THE PLANT AS A MACHINE 5 



These lateral leaflets, when in normal condition, go on con- 

 tinuously, and apparently spontaneously, executing approxi- 

 mately up and down movements, each of which takes from 

 two to four minutes to complete. 



The great difficulty in recording the j)ulsatory movements 

 of Desmodiuui arises from the extreme slenderncss of these 

 lateral leaflets. This is such that in attaching to them a 

 recording lever, however light, its weight, and the friction of 

 the writing-point, are sufficient to bring their movements to 

 a stop. I have, however, succeeded in overcoming this 

 difficulty by devising a recording Optical Lever. 



This lever consists of a very light aluminium wire, or, 

 which is still better, the stripped quill of a peacock's tail 

 feather, this being extremely light, and sufficiently rigid for 

 the purpose. The two arms of the lever are unequal. 

 The fulcrum rod rests on frictionless supports of glass or 

 agate. The same rod carries a light mirror. A thread of 

 cocoon silk is stuck to the motile leaflet by a minute drop 

 of shellac varnish. The far end of the thread is looped, 

 and fixed at any suitable notch on the arm B of the lever. 

 The other arm of the lever has a light sliding counterpoise. 

 It will thus be seen that by gradually shifting the silk loop 

 nearer the fulcrum, the magnification may be increased. 

 When the automatically moving leaflet executes a downward 

 movement, the arm B is pulled down, and there is a rotation 

 of the fulcrum rod with its attached mirror. A spot of light 

 reflected from the mirror is thus suddenly moved downwards 

 from its original position. It will be observed that by 

 moving the recording surface further away, the magnification 

 may be still more enhanced. A wide latitude of magnifica- 

 tion may thus be obtained, by changes in the effective length of 

 arm of the lever, and by variation of the distance of the record- 

 ing surface. Thus, for example, for purposes of demonstration, 

 with a screen at a distance of five metres, it is easy to exhibit 

 a pulsatory movement magnified to as much as one metre in 

 amplitude. But in the case of the illustrations in the present 

 book, it has not been found necessary to have any magnifica- 



