TRANSACTIONS OF TIIE SECTIONS. 
501 
it becomes indirectly a great source of inconvenience and error. 
From an extensive series of experiments with vibrating needles, 
in air and in a rare medium, he is led to conclude, that we can¬ 
not depend upon observations embracing extremely minute dif¬ 
ferences in the times of vibration, at different places, until our 
instruments of research have become more refined ; and that it 
is essential, in very delicate inquiries, to take the vibrations 
within very small arcs,—the least possible, so that they be con¬ 
sistent with accuracy of observation. With a bar of about 5 
inches long, Jth of an inch thick, and about T ^ths of an inch 
wide, vibrating in an exhausted receiver, from 200 to 300 vibra¬ 
tions may be obtained, before the arc of vibration becomes re¬ 
duced from 5° to 3°. The author thinks that the mean of 100 
vibrations, determined in the way recommended by Prof. Han- 
steen, and taken within an arc of 5°, is more to be depended on 
than a much greater number, taken in larger arcs. 
The following is a brief account of the apparatus employed 
by Mr. Harris for carrying on experiments, with vibrating mag¬ 
nets, in an exhausted receiver. 
There is a firm base, somewhat resembling in form a sector 
of a circle, sustained on levelling-screws; the diameter of this 
sector is placed as nearly as possible in the direction of the 
magnetic meridian : this base supports a parallelogram of wood, 
on which are fixed the magnetic apparatus and air-pump; the 
whole of the latter part is moveable, for a short distance, upon 
a pin concentric with the centre of the sector, and with the 
suspended bar and cord: by this is obtained the final adjust¬ 
ment, requisite to place the zero of the cord exactly in the 
meridian. 
The magnetic bar is suspended in a light vertical frame of 
wood, supported on a circular block, which last carries a gra¬ 
duated circle of stout card-board. The block is fixed, by means 
of a nut and screw, to a circular plate of fine-grained slate, 
having an air-tight communication, by a horizontal pipe, with 
the air-pump barrels; this plate of slate is found to answer 
admirably as a pump-plate,—thus avoiding a large mass of 
metal in the vicinity of the bar. 
The centre of the wood block is hollowed to about three 
inches of its diameter, in order to admit of the operation of a 
sort of forked lever, moveable through the centre of the pump- 
plate, in an air-tight collar; by means of this, the bar can be 
deflected and set free at any given point on the graduated cord. 
There is a gauge, of an extremely simple kind, for indicating 
the degree of exhaustion, and a thermometer, both attached to 
the wooden frame carrying the suspended bar, and the whole 
