Weaver: effects of rusts upon transpiration 383 



tainers of both rust-infected and control plants were carried this far 

 in the experiment so that enough would be left should an accident 

 occur to any of them. In some cases where the first leaves were not 

 fully grown, the plant was kept on the greenhouse bench over wet 

 sand for a day or two, and then a few hours before sealing equal 

 amounts of water (10 to 25 cc.) were added to each glass. Several 

 trial tests showed that containers without plants but filled with soil 

 and covered with the wax seal lost no measurable amount of water 

 during a period of several weeks. ■ 



It is believed that by the use of these methods and by keeping 

 the containers side by side throughout the experiment, the root en- 

 vironment, as regards kind of soil, its volume, water content, tem- 

 perature and aeration, was made the same for all plants. 



The method of measuring the amount of transpiration was by 

 determining the loss of weight of the containers. The first weighing 

 was made soon after the glasses were sealed. Weighings were made 

 on a balance sensitive to 2 milligrams under the load thus imposed 

 upon it, about 500 grams. The second weighing was usually made 

 about 24 hours later, care being taken to keep the plants in a fairly 

 humid atmosphere during this first interval, so that they would not be 

 injured by a too sudden change from the moist air of the bell jars. 

 Further weighings were taken, usually, at intervals of about 48 

 hours, until the plants, which were growing in the greenhouse, 

 showed indications of the leaf tips drying vip. 



In order to keep the transpiring area approximately constant, 

 only the first leaf was allowed to grow. The second leaf of each 

 plant w^as clipped daily. At no time was the humidity so high as to 

 produce guttation, and although the growth of the second leaf was 

 about 1 cm. each 24 hours, this was entirely comparable in all the 

 plants and the error of varying area thus introduced is believed auto- 

 matically to have checked itself out. 



Usually four containers of infected plants and four of control 

 plants constituted a battery for a single transpiration determination, 

 although the number was sometimes less. All of the plants m a bat- 

 tery were kept under exactly the same conditions throughout the 

 experiment. A constant record was kept of the light condition (sun- 

 shine or cloudiness), temperature, humidity, and the evaporating 

 power of the air. A few of the later experiments were carried on 

 in a large glass case of about 1.3 cu. m. capacity, and here the fac- 

 tors of humidity and temperature could be kept under better con- 



