IIl4 THE PLANT 
compared with those obtained from absorption experiments made with 
plants rooted in soil before they can be relied upon. This necessity practi-~ 
cally puts the potometer out of commission for accurate work, unless future 
study may show a somewhat constant ratio between the absorption of a plant 
in its own soil and that of a piant placed in a potometer. 
4. Measuring absolute humidity. The cog psychrometer makes it 
possible to determine the increased relative humidity produced within a 
glass cylinder or special tin chamber by a transpiring plant. From this 
result the absolute humidity is readily obtained, and by means of the latter 
the actual amount of water given off. The evident drawback to this 
method is that the increasing humidity within the chamber gives results 
entirely abnormal for the plant concerned. 
5.. Self-registering instruments, There are various methods for regis- 
tering the amount of transpiration, based upon weighing, or upon the poto- 
meter. The Richard recording evaporimeter has all the advantages of 
weighing, inasmuch as the water loss is measured in this way, and in addi- 
tion the amount is recorded upon a revolving drum, obviating the necessity 
of repeated attention in case it is desirable to know the exact course of 
transpiration. On the other hand, methods which depend upon the poto- 
meter, while graphic, are not sufficiently accurate to be of value.. 
6. The use of hygroscopic materials. Hygroscopic substances change 
their form or color in response to moisture. As they indicate comparative 
water joss alone, they are of value chiefly in the study of the stomatic 
surfaces of leaves. F. Darwin‘ has used strips of horn, awns of Stipa, and 
epidermis of Yucca to construct small hygroscopes for this purpose. In 
these instruments the error is large, but as no endeavor is made to obtain 
exact results, it is negligible. Filter paper impregnated with a 3-5 per 
cent aqueous solution of cobalt chloride is deep blue when dry. If a strip 
of cobalt paper is placed upon a leaf and covered with a glass slip it turns 
bright rose color, the rapidity of the change affording a clue to the amount 
of transpiration. 
158. Field methods. The conditions which a satisfactory field method 
of measuring transpiration must fulfill have already been discussed; they 
are accuracy, simplicity, and normality. These conditions are met only by 
weighing the plant in its own soil and habitat. This has been accomplished 
by means of the sheet-iron soil box, already described under the determina- 
tion of the chresard. The method is merely the familiar one of pot and 
balance, slightly modified for field use. The soil block, which contains the 
1Qbservations on Stomata by a New Method. Proc. Camb. Phil. Soc., 9:303. 1897. 
