closing in many plants may indeed take place under direct insolation, the "protection afforded" is less than "is often imagined." Thus, Bergen has found that the older leaves "of the Myrtus type" transpire more than the younger (1904b) and that, in a number of thick, xerophylous-leaved evergreens, the sun leaves transpire more than the shade leaves, and as much or more than many mesophytes (1904a). These rather unexpected results were attributed by Bergen to differences in the physical conditions within the leaf, such as the more highly developed vascular tissues and more extensive evaporation surface of the sun leaves as compared with the shade leaves, though it was believed by him that this explanation is not wholly adequate. The conclusion was drawn, which accords well at all points of contact with my own studies, that "xerophytic leaf structure is not always incompatible with abundant transpiration, but sometimes exists only for use in emergencies to protect the plant from injurious loss of water." Aside from the consideration that shade leaves may cease to transpire before they show any signs of wilting, in which, however, wilting may have proceeded far enough to result in the closure of the stomata, if this did not intervene from other cause, Bergen's observations indicate the truth of Schimper's above-quoted remark, and that the stomata are not by any means as effective in suppressing transpiration as supposed. In a word, the protection of the plant as regards its water content, by stomata, is chiefly or entirely confined to their closed condition.

Haberlandt (1904, p. 407) sees a difficulty in regarding the closure of stomata during wilting as a simple result of the reduction of turgor within the guard-cells, since, he observes, this concentration would, because of the reduction of the amount of water, lead to a higher concentration of sap, which would in turn lead to opening. It is evident, however, that wilting robs all of the cells of water, and this alone would lead to a concentration of solution in all cells at the same time. The resultant lowering of tensions would, then, be shared by the stomata, and there would result a closure. This closure may, however, not be as rapid as the wilting itself, though it is not long delayed. This delay may perhaps be referred to a slightly slower loss of water by the stomata. If, then, in conclusion of this portion of the study, the behavior of stomata in respect to regulation of water loss, both during the normal daily march of events and during wilting, is such as described, it is very natural to ask what is the function of these organs. The answers which have been given at various times have already been indicated. More recently C. E. Bessey (1898) has argued that the stomata have made possible the exchange of gases which would otherwise have been suppressed by the means adopted by the plants for conserving the water supply upon their emergence from a watery environment, and that therefore we must regard them primarily as organs for the interchange of oxygen and carbon dioxid.

He states:

The epidermis which prevents the escape of water-vapor also prevents the absorption of carbon dioxid. This difficulty was surmounted by the formation of stomata. A leaf without stomata, or what is the same thing, with its stomata permanently closed, as with wax, will not lose water, but it will starve for want of carbon dioxid. The stomata are open so long as there is no danger of such a water loss as would result in loss of turgidity, but when the cells show an approach to flaccidity the stomata close [italicizing mine].

Originally the loss of water was harmful, but many plants have in a secondary way made use of the stomata for the purpose of obtaining with the transpiration stream a supply of materials for use in building up the tissues and for other purposes. To the support of Bessey's view may be brought the lack of regulatory function of water loss by the stomata, at least to the extent that, though they allow the escape of water, they have not become so far adapted secondarily to regulate it. But, after all, the harmfulness of water loss by primitive plants is an assumption, as there may never have been a time in the history of plants when some transpiration was not a useful process, the hindrance of which would have delayed the progress of necessary physiological change within the plant. It seems to be simpler and more in accord with the known facts to regard the stomata as a means of communication between the interior of the plant and the gaseous environment, and that their function is to allow the exit and entrance of gases, inclusive of water-vapor.

RELATIVE HUMIDITY AND STOMATAL MOVEMENT.

RELATIVE HUMIDITY-ITS INFLUENCE UPON THE CONDITION OF STOMATA.

In some previous experiments (Sudden Change, p. 45) we have been concerned especially with the responses of transpiration and of stomata to change of illumination. In subjecting the plants to such changes other conditions should remain constant, or as nearly so as circumstances will permit. During the times when the experiments were made the relative humidity was very low, often as low as 5 per cent, while the difference between the dark room and the outside was, in this regard, not more than a very few per cent, a negligible amount. The fact that the stomata are found quite open during a very high temperature and a very low relative humidity-a typical desert conditionleads one to suspect that the latter has very little influence upon the stomata themselves, though, of course, the rate of transpiration may be very greatly influenced by it, just as the rate of evaporation is. If the conditions are such as to bring the plant nearly to the point of wilting, a high relative humidity will have the effect of making it easier for the plant to recover than if a low relative humidity prevailed; so that a plant may be so far wilted that the stomata are closed, when a rise in the relative humidity may then afford the condition for a returgescence and a consequent opening of the stomata.

The supposed regulatory effect of stomata upon transpiration is said to rest upon their great sensitiveness to changes of humidity (Haberlandt, 1904, p. 406) but neither the opening of stomata during returgescence, because of the dampening effect of humidity upon transpiration, nor the closure of stomata during wilting, which may but does not necessarily follow a low relative humidity, may be regarded as an expression of irritability. Indeed, from a teleological point of view, the stoma should, at any rate, close before the danger of water loss has become felt, or, as it has been expressed, before visible wilting ensues (Leitgeb). "Visible wilting" is, however, a purely artificial criterion. The danger-point, which, during wilting, we do not know much about, may be reached and passed before visible wilting is manifest. This phase of the subject is, however, treated in another place. The consideration before us is the supposed sensitiveness of stomata to change in relative humidity.

