Recoil Curves as shown hy the Hot- Wire Microphone. 



291 



diaphragm displacement in the microphone container are here exceedingly 

 minute, and we are thus able to assume that they produce no appreciable 

 change in the electrical resistance of the microphone. 



We are able to show, however, that breathing does produce a marked effect 

 in systolic discharge, as will be clearly indicated in the records described 

 later. 



Experimental Work. 



The earliest records were obtained by the simple process of seating the 

 patient on a microphone container. It was found that the " spring " of the 

 walls sufficed to produce measurable effects. In order to avoid change in zero 

 of a balanced electric circuit, the resistance variations were conveyed to the 

 galvanometer through the agency of a transformer. 



The next development was that in which the patient was seated in a chair 

 which was bolted to a wall of the container at a point beneath its centre of 

 gravity. Both of these methods involved careful balance of the body, and 

 in any case could give no quantitative results. The chair was then slung by 

 chains from the under surface of the diaphragm in the apparatus described 

 earlier in the paper, and a foot rest was fitted so that the patient could be 

 seated at ease. Here, again, the effect, though more consistent, and though 

 to some extent capable of comparable measurement, depended on the attitude 

 of the subject, and records showed that marked differences were given when 

 the knees were bent to different extents, or when the head was erect or 

 bowed. Finally, the subject to be tested was placed on the swinging platform, 

 in an absolutely erect position, with a shoulder rest to steady the body. 



A change was also made in the electrical arrangements. It was found 

 possible to replace the transformer method by that of the simple Wheatstone 

 bridge. This latter arrangement enabled us to get more faithful records of 

 the resistance changes of the microphone. Contrary to expectation, the zero 

 of the record was found to be quite steady. 



Consideration and Analysis of the Curves. 



The Normal Recoil Curve. — The normal recoil curve is undoubtedly very 

 similar to that shown in fig. 5. This curve and the majority of those 

 reproduced in this paper have all been obtained from one subject 

 (Mr. A. Heading), whose general physical standards are equal to those found 

 in the best type of pilot. In addition, the electro-cardiogram and ortho- 

 diagrams of this case are normal (Plate 5). 



Very similar records, which need not be reproduced here, were given by 

 many other normal subjects. 



As will be seen from the figure, the main features of the normal curve 



