776 PROFESSOR JOHN G. M'KENDRICK 



The slow method of recording vibrations, therefore, whilst it is the method by which data 

 can be obtained that have to do with pitch and quality, will fail in giving us a record of 

 variations in intensity. This aspect of the matter came under my notice at an early 

 period of the investigation. So far as I am aware, no one has attacked this side of 

 the problem. Nothing is more striking in listening to the phonograph when it is 

 reproducing either human speech or musical sounds than the way in which it catches 

 every inflection of the voice or the slightest emphasis, diminuendo, and crescendo of 

 the sound. This must be due to variations of pressure. How may these variations be 

 recorded ? 



24. The most evident method is to attempt to record mechanically the variations 

 in an electro-magnet produced by pressures on a variable resistance apparatus in the 

 same circuit. My first attempt was to place Graham's transmitter over the glass disk 

 of the phonograph and to place in the same circuit an electro-magnetic marker such as 

 is used for physiological purposes. This gave poor results, but still they were 

 encouraging. On placing a Breguet's chronograph in circuit the results were much 

 better, and it was evident that there was a movement of the vibrator of the chrono- 

 graph for each note or chord emitted by the phonograph. I then heard of an ingenious 

 apparatus devised by Heurtley of Breslau, by which he has succeeded in recording by 

 electrical and mechanical arrangements the sounds of the heart. His apparatus consists 

 essentially of a large stethoscope on which a peculiar resonator is fixed. The resonator 

 carries a small wooden tuning-fork, between the prongs of which is fixed a simple micro- 

 phonic contact of two carbon buttons. This is one half of the apparatus. The other half 

 consists of an electro-magnet, over the poles of which is fixed, face downwards, a shallow 

 tambour, of the Marey pattern, having on its under surface a broad ferrotype plate. 

 This tambour is then connected with an extremely delicate recording tambour. When I 

 heard of this apparatus I at once saw that the second half of it was exactly what I 

 wanted for the phonograph work, and, by the kindness of Professor Heurtley, the 

 apparatus was made for me in Tubingen without delay. When placed in the circuit 

 along with the carbon transmitter the pen of the recording tambour moves at right 

 angles to the line of revolution of the cylinder with each tone and chord played by 

 the phonograph. When the ear perceives tones of considerable intensity the lever 

 point is seen moving through a greater distance than when the tones are weaker ; con- 

 sequently we have a graphic record of the variations in intensity. If the recording 

 cylinder is timed to travel at the same rate as the cylinder of the phonograph, then the 

 curves on the former exactly correspond to the ensemble of the minute marks on the 

 latter corresponding to a particular variation in intensity. When the recording cylinder 

 is caused to travel as fast as the phonograph cylinder, the variation in the heights of 

 the curves recorded on the revolving cylinder is not so apparent as when the recording 

 cylinder travels more slowly. It is easy, however, to time the rate of revolution of both 

 cylinders by a chronograph. Thus I have found that when the recording cylinder is 

 travelling at such a rate that £ inch of surface corresponds to one-fourth of a second, an 



