RESEARCH METHODS IX STUDY OF FOREST ENVIRONMENT. 167 



the same time permits the entrance of air, which may be obtained 

 by the roots through the walls of the pot, insuring their mainten- 

 ance in a healthy condition. 



For the best distribution of moisture in the soil, it is desirable 

 that a second tube, or a tee connection on the longer one, should be 

 placed so as to extend about an inch below the soil surface. Since 

 there will be some escape of vapor from each of these tubes, a pot 

 similarly prepared, but containing no tree, should be run as a 

 measure of such loss, under the various conditions to which the 

 trees are exposed. 



The weight of the can, pot, and glass tube is first obtained. The 

 seedling to be potted is then weighed with the minimum of exposure 

 to the air. The seedling is then placed in the can, which is filled 

 with moderately dry soil, and the weight of the whole is obtained 

 immediately after potting, following which water may be given to 

 the plant. Having now the weight of the air-dry soil which has 

 been placed in the pot, its net oven-dry weight may readily be com- 

 puted, after drying small samples of the same soil; and from this 

 net weight may be calculated the amount of water which the pot 

 should contain at all times to maintain a moisture equal to 50 per 

 cent, let us say, of the saturation capacity of the soil. Once the soil 

 is well settled by watering, the top of the can is sealed with a mix- 

 ture of paraffin and vaseline. A measured amount is used, so that 

 the weight of this substance may also be included in the total weight 

 which the outfit should show at the desired moisture condition. 



Knowing the weight which the outfit should have at a certain 

 moisture condition, the simplest method of measuring the transpira- 

 tion is to put the can on one side of the scales and the desired weights 

 on the other side, and to inject water from a burette until a balance 

 is obtained. The water may alternately be injected through the long 

 and the short glass tubes. 



The transpiration results will be most expressive if given in terms 

 of transpirai ion per unit of leaf area : but since, with coniferous seed- 

 Lings, the determination of leaf area with any precision is next to 

 impossible, the plan of computing (he loss per unit of weight of the 

 plant may he considered. 



Where J, number of plants, even though of the same species and 

 grown under the same conditions, are to be placed in potometers for 

 transpiration study under a variety of field conditions, the plants 

 should by all mean- he calibrated under the same conditions before 

 being distributed, since extremely great variations in individuals 



-rem 1. 1 he inherent . 



