qualitative character and most of the quantitative measurements have been meaningless, either by reason of the use of fragments of plants or because the calibrations were of a plausible rather than of an analytical character. As a matter of fact, this loss of water from the plant, like many organic activities, is complicated by indirect and accessory functions in such manner that the main processes are difficult to evaluate. It is desirable in all such cases to construct a physical analogue which will reproduce the essential feature of the process to be measured. Notable success in this case has been attained by Professor Livingston who has devised and perfected an evaporimeter consisting of a porous clay cylinder closed at one end, which is kept filled with water (see Fig. 3). The liquid saturates the walls of this vessel and evaporates after the manner in which it would in the plant, excepting for the modifications induced by light and by the incidental structural features of the plant. The exposure of the instrument during any given set of conditions for any period gives data from which the actual evaporation may be calculated. Comparison with measured areas of leaf-surface shows that departures from the normal evaporation are made by the plant, and the departure may be expressed as the relative transpiration. This relative transpiration is never more than seven tenths of the evaporation from the instrument, and is generally much less. The evaporimeter has given us a standard and means of measurement by which all of the phases of water-loss with reference to diverse environment, in widely separated localities, and in the different stages of development of the individual may he measured with exactness.
As was fully expected, the exact calibrations have yielded data upon which new conceptions have been erected and new generalizations formulated. Among these may be mentioned that of "incipient desiccation." When the water-loss from an evaporimeter and from a plant is followed throughout the course of a June day at the Desert Laboratory, it is found under certain conditions that the midday maximum of temperature is accompanied by a maximum loss of water from the instrument. When this was compared with the loss from a plant it was seen that in certain cases the increased loss of water from the latter toward the middle of the day was checked.
All other means of interpretation of this lessened water-loss, including a consideration of the partial closure of the stomata, as determined by Lloyd, failing to explain the fluctuation in the middle of the day, recourse was had to determining the amount of water actually present in the leaves at such times. This revealed a deficiency. Briefly put, water was being lost from the membranes of the plants faster than it was being supplied to them with the result that vaporization slackened. This condition was designated as incipient desiccation and, as it is not accompanied by any external indications, its discovery was taken to be of great importance, both scientifically and economically, since the
