166 



results obtained experimentally with the magnetic force of the earth. 

 For this purpose he reverts to the thick wire galvanometer before 

 described, and points out the precautions respecting the cleanliness 

 of the coils, the thickness and shortness of the conductors, the per- 

 fect contacts, effected either by soldering or cups of mercury ; and 

 marks the value of double observations, i. e. observations afforded 

 on both sides of zero. The nature of the impulse on the needles is 

 pointed out ; being not that of a constant current for a limited or 

 unlimited time, but of a given amount of electricity exerted, either 

 regularly or irregularly, within a short period ; and it is shown ex- 

 perimentally that such impulses produce equal results of deflection, 

 and also that when two or more such impulses are given within a 

 limited time, the whole arc of swing is nearly proportional to their 

 number; so that the amount of deflection, within certain limits, in- 

 dicates directly, nearly the proportion of electricity which has passed 

 as a current through the instrument. 



If a wire be formed into a square of 12 inches in the side, and 

 then fixed on an axis passing across the middle parallel to two of its 

 sides, and if, when that axis is perpendicular to the line of dip, the 

 whole is rotated, then two of the sides of the rectangle will, in one 

 revolution, twice intersect the lines of force of the earth passing 

 across or through one square foot of area. The currents then tend- 

 ing to move in the upper and lower parts of the rectangle, will con- 

 join to urge one current through the wire ; and if this wire be cut 

 at one place close to the axis, and be there connected with a com- 

 mutator of simple construction, which is described in the paper, the 

 currents round the rectangle may be conveyed away to the galva- 

 nometer, and there measured. Such a rectangle, constructed of cop- 

 per wdre one-twentieth of an inch in thickness, gave a certain arc of 

 swing for one revolution. If five or ten revolutions were made, 

 within the time of vibrating of the needle, nearly five or ten times 

 this amount of deflection was produced : the mean result, in the pre- 

 sent case, was 2°* 624 per revolution. When the same length of the 

 same wire was arranged in oblong or oblate rectangles, so as to 

 diminish the inclosed area in different directions as regarded the axis 

 of revolution, still the deflection was in every case proportional to 

 the areas included ; showing that the effect produced was propor- 

 tional to the number of lines of force intersected by the moving wire. 

 The same result was obtained when two squares having areas in the 

 proportion of 1 to 9, were employed. 



When squares of the same area were formed of copper wire of 

 different thicknesses, then the efi^ects of obstruction in the conducting 

 part of the system v/ere brought out and measured. Thus, with 

 wires which were 0"05, O'l and 0*2 of an inch in diameter, and 

 therefore in mass as 1, 4, and 16, the deflections were 1, 2*78, and 

 3*45 ; a result almost identical with that obtained for the same wires 

 by the use of loops and a local magnet in the former researches. 

 When two equal rectangles were compared, one containing a single 

 circuit of 4 feet of wire 0*1 in thickness, and the other four circuits 

 of 16 feet of wire 0*05 in thickness, then the first was found to 



