the actual stimulus by virtue of the sum of preponderant intensities. The periodicity of daylight is a stimulus to the guard cells of stomata, but its relation to intensity in this connection is not clear. The amount of change in light intensity necessary to constitute an efficient stimulus seems to depend upon the existing intensity as well as upon the plant concerned. Apparently, a certain relative decrease is more efficient for sun plants than for shade plants. At least, many species sooner or later reach a point where a difference larger than that which has been efficient no longer produces a structural response. This has been observed by E. S. Clements (_l. c._) in a number of shade ecads. For example, a form of _Galium boreale_, which grew with difficulty in a light value of .002, showed essentially the leaf structure of the form growing in light of .03, while the form in full sunlight showed a striking difference in the leaf structure. In considering the light stimuli of habitats, it is unnecessary to discuss the stimulus of total darkness upon chlorophyllous plants, although this is of great importance in experimental evolution and in control experiment. The normal extremes of light intensity, i. e., those within which chlorenchym can function, are full sunshine represented by 1, and a diffuseness of .002, though small flowering plants have once or twice been found in an intensity of .001. The maximum light value, even on high mountains, never exceeds 1 by more than an inconsiderable amount, except for the temporary concentration due to drops of dew, rain, etc. It seems improbable that the concentrating effect of epidermal papillae can do much more than compensate for the reflection and absorption of the epidermis. Experimental study has shown that the maximum intensity in nature may be increased several, if not many times, without injurious results and without an appreciable increase in the photosynthetic response, thus indicating that the efficient difference increases toward the maximum as well as toward the minimum.