TRANSACTIONS OF THE SECTIONS. 35 



care that the slack cord is placed clear of all obstacles, and that it is free from kiiilcs, 

 the operator holds the gyrostat steady, so that its case is prevented from turning, 

 ■while an assistant pulls the cord through by running, at a gradually increasing pace, 

 away fi-om the instrument, while holding the end of the cord iu his hand. Suili- 

 cient tension is applied to the entering cord to prevent it from slipping round on 

 the shaft. In this way a very great angular velocity is communicated to the fly- 

 wheel, sufficient, indeed, to keep it spinning for upwards of twenty minutes. 



If, when the gyrostat has been spun, it be set on its bearing-edge with the centre 

 of gravity exactly over the bearing point, on a smooth horizontal plane such as a 

 piece of plate-glass lying on a table, it will continue apparently stationary and iu 

 stable equilibrium. If while it is in this position a couple round a horizontal axis 

 in the plane of the flj'-wheel be applied to the wheel, no deflection of this plana 

 from the vertical is produced, but it rotates slowly round a vertical axis. If a 

 heavy blow with the flst be given to the side of the case, it is met by what seems 

 to the senses the resistance of a very stiff" elastic body, and, for a few seconds after 

 the blow, the gyrostat is in a state of violent tremor, which, however, subsides 

 rapidly. As the rotational velocity gradually diminishes, the rapidity of the tremors 

 produced by the blow also diminishes. It is very curious to notice the tottering 

 condition, and slow, seemingly palsied tremulousness of the gyrostat when the fly- 

 wheel has nearly ceased to spin. 



In the liquid gyrostat the fly-wheel is replaced by an oblate spheroid, made of 

 thin sheet-copper and lilled with water. The ellipticity of this shell in the in- 

 strument exhibited is yV — that is to say, the equatorial diameter exceeds the polar 

 by that fraction of either. It is pivoted on the two ends of its polar axis in bear- 

 ings fixed in a circular ring of brass surrounding the spheroid. This circle of brass 

 is rigidly connected with the curved polygonal bearing-edge which lies iu the 

 equatorial plane of the instrument, thus forming a framework for the support of 

 the spheroidal shell. In fig. 3 a section is represented through the polar axis to 

 show the ellipticit}', and fig. 4 gives a view of the gyrostat as seen from a point in 

 the prolongation of the axis. To prevent accident to the shell, when the gyrostat 

 falls down at the end of its spin, cage-bars are fitted round it in such a way that no 

 plane can touch the shell. 



The method of spinning the liquid gyrostat is similar to that desci-ibed for the 

 solid g^TOstat, differing only in the use of a very much longer cord and of a large 

 wheel for the purpose of pidling it. The cord is first wound on a bobbin free to 

 rotate round a fixed pin. The end of it is then passed two and a half times round 

 a little pulley, and thence to a point in the circumference of a large wheel to 

 which it is fixed. An assistant then turns the wheel with gradually increasing 

 velocity, while the frame of the gyrostat is firmly held, and the requisite tension 

 applied to the entering cord to prevent it from slipping round the pulley. 



Secular Illustration of the Laws of the Diffusion of Liquids. 

 By Sir W. Thomson, D.C.L., F.R.S. 



On a new case of Instahility of Steady Motion. 

 By Sir W. Thomson, D.C.L., F.B.S. 



On the Nutation of a Solid Shell containing Liquid. 

 By Sir W. Thomson, D.C.L., F.B.S. 



