296' lioyal Society. 



one or the other more convenient), a plane mirror is fixed, the sur- 

 face of which, by means of two adjusting screws, may be placed per- 

 pendicular to the axis of the needle j but scrupulous attention need 

 not be paid to this adjustment, as any deviation may most exactly be 

 measured by the observations themselves, and taken into account as 

 errors in collimation. The needle thus balanced is enclosed in ft 

 wooden cylindrical box, which, besides the small aperture in the lid 

 for the passage of the thread, has a larger one in the side, which is 

 rather higher and wider than the mirror already mentioned. 



Opposite to the mirror, a theodolite is placed, the vertical axis of 

 which is in the same magnetic meridian with the thread of suspension, 

 and at a distance from it of about sixteen Parisian feet. The optical 

 axis of the telescope is placed rather higher than the needle, and in- 

 clined in the vertical plane of the magnetic meridian, so as to be 

 directed towards the centre of the mirror on the needle. 



To the stand of the theodolite is fixed a horizontal scale of four feet 

 in length, divided into single millimetres : it makes a right angle with 

 the magnetic meridian. That point of the scale which is situated in the 

 same vertical plane with the optical axis of the telescope, and which, 

 for the sake of brevity, may be denominated the zero point, is marked 

 out by a fine thread of gold depending from the middle of the object- 

 glass, and charged with a weight. The scale is fixed at such a height 

 that the image of a portion of it is seen in the mirror through the 

 telescope, the eye-glass of which is adjusted accordingly. At the 

 opposite side from the needle, in the pame vertical plane, and at a 

 distance from the telescope equal to that of the image, a mark is fixed, 

 serving every instant to ascertain the unchanged position of the theo- 

 dolite. 



It is obvious, that if all these conditions be fulfilled, the image of 

 the zero point on the scale will appear exactly on the optical axis of 

 the telescope, and that, so far as an object of known azimuth is visible 

 at the place of the theodolite, we may, by means of this instrument, 

 immediately find the absolute magnetic declination. If, on the other 

 hand, those conditions are only partially fulfilled, then, generally 

 speaking, the image, not of the zero point, but that of another point 

 of the scale, will appear on the optical axis j and if the horizontal 

 distance of the scale from the mirror have been measured with exact- 

 ness, it will be easy to reduce the amount of the divisions of the scale 

 to the corresponding angle, and thus to correct the result first obtain- 

 ed. By turning the needle in the stirrup (so that the upper surface 

 becomes the lower), the amount of the error of collimation of the 

 mirror may be ascertained with great ease and precision. In both the 

 apparatus, one part or division of the scale is equal to nearly twenty- 

 two seconds j an interval which even the least practised eye may 

 easily subdivide into ten parts. 



By this mode of operating, therefore, the direction of the needle 

 and its variations are determined with the greatest possible precision. 

 It is by no means necessary always to wait till it is at rest j as the two 

 elongations to the right and the left may be observed with great accu- 

 racy, and their combination, properly managed, will indicate the cor- 



