324 PROCEEDINGS OF THE AMERICAN ACADEMY 



for two secondaries and one primary. Though the new method of 

 procedure was open to the same objections as the old, the figui-es all 

 agreed, within sufficiently narrow limits, both with themselves and 

 with those before obtained. A telephone and musical contact-breaker 

 were now substituted for the galvanometer and key. It was expected 

 that the two sets of waves induced by a single primaiy in two second- 

 aries, or by two primaries in a single secondary, could be made to 

 interfere. 



It was, of course, jDOSsible (uj)on any theory) so to place the two 

 coils that each should neutralize very nearly the effect of the other, 

 the telephone then being silent, and any considerable deviation from 

 this position of the two coils must necessarily be accompanied by the 

 production of a musical note ; but, after long and unsuccessful search 

 for other maxima and minima, the inevitable conclusion was reached 

 that no true phenomenon of magnetic wave-interference could be 

 found, and it was feared that all the previous results might have been 

 delusive. 



It was afterward discovered that, although the wave suffers a re- 

 tardation, this retardation diminishes with the wave-length, so that it 

 cannot produce complete interference under any condition whatsoever. 

 "We shall see, in fact, that the phenomena which have been described 

 are perfectly consistent, the methods sound, and even that the rough 

 measurements of these earlier experiments may be considered as 

 close approximations to the truth. 



The apparatus, principally employed in later and more precise deter- 

 minations (see Diagram), consisted of a wooden shaft, one fourth inch 

 in diameter, and about one foot long, carrying a pair of commutators 

 (cut from a convenient size of brass tubing), — one of which was of 

 the ordinary sort, the other different from it, in that its two arms were 

 greatly prolonged, and at the same time twisted, corkscrew fashion, 

 round the shaft one and one half turns each. The two lines of sep- 

 aration formed a pair of spirals of nearly constant pitch, so that by 

 means of sliding contact-pieces this commutator might be made to act 

 either simultaneously with its neighbor, or any fraction of a revolu- 

 tion sooner or later,* and this without stopping the machine. One of 

 the commutators was put in the primary, the other in the secondary, 

 circuit. The primary, or magnetizing coil, was of No. 16 insulated 

 wire, wound to a depth of half an inch over the first three inches of the 



* This fraction of a revohition is called, later on, the inclination of, or angle 

 between, tlie two commutators. 



