60 Proceedings of the Royal Irish Academy. 



•experiments, the indicator and its'mica plates were always perfectly 

 balanced. 



Up to the present, water has been the liquid exclusively employed. 

 I hare sometimes used the common water supplied from the water- 

 works, but recently, distilled water. In making the experiments, 

 great care was necessary in the introduction of the mica plates into 

 the water. If water was a perfect fluid, the introduction of such 

 slender disks would be accompanied by no sensible resistance, but this 

 was so far from being true, that unless both plates were plunged into 

 the liquid at the same time, and to almost equal depths, the equili- 

 brium of the indicator became totally deranged by the difference of 

 the resistance experienced by these disks plunged edgeways. After 

 they had been plunged into the liquid, and allowed to adjust them- 

 selves, it was manifest that each disk exercised attraction on the 

 adjacent liquid, from the concave meniscus of liqiiid heaped up at 

 (?ach side. But this is precisely what is desired, for we wanted an 

 indicator moving as much as possible with the liquid, and immediately 

 partaking of its changes of velocity. 



If the air contained in the vessel were absolutely at rest during 

 the rotation, it would resist in a slight degree the motions of the 

 wooden cross-bar and of the pieces of the mica disks above water ; but 

 as it is certain that air is endowed with some viscidity, or, as some 

 prefer to call it, internal fi'iction, it must partake in some measure of 

 the motion of the vessel. If it did not, its tendency would manifestly 

 be to keep the indicator in its original place. In this case the moving 

 force acting on the indicator would be that of the strata of liquid 

 pressing against the mica disks, and the resistance, that of the air in 

 which the indicator was partly plunged. Prom the weight and 

 dimensions of the wooden indicating bar and the mica disks, which I 

 have already given, it is evident that the moment of inertia of the 

 whole is very small, and as yet I have not deemed it necessary to cal- 

 culate its exact amount. In reducing the results of observation, in 

 order to determine numerical co-efficients for the friction of the fluids 

 against solids, or for internal viscidity, this moment must be esti- 

 mated. 



The apparatus for making the experiments being adapted for 

 working with very slow or rapid motions, it has occurred to me to 

 divide the investigation into two parts. The first refers only to the 

 phenomena of slow rotation. By the adaptation of the pendulum, the 

 vessel containing liquid has been made to revolve at velocities of from 

 about one turn in two hours and a half to one turn in four hours, or 

 six turns in a day. 



The method of observation adopted was the following : The vessel 

 was filled with water to a depth of twelve centimetres, which enabled 

 the mica disks to plunge tlu'ee centimetres into the liquid, leading one 

 -centimetre between the surface of the liquid and the indicating bar. 

 After this had come to perfectly steady condition, the positions of the 

 mica disks were sighted on the graduated circle, and when the sus- 



