182 



Prof. D. E. Hughes, 



[May 10, 



obliged to use similar means, such as torsion, and also what would be 

 a repetition of several experiments tried long since by others, if it was 

 not for the fact that the induction balance enabled me to observe all 

 the effects of stress on iron by means far more sensitive than any 

 hitherto employed, and notice effects which, without its aid, would 

 pass unperceived. Some idea of this extreme sensitiveness may be 

 inferred from the fact that if we balance two pieces of iron to a com- 

 plete zero, the addition or subtraction of the -500V00" P ar ^ °f iron, 

 or the smallest filing to a large balanced mass of iron, will give out at 

 once loud tones, which we can measure and appreciate. 



In my previous experiments upon molecular magnetism I made use 

 of an iron wire, but I have found that in all experiments upon 

 magnetism it is far preferable to use thin flat strips, known as hoop- 

 iron ; ' as we can then have any desired amount of surface in the form 

 of a thin flat bar. 



Jamin had previously recognised the importance of using thin and 

 wide steel pieces for his well-known magnets. We can easily mag- 

 netise them to saturation, and where torsion is applied the contour of 

 the spiral becomes visible. Again, not only is the magnetism more 

 equally distributed throughout their mass, but we are enabled to 

 study the conditions of neutrality without the constant hindrance of 

 superposed magnetism, as we constantly find in the case of a wire or 

 bar of iron. 



I cannot, within the limits of this paper, describe all the varied 

 forms of coils and methods of obtaining perfectly defined zeros, 

 which are necessary in any varied research with the induction balance; 

 description of some of these will be found in my already published 

 papers. I will now describe a simple form which will suffice for the 

 experiments herein mentioned, and the following diagram shows its 

 electrical communications. 



A coil, having a large aperture, is fixed to a board ; two small 

 abutments or supports, A' A", at a few inches distance on each side of 

 the coil, allow us to suspend or fix an iron band or strip passing 

 through the aperture, which then becomes the core of an electro- 

 magnet. This forms the essential portion of the apparatus. The 

 iron or copper strip rests upon the two supports A' A", which are 

 20 centims. apart ; at one of these it is firmly clamped by two binding 

 screws, while the opposite end at A" can turn freely. The strip of 

 iron J' J", upon which the researches are made, is 22 centims. long, 

 and of any desired width and thickness ; it is fastened by means of a 

 binding screw B" to the projecting end of an axle, which has a key 

 or arm G, serving as a pointer moving upon a circle, and giving the 

 degree of torsion which the wire may receive ; a binding screw 

 allows us to fasten the wire, after turning the pointer to any degree 

 of torsion, and thus preserves the required stress as long as is necessary. 



