We shall in the succeeding series of papers deal with the subject of tropism in general. Different plant organs undergo curvature or bending, sometimes towards and at other times away from the stimulus which induces it. The problem is very intricate; the possibility of its solution will depend greatly on the accurate determination of the immediate and after-effects of various stimuli on the responding organ. The curvature induced in the growing organ is brought about by variation, often extremely slight, of the rate of growth; the result, moreover, is liable to be modified by the duration and point of application of stimulus. The difficulties connected with the problem can only be removed by the detection and measurement of the minutest variation in growth, and by securing a continuous and automatic record of the entire history of the change. In the chapter on High Magnification Crescograph an account is given of the apparatus which I have devised by which the rate of growth may be magnified from ten thousand to ten millions times. It is thus possible to measure the imperceptible growth of plants for a period shorter than a single second. The variation of normal rate of growth is also found by measuring successive growth records on a stationary plate at regular intervals, say of ten seconds, or from the flexure in the growth-curve taken on a moving plate (p. 163). I was next desirous of exalting the sensitiveness to a still higher degree by an independent method, which would not only reveal very slight variation induced in the rate of growth, but also the latent period and time-relations of the change. For this purpose I at first devised the Optical Method of Balance[1] which was considered at the time to be extremely sensitive. The spot of light from the Optical Lever (which magnified the rate of growth) was made to fall upon a mirror to which a compensating movement was imparted so that the light-spot after double reflection remained stationary. Any change of rate of growth--acceleration or retardation--was at once detected by the movement of the hitherto stationary spot of light in one direction or the other. [1] "Plant Response"--p. 413. A very careful manipulation was required for the adjustment of the Optical Balance; the record moreover was not automatic. For these reasons I have been engaged for several years past in perfecting a new apparatus by which, (1) the balance could be directly obtained with the utmost exactitude, (2) where an attached scale would indicate the exact rate of growth, and (3) in which the upsetting of the balance by external stimulus would be automatically recorded, the curve giving the time relations of the change. PRINCIPLE OF THE METHOD OF BALANCE.