9S Compass for accurate Observations [Aug. 



This needle took the direction of the true meridian, and not of 

 the magnetic meridian. Its oscillation from six in the morning 

 till two o'clock (even in cloudy weather) usually amounted to 

 three degrees. I observed its direction for three weeks. I have 

 not yet succeeded in finding an explanation of a result so extra- 

 ordinary as that of the fourth and sixth experiments. 



Article III. 



Description of a Compass for accurate Observations on the Mag- 

 netic Variation. By Col. Mark Beaufoy. 



(With a Plate.) 



The principal part of this instrument, viz. its needle, with 

 the box and divided arch, to measure the angle of variation, is 

 very similar to the instrument in the possession of the Royal 

 Society ; but it is greatly improved by the addition of a small 

 transit telescope, which very readily and accurately determines 

 the true meridian, or the zero from which the degree of the 

 variation is to be measured. This addition renders it particularly 

 well adapted for travellers who wish to make accurate observa- 

 tions in different parts of the globe, as it contains every thing 

 necessary for the observation within itself; and as every part 

 admits of adjustment, by reversing, it may be depended upon 

 for accuracy. 



Plate IX. fig. 1, contains a representation of the whole instru- 

 ment in perspective, explaining all its parts at once. The box con- 

 taining the needle is not fixed, as in many compasses, but turns 

 horizontally on a centre, and has an index fastened to it, point- 

 ing to a divided arch, on the brass frame on which it turns; and 

 the method of observing is to move the box till a line drawn on 

 it points exactly to the end of the needle, which being done the 

 angle of variation is shown by the divisions on the arch. F F is 

 a mahogany board, which is the support of the whole instru- 

 ment; it stands on the points of three screws, G, H, I, by which 

 it can be levelled. Above this is a flat plate of brass, E E, attached 

 to the board by a centre pin, and resting upon three studs, pro- 

 jecting from the board, to ensure its having a perfect bearing, 

 whilst it admits of a small horizontal motion round the centre by 

 means of the screw W X. The plate EE has the divided arch, m, 

 fixed upon one end of it ; and a vernier, D, traverses against 

 the divisions; it is divided upon a projecting part of a brass plate, 

 C C, which moves on the common centre pin of the plate E, 

 and also the needle A A B. The plate C C has two segments 

 of Lrass at each end of the needle, and these have the centre 



