44G PROCEEDINGS OF THE AMERICAN ACADEMY 



1°. If we admit the theory of solenoids, the action of parallel cur- 

 rents ought to be added, and the amount of magnetic intensity to be 

 increased. The reverse takes place. 



2°. When the currents of the coils are sent in opposite directions, 

 they ought to act inversely on the particular currents of the iron, and 

 the results should diminish each other. On the contrary, they are added. 



3°. The action of the bobbins should, in this case, be nothing at the 

 middle point of the bar. It is not so. We cannot say that there is, 

 at this point, a resultant pole, for it would manifest itself by a point 

 of repulsion. 



IVI. Jamin then states that these results seem to him to require a 

 modification in the theory of solenoids. 



Mr. D. Sears has (American Journal, July, 1874) measured the dis- 

 tribution of magnetism in an iron bar which formed the armature of 

 the cores of two coils, by sliding a coil of fine wire, whose terminals 

 were connected with a galvanometer, along this armatui-e, and meas- 

 urino- the instantaneous current induced in this secondary coil when 

 the armature was magnetized by sending a current through the pri- 

 mary coils. His results are opposed to those of Jamin. The case, 

 however, is not exactly that of Jamin, and I have therefore, after 

 repeating Mr. Sears's experiments with similar results, applied this 

 method of measuring the distribution of magnetism, by means of a coil 

 of fine wire, to Jamin's apparatus, as follows : — 



' I made a bar of soft iron, 50 cm. long, the core of two coils, as in 

 Jamin's experiment, and so connected the coils with a battery that a 

 current could be sent through a single coil, or through both^ coils in 

 the same or in opposite directions. ( )ne of Farmer's thermo-batteries 

 was used as a source of electricity, because of the very great constancy 

 of its current. A small coil of fine wire, like that used by INIr. Sears, 

 was arrauiied to slide along the bar, and its terminals were connected 

 with a Thomson's galvanometer. When a current was sent through the 

 primary coils, magnetism was induced in the bar, and this, in its turn, 

 induced an instantaneous current in the coil of fine wire, and so caused 

 a deflection of the galvanometer. Although the secondary coil was 

 parallel to the primaries, I found, by substituting a glass rod for the 

 iron bar, that the direct action of the inducing coils on the secondary 

 coil was exceedingly small, excepting when these were brought very 

 near together, which it was not necessary to do in this experiment. 



The method used in these experiments is more delicate than Jamin's, 

 as may be shown by constructing curves from the observations in Table 

 I., or by the smallness of the differences in the last column of that table ; 