There is a small amount of evidence that a high relative humidity favors, as a condition, the wider opening of the stomata in the ocotillo. This plant has, as has been already noted, two kinds of leaves, distinguished by their position as regards the chief shoot and by their thorn metamorphosis. The primary leaves are of a briefer life than the secondary leaves, possessing less power of resistance. It is also true that in general the stomata are larger and open more widely, and it may readily be supposed that this may con

tribute to the brevity of the life span of the primary leaf, though I do not for my own part believe so.

In table 46 measurements are given of the stomata of both primary and

secondary leaves.

TABLE 46.-Measurements of stomata of primary and secondary leaves.

I was unable to take the relative humidity in these days, but judging from other experiences a difference of 30 per cent between July 25 and 26 would be a very conservative guess, and it is not improbable that it was as great as 60 per cent. By the morning of the 28th the humidity would have dropped back to 20 per cent or thereabout. Assuming these figures to be in the main correct, it would appear that, after a showery night, the high relative humidity made possible, on July 26, the opening of the stomata of both primary and secondary leaves to an average transverse measurement twice that on the day previous at the same hour. It is, of course, possible that this was the result of an increased supply of soil-water or of water in the stem absorbed through the areas immediately contiguous to the basis of the rosettes of secondary leaves, or, again, of a conservation of the supply of water in the stem by the wetting and relative humidity. Facts are already at hand to show that the ocotillo is especially sensitive to increased soil-water (Cannon, 1905), to meteoric water (Lloyd, 1905), and probably also to a high relative humidity independently of these (Cannon, 1906). It will be seen, however, that on the following two days, July 27 and 28, the stomata of the secondary leaves gradually increased their dimensions at the same hour (from 3 to 6 micra) while those of the primary leaves decreased (from 9 to 7 micra), and this under a much reduced relative humidity. Even allowing for individual differences in leaves and errors of measurement, it appears conclusive that the two sets of stomata behaved differently, and this is difficult to explain. The behavior of the stomata of the secondary leaves suggests the effect of soil-water, but why this factor did not influence the primary-leaf stomata similarly is quite unaccountable. It is possible that the formation of scission tissue had set in, which results in the separation of the leaf-blade from the endogenously formed spine (Robinson, 1904), but I endeavored to guard against this possibility

by choosing mature but not moribund leaves. The plant itself was in vigorous condition, and by its strong growth indicated a plentiful supply of soilwater, and this makes any explanation involving an increase of soil-water improbable. On the other hand, that such a difference in the relative humidity as above described should affect the stomata of a plant well supplied with water seems, in the light of other experiments, even more improbable. The following data will be seen to bear out this view:

On July 13, 1904, there was a hard shower lasting an hour, during which 1.1 inches of rain fell. In spite of this, however, the stomata showed no change in dimensions, measuring o to 3 micra before and after the shower. The plant had previously been irrigated.

A few days before this (July 8) an attached branch was allowed to remain covered by a bell-jar for one hour and a quarter, during which the relative humidity rose from 42 per cent to 87 per cent. At the close of the period, at 4h24m p. m., the stomata measured exactly as those of a control branch.

On July 11, 1904, at 9h30m p. m., the stomata of the same branch measured o to 2 micra. The branch was left covered by the bell-jar for 12 hours, when at 9h30m a. m. the stomata measured 2 to 6 micra, as also did, however, the stomata of the control. It is to be concluded, therefore, that the opening was normal and unaffected by the continuous high relative humidity.

The same behavior has been shown by Verbena ciliata. An exposure of 2 hours to a high humidity under a bell-jar on July 3, 1905, at mid-day (experiment 140) failed to produce any changes in the stomata, though at the close of the period the rate of transpiration was apparently greater. This result, it is true, may be incorrect, since there may have been moisture deposited on the leaves or absorbed by the glandular hairs, which, upon exposure to the day air beneath a fresh bell-jar, was sufficient to affect the hygrometer.

The same behavior of the stomata was shown also by all the Verbena experiments in the 1906 series (200 to 300), in which bell-jars were used. In no case was I able to detect obvious differences between the experimental part, detached from the plant, and the control part.

Similarly the method used in my experiments with Ampelopsis to determine the presence or absence of an induced rhythm in stomatal movement made it necessary to confine leaves in a small chamber, which, while lighttight, produced a high relative humidity. After 12 to 14 hours under these conditions, in the total absence of light, no tendency on the part of the stomata to open was observed, while exposure to darkening was followed by closure in spite of increasing humidity. One instance only (July 19, 1905) shows a behavior which, if not attributable to personal error, might have been due to the increased humidity.

It is evident, though, that the further opening of stomata under increased humidity is of less importance, for the teleological point of view, than closure