under the Operation of Feeble Magnetic Forces. 243 



recent evidence, that saturation is attainable only in a very 

 rough sense. The question was too important to be left un- 

 decided, but it was difficult to deal with by the magneto- 

 metric method. If the arrangement is sensitive enough to 

 allow the effect of the small force to be measured with reason- 

 able accuracy, it is violently disturbed by the occurrence of 

 high degrees of magnetization. Moreover it is undesirable 

 to depend so much, as in this method, upon what may happen 

 near the free extremities of the iron rod, where the magnetic 

 forces must vary rapidly. The " ballistic method," in which 

 the changes of magnetization are indicated by the throw of 

 a galvanometer-needle in connection with a secondary coil 

 embracing the central parts of the rod, has the great advan- 

 tage for this purpose, that the reading is independent of the 

 stationary condition of the iron. In the first experiments by 

 this method the magnetizing helix was similar to one already 

 described (p. 234); and the small, as well as the large, altera- 

 tions of force were effected by varying the resistance of the 

 circuit. By suitably choosing the resistances from a box, the 

 small alterations of current could be obtained with sufficient 

 suddenness by the simple introduction or removal of a plug, 

 and were taken of the same order of magnitude at different parts 

 of the scale. A comparison of effects (with the aid of a table of 

 reciprocals) proved that a pretty strong total force * or mag- 

 netization did not interfere much with the response of the 

 iron to a given force of small magnitude. 



This arrangement did not well allow of the investigation 

 being pushed further so as to deal with stronger magnetizing 

 forces. If, with the view of increasing the current, we cut 

 down the german-silver resistance too closely, the estimate of 

 total resistance depends too much upon the battery, and the 

 current becomes uncertain. This difficulty is evaded by the use 

 of a double wire — one conveying the strong current, of which 

 the measurement does not require to be very exact; the other 

 conveying the weak current, of which the effect at different 

 parts of the scale is to be examined. 



In order to obtain a satisfactory ratio of length to diameter, 

 without the loss of sensitiveness that would accompany a 

 diminution in the section of the iron, a helix was prepared 

 of length 59*6 centim. It was wound upon a glass tube 

 with a double wire in three layers, the whole number of turns 

 of each wire being 1376. The magnetizing force due to unit 

 current in one wire is therefore 



4ttx 1376/59-6 = 290-1. 



* Up to about 6 C.G.S. The iron was uiiannealed Swedish, 3*3 millim. 

 in diameter. 



S2 



