cell to cell, though this is not invariably true. Since air passages are necessarily absent where cell walls touch, it is possible that this disposition of the plastids is likewise due to the lack of aeration. Translocation is directly connected with the appearance of starch. As long as all the sugar made by the chloroplasts is transferred, no starch appears, but when assimilation begins to exceed translocation, the increasing concentration of the sugar solution results in the production of starch grains. The latter is normally the case in all flowering plants, with the exception of those that form sugar or oil, but no starch. The constant action of translocation is practically indispensable to starch formation, since an over-accumulation of carbohydrates decreases assimilation, and finally inhibits it altogether. In consequence, translocation occurs throughout the day and night, and by this means the accumulated carbohydrates of one day are largely or entirely removed before the next. [Illustration: Fig. 39. Position of chloroplasts in aerial leaf (1) and submerged leaf (2) of _Callitriche bifida_. × 250.] =176. The measurement of responses to light.= Responses, such as the periodic opening and closing of stomata, which are practically the same for all leaves, are naturally not susceptible of measurement. This is also true of the transpiration produced by light, but the difficulty in this case is due to the impossibility of distinguishing between the water loss due to light and that caused by humidity and other factors. If it were possible to determine the amount of chlorophyll or glucose produced, these could be used as satisfactory measures of response. As it is, they can only be determined approximately by counting the chloroplasts or starch grains. The arrangement of the chloroplasts can not furnish the measure sought, since it does not lend itself to quantitative methods, and since the relation to light intensity is too inconstant. Hesselmann (_l. c._, 400) has determined the amount of carbon dioxide respired, by means of a eudiometer, and has based comparisons of sun and shade plants upon the results. As he points out, however, light has no direct connection with respiration. Although the latter increases necessarily with increased nutrition, the relation between them is so obscure, and so far from exact, that the amount of respiration can in no wise serve as a measure of the response to light. As a result of the foregoing, it is clear that no functional response is able to furnish a satisfactory measure of adjustment to light, though one or two have perhaps sufficient value to warrant their use. Indeed, structural adaptations offer a much better basis for the quantitative determination of the effects of light stimuli, as will be shown later.