572 . REPORT— 1894. 



changes of temperature, a certain displacement would take place once for all, but 

 no rotation such as is required to produce the secular variation. The constantly 

 renewed displacement of the magnetic axis, if it is due to the suggested cause, 

 must be produced by a continuous change inside the earth. A possible explanation 

 suggests itself in the secular cooling of the earth, owing to which more and moi"e 

 of its iron becomes susceptible to magnetisation. This virgin iron will be free to 

 adjust its magnetic axis to the magnetic forces acting on it, and a continuous dis- 

 placement of the axis must result. It must be acknowledged, however, that, even 

 taking account of this possibility, it is difficult to see how a rotation of the axis can 

 be kept up for several revolutions. Some light might be thrown on the question 

 by an experimental investigation to discover a possible effect of great pressure on 

 the temperature at which iron loses its magnetic properties. Whether there are 

 any appreciable magnetic eifects due to electric currents outside the earth may, of 

 course, be established by an analysis of the magnetic forces on the surface of the 

 earth. I understand that A. Schmidt and Neumayer have already made much 

 progress with such an analysis, and I am preparing an independent calculation with 

 special reference to the outside forces. 



The results of this investigation may be stated tlius: — 



(1) The mechanical reactions on the earth of currents induced in space, assumed 

 to be a conductor of electricity by the rotation of its magnetic system, are insuf- 

 ficient to produce an appreciable lengthening of the day in historical times, unless 

 the conductivity lies within certain narrow limits. The absence of any marked 

 effect cannot therefore be brought forward as an argument against the conduc- 

 tivity of space. 



(2) The magnetic reactions of the same current, taken in conjunction with the 

 secular cooling of masses of iron inside the eartli, tend to produce a displacement 

 which in kind is the same as that actually observed in the secular variation. But 

 whether quantitatively the variation can be explained in this way is very 

 •doubtful , 



3. On the Construction of a Delicate Galvanometer. 

 By Arthur Schuster, F.R.S. 



Maxwell has shown how a galvanometer must be wound in order to produce 

 the maximum magnetic field for a given resistance. The present communication 

 is intended to show what the advantage gained would be if the winding which is 

 theoretically best could be adopted. Assuming the cavity in which the magnet is 

 suspended to be cylindrical, the sensitiveness of the galvanometer will vary with 

 the square root of the resistance, and inversely as the square root of the radius of 

 the cavity. The smallest angle which can be read without making the system 

 astatic will also depend on the diameter of the mirror ; a point which has been 

 neglected in the construction of some recent galvanometers. Taking the resistance 

 of the galvanometer to be ono ohm, the diameter of the cylindrical cavity to be 

 one centimetre, and the widtii of the reflecting mirror to be also one centimetre, 

 the smallest current which a galvanometer will show with a horizontal force of 

 0'17 is under different conditions as follows : — ■ 



Amperes 

 If wound In the theoretically best possible way . . . 160 x 10"" 

 If wound on a bobbin of rectangular section of the best 

 dimensions, the wire being uniform in each layer, but 

 changing from layer to layer so as to give the best results . 200 x 10^* 

 If wound on a bobbin of rectangular section with uniform 

 wire, the dimensions of the bobbin being the most favourable 2'14 x 10~* 



The advantage gained by having the winding in the way which is theoretically 

 ithe best is therefore not very marked. 



4. On the Minimum Current audible in the Telephone} 

 By Lord Rayleigh, Sec. B.S. 



^ Published in the Phil. Mag , xxxviii. pp. 285-295, September 1894. 



