312 James Elliot on certain 



upon its peg, or axis, to admit of its being raised or lowered 

 at pleasure. I also confined it to one place by forming a 

 small cavity on the support for the point of the peg to run 

 in. This being done, I was much pleased to find my expec- 

 tations exactly realized. By adjustments of the screw the 

 conical revolution could be quickened, retarded, annihilated, 

 or reversed, as might be desired ; and all its motions were 

 brought under perfect control. At the same time it was sur- 

 rounded by a fixed plane to represent the ecliptic, its own 

 equator being marked upon it ; and, by forming the axis of 

 hard steel, and giving it a support of agate, its velocity could 

 be kept up without much abatement for a long time.* 



The rotation is produced in the ball by means of a string 

 and handle, much in the same way as that in which a hum- 

 ming-top is spun. 



The case in which, from the two centres coinciding, the 

 axis remains fixed in one direction without any conical re- 

 volution, enables us to illustrate clearly what is meant in 

 astronomy by the Parallelism of the Earth's Axis, since the 

 model may be carried by the hand slowly round in any cir- 

 cular or elliptic orbit, without any perceptible deviation of 

 the axis from its original direction. 



But, when the centre of gravity is brought slightly below 

 the point of support, we are then enabled to show the devia- 

 tion from parallelism which arises in the direction of the 

 earth's axis after a long period of years, the same motion 

 exhibiting the Precession of the Equinoxes. With the centre 

 of gravity so placed, if the ball is made to rotate in the di- 

 rection marked by the upper arrow, on the figure, or from 

 west to east, the equinoctial point, E, is observed to move 

 slowly in the direction marked by the lower arrow, from east to 

 west. The latter motion may be made as slow as we please ; 



* Since the model described was constructed, my attention has been directed 

 to Bohnenberger's instrument for the same purpose, of which I was not pre- 

 viously aware. While that instrument is exceedingly beautiful, and adapted 

 to various experiments on rotatory motion, for which the model described above 

 is not intended, it wants (as will readily be admitted) the simplicity and capa- 

 bility of precise adjustment of the latter, and is not so well adapted for the 

 particular purpose of astronomical illustration. 



