefore, was that of measuring the loss of water from 

 cut shoots in a potometer. The burette form of potometer described by Lloyd 

 (1908) was adopted. This instrument is capable of measuring the loss of water 

 to within 05 c.c, an accuracy which is not attainable with most forms of poto- 

 meter balance. Further, the burette type has been used by students in the 

 University under the writer's supervision, and has been found to give very 

 uniform results with the one species of plant. This form of potometer has 

 also the advantage of rapid and accurate reading, and it is believed that the 

 results would differ very little from those obtained by other methods provided 

 that the precautions mentioned below be observed. 



The shoots were cut under water early in the morning, carefully trimmed 

 with a sharp knife, and the green cortex stripped away to prevent clogging of 

 the vessels by bacterial decomposition. The wood was inserted into rubber 

 tubing of 5 mm. diameter previously filled with water and attached to the 

 burette. By means of the rubber tubing a good water-tight joint is obtained. 

 The shoots were supported vertically. 



No readings were taken for the first hour and a half in order that the 

 shoots might become turgid. In any case, Lloyd (1908), using American desert 

 plants, has found that the difference between the amount of water transpired 

 and the amount absorbed is practically negligible. Thus the loss from the 

 burette measures directly the amount of water transpired by the shoots. Although 

 . it is not claimed that the burette potometer gives absolute readings for the 

 transpiration rates, it does give relative readings, so that the rates are strictly 

 comparable, which is the main requirement for the present considerations. 



The tops of the burettes were covered with small specimen tubes to prevent 

 evaporation from the meniscus. The battery of potometers was erected on a 

 stand improvised on the spot, and is shown in pi. xxi., fig. 1. 



2. Measurement of the Climatic Factors. 



(a) Temperature. — The temperature was measured in degrees Fahrenheit by 

 means of a thermometer attached to the potometer stand. 



(h) Light Intensity. — The measurement of light intensity has been neg- 

 lected by many workers on transpiration and photosynthesis, such terms as 

 "dull," "bright," etc., having little value. Inasmuch as it is a factor in stomatic 

 control it cannot be neglected. The intensity of illumination was measured by a 

 simple type of actinometer, namely, by noting the time taken for a strip of 

 sensitive paper to become the same shade as a standard fixed disc. The relative 

 intensities are plotted on the curves in seconds. Readings were taken every 

 half-hour, as shown on the curves ; only the hourly readings appear in the 

 table, fig. 2. 



(c) Relative Humidity. — This was determined by means of wet and dry 

 bulb thermometers which were protected from the wind. The wet and dry 

 bulb readings were reduced to "relative humidity" by means of Glaisher's formula. 



(d) Evaporating Poiver of the Air. — A method sensitive to slight variations 

 in evaporating power from hour to hour was necessary, and the type of atmo- 

 meter used by Livingston (1906) was employed. The apparatus consisted of a 



