The investigation of this subject was rendered possible by the successful device of my Oscillating Recorder. A very light glass fibre was used for the construction of the lever, which was supported on jewel bearings. The short arm of the lever was 2 cm. in length, and the long arm 8 cm. This gave a magnification of 4 times. But it is quite easy to increase the magnification to 10 times or more. The pull exerted by the pulsating leaflet is extremely slight, and the relatively heavy lever made of steel wire used in the Resonant Recorder is not well-suited for our purpose. The pulsation of the leaflet is relatively slow, being once in two minutes or so. The intermittent contact of ten times in a second, given by the Resonant Recorder, is therefore too quick. In the Oscillating Recorder the intermittence was, therefore, reduced to once in a second, or once in five seconds, the recording plate itself being made to move to-and-fro at this rate. The carrier of the plate-holder slides backwards and forwards on ball bearings; a wheel in the clockwork connected with an eccentric is released periodically, at intervals which may be varied between one and five seconds. By the action of the eccentric, the plate carrier approaches the writing lever with diminishing speed till the movement is zero at the contact. This contrivance is essential, since any sudden shock of the plate against the lever is apt to give rise to after-vibrations of the writer. The plate carrier is quickly withdrawn after the production of a dot on the smoked glass plate by contact with the writing lever. The clockwork is governed by a revolving fan which can be gradually opened out by a regulating screw. The speed can thus be adjusted within wide limits, and maintained constant and at any desired speed. A second set of wheels connected with the clockwork moves the plate-holder in a lateral direction. A series of records may thus be taken for fifteen minutes, half an hour, or an hour. The record obtained in this way is very perfect. Not only is the effect of an external agent shown by variation in the amplitude and frequency of pulsations, but the change of speed in any phase of the pulse becomes automatically recorded. RECORD OF PULSATION OF _DESMODIUM GYRANS_. The whole plant can not be conveniently manipulated for different investigations. It is, however, possible using the precautions described below to use the detached petiole carrying the pulsating leaflets. The terminal large leaf may also be removed. The necessary amputation is often followed by an arrest of pulsation. But as in the case of isolated heart in a state of standstill, the movement of the leaflet may be revived by the application of internal hydrostatic pressure. Under these conditions, the rhythmic pulsations may easily be maintained uniform for many hours.