98.6 per cent of the amount absorbed; in other words, only 1.4 per cent of the water absorbed is retained by the plant. From this it is evident that the simplest and most convenient measure of absorption under normal conditions can be obtained through transpiration, since the discrepancy between absorption and transpiration is scarcely larger than the error of any method applicable to the field. Conversely, the measure of absorption obtained by the process described in the preceding section serves also as a measure of transpiration. The determination of the latter in the field is so much simpler, however, that it is rarely desirable to apply the absorption method. =157. Measurement of transpiration.= The water loss of a plant may be determined absolutely or relatively. Absolute or quantitative determinations are by (1) weighing, (2) collecting, or (3) measuring the water absorbed; relative values are indicated by hygroscopic substances. A number of methods have been employed more or less generally for measuring transpiration. The great majority of these can be used to advantage only in the laboratory, and practically all fail to meet the fundamental requirement for successful field work, namely, that the plant be studied under normal conditions in its own habitat. The following is a summary of the various methods, the details of which may be found in Burgerstein. 1. _Weighing._ This is the most satisfactory of all methods for determining water loss. It is more accurate than any other, and is unique in that it does not place the plant under abnormal conditions. On the score of convenience, moreover, it excels every other method capable of yielding quantitative results. Various modifications of weighing are employed, but none of these have all the advantages of a direct, simple weighing of the plant in its own soil. 2. _Collecting the water transpired._ This may be done by collecting and weighing the water vapor exhaled by a plant placed within a bell jar, or by weighing a deliquescent salt, such as calcium chloride, which is used to absorb the water of transpiration. The decisive disadvantage of these methods is that transpiration is carried on in an atmosphere far more humid than normal. If an excessive amount of salt is used, the air is abnormally dry. In both cases, the water loss decreases until it reaches a point much below the usual amount. Finally, all methods of this kind are open to considerable error, and are inconvenient, especially in field work. They are of relatively slight value in comparison with weighing.