[ 251 J 



XXIY. Experiments with Rotating Viscous Liquids. By 

 John Buchanan, JD.Sc. {Land.), and Henry W. Malcolm, 

 B.Sc. Carnegie Scholar of Aberdeen University*. 



[Plate IV.] 



f I ^HE object of these experiments was to study the behaviour 



X of some viscous liquids when placed in a cylindrical 

 drum which is subjected to angular acceleration about a 

 horizontal axis. 



Let us imagine such a drum to be mounted similarly to 

 the wheel of an Atwood's machine, and, like the wheel, to be 

 caused to rotate about a horizontal axis by means of a falling 

 load attached to a cord wrapped round the axle. Provided 

 the liquid fills the drum completely, it is not difficult to 

 see beforehand, in a general way, what kind of results may be 

 expected. 



When the drum is undergoing angular acceleration bv the 

 pull of the load on the cord, the more viscous is the liquid, 

 the more closely must it follow the motion of the drum. The 

 viscous is the liquid, however, so much the more slowly 

 will the various parts of the liquid take up the motion of the 

 drum. In other words, there will be more relative motion 

 between the neighbouring parts of the liquid, and therefore a 

 greater retarding effect, in the latter case. 



Experiment bears out these conclusions. The motion of 

 a drum filled with castor-oil could hardly be distinguished 

 from the motion of a solid body; with the drum full of water, 

 the presence of the liquid showed itself in a very marked 

 degree. 



The conditions are completely altered, however, if the 

 drum be only partially filled with liquid. As the rotation of 

 the drum round a horizontal axis proceeds, there will be a 

 continual lifting of the centre of gravity of the liquid by 

 an amount depending on the viscosity of the liquid, as 

 the various layers cling to one another and to the walls of the 

 drum. Clearly, the greater is the viscosity of tbe liquid, the 

 greater will be the retardation due to this cause, — except, 

 indeed, in the extreme case where the motion is so rapid that 

 the liquid is maintained, by centrifugal force, as a ring of 

 invariable form carried bodily round with the walls of the 

 drum. 



Under ordinary circumstances the motion inside the drum 

 18 BO exceedingly complex that only by experiment can the 

 amount of retardation due to a specified volume of a given 



* Communicated bv the Authors. 



