Electro-Measuring Instruments. 311 



of the deflecting-field and also into a weaker part of the con- 

 trolling-field) is the one we find the best, and it is the one we 

 are using in all our latest experiments, the interior of one of 

 which is shown in fig. 3. 



Change of Strength of the Controlling-Magnet. — In some of 

 our instruments we have found that the constant did not 

 suffer sensible change during a period of many months ; in 

 others that the change during a much less time, arising from 

 the falling-off of the strength of the controlling-magnet, was 

 very serious. This latter effect we believe to be partly due to 

 inferiority in the steel, it being very difficult to obtain succes- 

 sive deliveries of magnet-steel from the same manufacturer 

 equally good in quality: and partly from the way in which the 

 armature, attached to the instrument when not in use, was 

 removed from the magnet when about to be used. In some 

 of our older instruments, from the form of construction, the 

 mode of removal of the armature took rather the form of a 

 pull, which is known to be bad for the permanency of a 

 magnet. Consequently our new Non-Commutator Instru- 

 ments are so arranged as to make it impossible to use arma- 

 tures at all, while in our Commutator Instruments armatures 

 are attached ; but the instrument is so constructed that the 

 armature A (fig. 2) can only be removed by a pure sliding 

 motion ; and we shall thus be able to ascertain experimen- 

 tally whether no armature at all, or an armature put on and 

 removed by a pure sliding action, is the better for keeping the 

 controlling-magnet constant in its strength. We anticipate, 

 however, that no armature at all will be the preferable plan. 



Olimmeters. 

 An ohmmeter, as we have already said, we define to be an 

 instrument in which the pointer points at once to the number 

 of ohms required to be measured. The method we employ 

 for making such an instrument consists in fixing two coils, at 

 right angles to one another, acting on the same soft-iron 

 needle. One of these coils, having terminals TT (PI. XII. 

 fig. 6), is made of thick wire, and is placed in series with the 

 resistance to be measured ; while the other, having terminals 

 1 1, is composed of very fine wire, and is put as a shunt to the 

 unknown resistance. Hence the main current produces its 

 effect by means of the thick wire coil, and the difference of 

 potentials at the terminals of the unknown resistance on the 

 fine wire coil. Now by properly proportioning the shapes of 

 coils, and by winding these coils in a definite way, we have 

 obtained an instrument in which the deflection of the pointer 

 is exactly proportional to the number of ohms, and which we 



