MOVEMENTS OF GROWTH. 379 



attached to the plant is taken over the small wheel, #, which is 

 cemented on the large wheel, r, and accurately centred about the 

 same axis. A thread fixed to the larger wheel, and wound round 

 it, carries the indicator, z 9 which in our arrangement sinks as 

 growth proceeds. The indicator is balanced by a weight, g, 

 fastened to a thread wound in the opposite direction. The drum, 

 t, coated with soot, is driven by clockwork, actuated by a spring, 

 and regulated by a conical pendulum, p. The clockwork is con- 

 tained in a heavy iron case. The drum, 70 cm. long, which may 

 be fixed centrally or excentrically by displacement of the support- 

 ing axis, /, may moreover with the axis be entirely removed 

 (at Z). It makes one revolution per hour. The indicator, which 

 must not be too light, is made of brass, and horizontal ; it is ar- 

 ranged, as the figure shows, to trail on the drum, against which 

 it is made to press by giving the thread a twist. If the drum is 

 fixed excentrically, the indicator only strikes it at intervals, and 

 between times slides on the catgut stretched between the movable 

 clips, b. In the arrangement here described, the auxanometer 

 magnifies the growth fifteen times. 



The thread (silk thread) can easily be attached to the plant by 

 making a loop at one end of it, passing the other end through it, 

 and finally laying the slip-loop thus made over the upper end of 

 an internode, immediately below the base of a leaf. 



In order to limit as far as possible sources of error arising from 

 hygroscopic peculiarities of the thread, it is advisable to employ 

 silk thread only for the part passing over the pulley, and connect- 

 ing with the plant, for the rest using fine silver or platinum wire, 

 sharply bent at both ends, so that small loops of the thread can be 

 hooked on. 



If the observations are made in the light, a mirror must be 

 placed in a vertical position behind the plant, and parallel with 

 the window, in order to exclude disturbing heliotropic curvatures. 

 The experiments, moreover, are to be carried out in a place subject 

 to the slightest possible variations of temperature. Temperature 

 readings must of course always be made (for details see 7.7). The 

 soil, in which the plants must have been rooted for a long time, is 

 well watered some time before the investigation, and must not get 

 dry during the observations. 



Before being used, the rotating cylinder of the apparatus has to 

 be coated with paper. To do this we lay on the table a sufficiently 

 large piece of paper, glazed on one side, go over the rough side of 



