518 DR. LAMONT ON THE MAGNETIC OBSERVATORY 
differs most from those in previous use. We owe to Poisso 
the highly ingenious idea of employing two equations, one of 
which shall contain the product, and the other the quotient, of 
the magnetism of the earth and of that of the bar, and by this 
means of entirely eliminating the magnetism of the bar in the 
calculation, and reducing the magnetism of the earth to abso- 
lute measure. Poisson proposed to obtain the requisite data 
for the two equations by means of experiments of vibration. 
Christie showed, on the other hand, that whilst the product 
might be obtained with great ease by vibrations, the quotient 
was more conveniently furnished by means of deflections. 
M. Gauss, in his treatise entitled Intensitas vis magnetice ad 
mensuram absolutam revocata, has shown how to obtaina more 
accurate determination of these two equations, avoiding all con- 
stant errors. We find in this memoir measurements which show 
the degree of accuracy of which Gauss’s method is susceptible ; 
I subjoin these results, and also the determinations made by 
M. Kreil at Prague in 1840, with instruments precisely similar. 
Gottingen. Prague. 
1832. May 21 1°7820 1840. Aug. 21 1°81414 
oo ©6224 1°7694 ose) 22 OS 21s 
June 4 11-7718 coe 24° 195596 
oo ~ 24, 28 1-7625 ace 25, ¢ gOS 
July 23, 24 1°7826 28 1°90970 
oo. 25,26 1°7845 30 ~=1°93483 
Sept. 9 1°7764 Sept. 3  1°92898 
ee §=18 1-782] 
Klee T7965 
Oct. 15 1:7860 
A mere inspection of these numbers leads to the inference 
that the means hitherto employed for the measurement of the — 
intensity still leave much to be desired. The three first places 
of figures ought to agree (the effect of intervening changes 
of the earth’s intensity, which has been neglected in the calcula- 
tion, may show itself in the fourth place); but differences are 
found always in the third and sometimes in the second place. 
The principal reasons probably are,—1st, that only small angles — 
of deflection were used; and, 2nd, that in the observation of 
vibrations the cases were not always sufficiently closed against — 
currents of air. The experiments already narrated in this me-— 
moir have manifested that two such magnets never have a per- 
fectly parallel march, rendering it impossible to measure small 
