OF ARTS AND SCIENCES. 233 



employed in the usual manner to study the motion of a very minute 

 glass head mounted upon the diaphragm of a magneto-receiver, 

 through the coils of which was carried an alternating current. The 

 glass bead was illuminated by sunlight or the electric arc. The 

 rate of the fork was 128 complete vibrations per second. The alter- 

 nating current could be caused to flow through the telephone coils 

 for a definite time by the same circuit-making wheel which was 

 used in the experiments already described. As we desired to use a 

 far stronger current than that produced by the transmitting tele- 

 phone, we made use of the alternating current from a transformer 

 excited by a Thomson-Houston alternating current machine, mak- 

 ing 128 complete alternations per second. The strength of the 

 current through the coils was varied by a set of resistances made of 

 incandescent lamps from a maximum of about 100 milliamperes to 

 a minimum of about 50 milliamperes, an amount vastly in excess 

 of any telephone current. When the circuit-making wheel was at 

 rest and in the proper position the curves of Lissajous were clearly 

 seen. The motion of the bead was very nearly a simple harmonic 

 one, so that the curves observed were ellipses, passing into the 

 limiting oblique straight lines. Difficulty was met with from 

 variations in the rate of alternation of the current, arising from 

 very slight variations in the speed of the dynamo machine, so that 

 it was necessary to utilize such moments as were found to be 

 available when the steadiness was sufficient. It the circuit-mak- 

 ing wheel is in motion, it is clear that the appearance of the curve 

 seen will depend upon the duration of the make, or rather upon the 

 duration of the motion of the telephone diaphragm. If the har- 

 monic motion of the diaphragm persists for a period equal to one 

 complete vibration of the fork of the comparator, — or, more strictly, 

 for a period somewhat longer than this, and depending upon the 

 relative phase of the two vibrations when the current begins to 

 flow through the telephone, — it is evident that the curve should be 

 a complete ellipse; but if the. harmonic motion of the diaphragm 

 lasts for a less time, then only a portion of an ellipse can be seen; 

 and this will necessarily be deformed, as the curve must begin and 

 end on the same vertical line; which is, of course, the vertical line 

 given by the illuminated bead, as seen through the microscope of 

 the comparator when no current is passing through the telephone. 



The method has not yet been sufficiently perfected to give accu- 

 rate quantitative results ; but with the rate of alternation as stated, 

 viz. 128 per second, and with the strong current used, in some 



