MULTIPLE AND AUTOMATIC RESPONSE 279 



light to give a slightly magnified record of these movements 

 was some special kind of dry grass haulm, which combined 

 rigidity with excessive lightness. I have tried the finest 

 feathers from small birds, but these were not so efficient 



Fig. 1 3 1. ^Photograph of the Oscillating Recorder, reduced to one- 

 fourth the natural size. 



as the selected specimen of grass which I was so fortunate as 

 to obtain later. Though in this I was successful in obtaining 

 a recording-lever, there yet followed the difiiculty that owing 

 to its being non-magnetic it could not be thrown into 

 resonant vibration by the electro-magnet. I had therefore 



