The amount of water absorbed may be obtained directly by subtracting the final weight of the jar and moist soil from their first weight, but a desirable check is obtained by taking the dry weight of jar and soil from the first, and the final weight of these, and subtracting the one from the other as indicated in the table. A second check is afforded by daily weighings, from which the amount of water transpired is determined. Since the two sunflower plants made practically no growth during the period of experiment, the exact correspondence between water absorbed and water lost is not startling, though it can not be expected that the results will always coincide. This method has certain slight sources of error, all of which, it is thought, have been corrected in a new and more complete series of experiments now being carried on. The aeration of the soil is not entirely normal, as is also true of the capillary movements of the water, on account of the nonporous glass jar and the rubber cloth. Since the latter are necessary conditions of all accurate methods for measuring absorption and transpiration, the resulting error must be ignored. It can be reduced, however, by forcing air through the thistle tube from time to time. Sturdy plants, such as the sunflower, are the most satisfactory, since they recover more quickly from the shock of transplanting. Almost any plant can be used, however, if repotted in a loose sandy soil often enough. This permits the root system to develop normally, and also makes it possible to wash the soil away without injury to the root. The method is so recent that there has been no opportunity to test it in the field. It would seem that it can be applied without essential change to plants in their normal habitats. Very large herbs or plants with extensive root systems could not be used to advantage, and to be practicable the experiments would need to be carried on near the base station. The great value of the method, however, lies in its use as a check in determining the accuracy of other methods, and in practice it will often be found convenient and time-saving to use the latter, after they have once been carefully checked for different groups of species. This matter is further considered under measures of transpiration. [Illustration: Fig. 31. Absorption and transpiration of _Helianthus annuus_. I and II, plants repotted in soil of known weight and water-content; III, plant undisturbed in the original soil; IV, potometer containing plant with cut stem; V, potometer with entire plant.]