PULSATIONS OF DESMODIUM GYRANS 293 



different modes of adjustment. One up and down, the 

 second lateral, and the third, by means of the hinged support, 

 for the inclining of the specimen. The movement of the 

 leaflets, it must be remembered, does not always take 

 place in a vertical plane. The object of these mechanical 

 adjustments therefore is to enable us to place the specimen 

 at such an angle that its to-and-fro vibration when straight 

 shall be vertical, or have its long axis vertical when the 

 movement is elHptical. 



A light cover with mica windows can be made to enclose 

 the specimen. By means of electric current sent through a 

 spiral of German silver, the inside of the chamber may be 

 heated to any desired degree. The temperature can be 

 lowered, on the other hand, by sending through the chamber 

 a stream of cooled air : different vapours or gases could be 

 passed into the chamber for studying their effects on the 

 automatic pulsation. 



The arm of the recording-lever is attached by means of a 

 cocoon thread to a point about the middle of the leaflet, by 

 means of a drop of shellac-varnish. As the pull exerted 

 by these leaflets is very feeble, the writer has to be made 

 extremely light. The vibration of the Resonant Recorder 

 being about 10 times per second, the record taken with it 

 appears as continuous. In certain experiments it is desirable 

 to obtain data for accurate time-measurements of different 

 phasic movements of the leaflet. This I have been able 

 to secure by the employment of the Oscillating Recorder, 

 where the recording-plate, by means of an electric motor 

 provided with an eccentric, is made to execute a to-and- 

 fro movement. The intermittent dots thus produced may be 

 once each second, or once in 2 seconds. As the oscillating 

 recorder permits the employment of light grass haulm for 

 the recorder, we may easily obtain a fair magnification pro- 

 duced in the record. I have used both the methods — 

 resonant and oscillating — for obtaining the records : in the 

 former they appear continuous ; in the latter, dotted. 



As an example of the extreme regularity which can be 



