213 



to have been the production of an instrument that would, under all 

 variety of circumstances, give a correct (magmetic) meridian direc- 

 tion, or in some way indicate the amount of its own error ; and he 

 considers that he has, at least partially, succeeded in the attainment 

 of this object in the instrument described in this paper, and of 

 which drawings accompany the communication. 



In order to determine the nature of the action of a mass of iron on 

 a magnetic needle, the author constructed needles with the magnetic 

 bar M'hoUv on one side of the central support, counterpoised by an 

 arm carrying a weight on the other. In some of these the magnetic 

 bar was straight, in others it was bent in the middle, either the 

 upper or lower half being horizontal. He also constructed a com- 

 pound needle of two such bent bars, suspended concentrically with 

 two of their contrary poles remote, the other two poles being the 

 one above and the other below the points of support. In order to 

 give magnetic stabihty to the compound needle, an arc of soft iron 

 was attached to the extremity of the counterpoising arm of each 

 magnet, so as to be in close proximity, though not actually to touch 

 the other. Attached to the counterpoise arms were the halves of a 

 compass. card cut through the east and west points, and so adjusted 

 that the edge of one was vertically over that of the other when the 

 needles were in the same vertical plane, but crossed each other when 

 the needles deviated from the same du-ection. The author con- 

 sidered that a mass of iron attracting each pole of this compound 

 needle would cause the cards to cross each other, and thus give the 

 amount of deviation due to that mass, but found that, instead of the 

 cards crossing, the needles deviated in contrary directions so as to 

 remain in one straight line, as if they had been rigidly connected. 



To obviate this, for the magnet having its unmarked end pointing 

 south was substituted one having its mai'ked end in that direction, 

 but of less power than that of which the marked end pointed north ; 

 and two magnets rigidly connected in the same sti-aight line, with 

 their marked ends remote, were balanced concentric with the other 

 two. The author considers that the result of this combination would 

 be, that this astatic bar would, by the repulsion between its poles 

 and those of the other magnets, take up a position at right angles 

 to the magnetic meridian, the other magnets pointing due north and 

 south. Further, that the north pointing and south pointing ends of 

 the compound needle on the half cards beins: poles of the same name, 

 that is, both marked, any mass of iron which attracted one end 

 would also attract the other, and thus the half cards which they 

 carried would be caused to cross each other. So that the whole 

 being properly adjusted, the deviations caused by the iron on the 

 two needles will be equal, and the error of deviation of either one 

 win be half the angle between the two. A figure of this complicated 

 compass accompanies the description. 



In conclusion, the author states that the practicabilitv of rendering" 

 the instrument so simple and accurate as to be generallv useful, de- 

 pends on experiments and investigations yet to be 'made ; and 

 Proceedings of the Royal Society, ^'ol, VI. Xo. 90. ic 



