Intelligence and Miscellaneous Articles. 159 



that the variation of the attraction should not be visible on the bifilar 

 magnetometer. But it was evident also that an arrangement might 

 be made sufficiently sensible to show this variation ; and I tried in 

 1858-59, by means of native workmen, to make an apparatus in 

 which the magnetic force would be replaced by a constant torsion- 

 force. 



This trial did not succeed, on account of defects of workmanship, 

 and I was obliged to wait till my return to Europe. 



In June 1860 I communicated the principle of the apparatus to the 

 optician already named ; afterwards I sent him the necessary draw- 

 ings, and I believe the instrument was made in the first months of 

 1861. We found, however, that the spring (like the balance-spring 

 of a watch) employed under the weight acted badly, and I substituted 

 a simple gold wire in the same year. 



- I ought to add to this, that in the end of 1860 I sent a descrip- 

 tion of my instrument to be communicated to the Royal Society of 

 Edinburgh ; that the instrument was seen in 1861 by several persons 

 in Mr. Adie's workshop ; that I mentioned the apparatus and its 

 principle to different persons in Paris ; and I owe to M. Henri Robert, 

 the well-known watchmaker, the gold wire which he recommended 

 to me as the best for the objects of my instrument. 



1 ought, before ending, to notice a difference in M. Babinet's and 

 my method of observing. 



My apparatus consists of a cubic weight suspended by two gold 

 wires ; three sides of the cube have mirrors attached, the planes of 

 the two on opposite sides making each an angle of 93 degrees with 

 the plane of the third ; a single gold wire is hung from the centre of 

 the lower side of the cube. (I have also noticed the plan of fixing 

 the wire higher, given by M. Bab'net.) This wire supports a lever 

 by the middle, which lever can be turned by the hollow cylinder in 

 which the lever hangs without being fixed. On the exterior of the 

 cylinder there is another mirror. The instrument is adjusted in such 

 a way that a revolution of the cylinder (and of the lower wire) will 

 produce a movement of the weight of about 90 degrees ; the number 

 of degrees less than 90 depends upon the degree of accuracy desired, 

 and of- which the instrument is capable. 



It is evident that if W be the weight suspended and AW any 

 variation produced by a change of graviry, then 



AW 



-— — = cotan v Av ; 

 W 



where v is the angle of rotation of the weight at a given station or 

 time, and At; the variation of this angle at any other station or time, 

 the force of torsion being supposed constant. 



Thus, if we make v near enough to 90 degrees, we may perceive 

 an exceedingly small variation of W. 



A small telescope furnished with a prism of glass behind the wire 

 in the eyepiece shows accurately the coincidence of the images 

 reflected from the mirrors and of the wire seen directly by the eye- 

 piece (according to the ingenious arrangement in Dr. Lamont's 

 instruments) ; and two glass scales, one above the other, are em- 



