■62 Proceedings of the Royal Irish Academy. 



slow rotation. "When the vessel rotated once in thirty, or even in 

 twenty seconds, after an interval of about seven minutes the indicator 

 moved with the same angular velocity as the vessel. When the mica 

 plates had been long immersed in the water, this interval was 

 reduced, showing that the water took time to completely adhere to 

 the mica. 



As the radius of the vessel is 132 millimetres, the disks were 

 plunged in water at 42 millimetres from the glass sides in experiments 

 1, 2, 3, 4, 5, and 9. They were at 72 millimetres in experiments 6, 

 7, 10, and 11, while in experiment 8 they were at 92 centimetres 

 from the glass. In this way they enable us to observe the relative 

 velocities of the strata situated at very different distances. Yet in all 

 of these observations, from the moment when rotation commenced 

 until its conclusion, the index continued to steadily point to the gra- 

 duated circle. The adherence of the water to the mica disks was all 

 along distinctly manifest. The water also adhered to the glass at the 

 sides of the vessel, and all through the fluid the adjacent particles 

 were attracted to each other. With slow velocities of rotation, these 

 actions seemed to suffer no disturbance : each successive stratum of 

 Avater, from the sides of the vessel to the mica plates, revolved with 

 precisely the same angular velocity, and thus the vessel and its con- 

 tained liquid rotated as one mass. 



Besides the experiments recorded in the foregoing Table, I have 

 made several others with slow and quick velocities ; the former all lead 

 to the same result as those recorded, and I reserve the discussion of 

 the latter for the second part of my Report. The observations I have 

 already made, and those presented by other inquirers, clearly establish 

 tliat natural fluids possess internal viscidity to such an extent as to 

 vitally influence conclusions drawn from the mathematical treatment 

 ■of problems in hydi'O-mechanics. Capillary phenomena, where the 

 molecular action of solids and liquids at small distances is so strik- 

 ingly manifested, appear to have led to the notion that, because large 

 and uncapillary vessels do not clearly manifest such actions, the 

 motions of liquids within them could be considered as independent of 

 molecular action. But the force of cohesion, from particle to par- 

 ticle, exists throughout the whole of a mass of liquid, no matter how 

 great, and hence its motions are governed by the specific amount 

 ■of this force. It is singular that the idea of what is called a perfect 

 iluid should be so constantly in the minds of mathematicians who 

 have treated hydro-mechanical questions, when one of the best-known 

 ■elementary problems implicitly supposes a totally different fluid pro- 

 perty. A cylindrical vessel containing liquid is set into very rapid 

 rotation : required the shape of the concave surface of the liquid. The 

 ordinary equations of hydro-dynamics are used for solving the problem, 

 but the hypothesis is made that the angular velocity of each particle 

 of the liquid is the same. This could not be true in a perfect liquid. 

 The cylinder would tend to slip past the particles close to it, and if 

 these acquired any velocity from the roughness of the cylinder, the 



