344 Mr. Macdonald's Observations of the diurnal Variation 
perpendicular, and strong screws working in two plates, on 
the principle of the levelling screws of a theodolite. The cy- 
linder was. rendered perfectly perpendicular to the level plate, 
by means of three small adjusting screws on the tube. These 
screws acting laterally on the brass wire-pointed pin that gave 
the sun's shadow on the concentric circles, brought it perpen- 
dicular over the centre of the plate. A flat small brass' ruler 
was furnished with a socket, which received the point of the 
shadow- pin. This ruler had a horizontal arm to it, having a 
semicircle at one extremity coinciding with the tube on the 
centre of the plate. A pointed wire passed to the plate 
through the other extremity of the horizontal arm. This wire 
could be raised or lowered at pleasure by a small screw. It is 
evident that if this point touched, delicately , the plate of brass, 
in moving it round (by means of the socket above, and semi- 
circle playing round the tube below), the shadow -pin must be 
perpendicular. If the plate was acted upon evidently roughly 
by the pin in moving horizontally, it is clear the pin must be 
inclined to that side, and a delicate correction was given by 
the little brass screws fixed to the tube screwed on the centre. 
By this contrivance the shadow -pin was fixed exactly perpen- 
dicular to the plate. 
The box in which the magnet is suspended seems to have 
come out of the maker's hand unfinished, having on it no 
hairs through, or parallel to its axis, by which to apply it 
truly to a meridian. These hairs were fixed above and below 
the magnet, in a plane perpendicular to the graduated o° at 
both ends. Two very fine brass wires were fixed on the out- 
side, on the ends of the box. These brass wires were in the 
plane of the hairs, and when applied to the meridian, the haire 
