PULSE DURING MENTAL AND PHYSICAL WORK. 191 



of the pulse-curve by their mass, these contacts proved to be unsafe. 

 Occasionally no proper contacts were made. Occasionally the platinum 

 points stuck and failed to break. To avoid these difficulties we used 

 a mercury cup from which each systole raised the end of a fine platinum 

 \vire that was attached to the end of the tambour lever. This lever, 

 an aluminum arm about 10 cm. long, was counterbalanced by a bit 

 of wax, so as to produce the greatest amplitude of movement. 



When properly adjusted this apparatus functioned fairly well for 

 normal subjects. Operated chiefly by the capillary pulse, however, 

 it works for some subjects better than for others, and at some times 

 better than at others. On the whole we found it an exacting instru- 

 ment. In the first place, it must be carefully adjusted to each indi- 

 vidual subject, and the optimum relative position of the clamp, hand, and 

 transmitting capsule must be experimentally determined. In general 

 we found it better to leave the thumb free, so that only the fingers and 

 the palm of the hand were in contact with the capsule. The pressure of 

 the clamp that is necessary to get satisfactory results also varies with 

 the individual and his blood-pressure. The best conditions must be 

 found by trial. In cold weather the hand must first be warmed to secure 

 adequate capillary circulation. A third adjustment was the height 

 of the mercury cup. Our experience leads us to say that the optimum 

 height of the cup, when the systolic wave breaks the contact, is where 

 the contact is just broken when the whole system is in equilibrium. 

 If the device is so arranged that the systolic wave shall make the 

 contact, the surface of the mercury should be about 2 mm. below the 

 point of equilibrium. This latter arrangement has some practical 

 advantages, but it is theoretically less satisfactory than the former on 

 account of the normal fluctuation in the height of the systolic wave and 

 the consequent differences in the relative position of the moment of 

 contact. In the final form of this device we used only one mercury 

 cup, carrying the current through the axis of the recorder. We tried 

 using two mercury cups, connecting them with a transverse platinum 

 wire at the end of the recorder. That proved to be inexpedient because 

 of the relatively large, though intrinsically small, surface-tension 

 between the mercury and the platinum wire. Even with one cup, if 

 the mercury is a little too high, the surface tension is sufficient to keep 

 the platinum wire in contact, with consequent failure to record the 

 pulse. A fourth adjustment was necessary in order to avoid the 

 plethysmographic effects, by which the pointer either rises above the 

 mercury with increasing volume of the hand and fails to return to its 

 surface in the diastolic phase of the pulse-wave, or the pointer sinks 

 below the level of the mercury with decreasing volume of the hand and 

 fails to rise above its surface in systole. To avoid these plethysmo- 

 graphic changes we introduced two systematic leaks, as follows: 

 (1) the tambour membrane was pricked by a pin point so as not to be 



