1858.] on Rotatory Stability, 483 



velocity is communicated by a separate train of wheels, from which 

 it can be instantly detached. Thus from the principle of fixity of 

 the plane of rotation, it was expected that the reflecting surface 

 would preserve its level, notwithstanding the motion of the ship. 



It may be proper here to observe, that simple as this idea was, 

 the practical application of it involved several instances of those 

 mechanical resources which its inventor could so readily supply. 

 One of these (of importance in all similar instruments) is the means 

 of obviating the unavoidable inequalities of density in different 

 parts of the most accurately cast metallic disk, as well as defects in 

 accuracy of form, which affect the perfect equilibrium. These 

 were remedied by plugs, or small pieces of metal, screwed more 

 or less deeply into holes made for them at opposite parts of the 

 circumference, three vertically, and three horizontally placed, by 

 which, in successive trials, a complete compensation was effected. 



The method was, however, found practically to fail ; and the 

 failure has been since traced to another mechanical principle. 

 The pivot partakes in the irregular motion of the ship. When the 

 disk is not revolving, this motion is in turn communicated to the 

 disk ; and the centre of gravity being below, — the very circumstance 

 which gives it stability on land — causes it to acquire an oscillatory 

 movement. When in rotation, this does not show itself directly, 

 but is compounded with the rotation, and causes a precessional 

 motion, which is fatal to its use as a horizontal reflector. 



Hence, if the centre of gravity coincided with the point of sup- 

 port, as would be most readily done by suspending the revolving 

 disk in gymbals in the manner of Bonenberger's machine, this 

 cause of irregularity would be avoided. By this means it would 

 preserve its original level; but this would not necessarily nor 

 usually be the true horizontal level. 



To obtain the true horizontal point another contrivance is 

 necessary : this (as not being connected with rotatory motion) can 

 here only be briefly adverted to. It consists essentially in the bub- 

 ble of a spirit-level reflected from a diagonal mirror above it, seen 

 through a coUimating lens, so adjusted that in one inclination the 

 apparent position of the bubble coincides accurately with the true 

 horizontal direction : whence it is easily demonstrated that in 

 every other inclination within certain small limits, the apparent 

 place of the bubble will deviate from its former place by the same 

 angle as that by which the base is inclined, or will in all positions 

 give the true horizontal direction : the level-tube being bent into 

 a small arc of a circle, whose radius is equal to the principal focal 

 length of the lens.* 



But for other classes of observations on board ship which involve 



* This contrivance has been fully described in the Notices of the Royal 

 Astronomical Society, Vol. xviii., p. 65. Jan. 1858. 



