326 Mr Taylor, Experiinents with Rotating Fluids 



Experiments with Rotating Fluids. By G. I. Taylor F.R.S 



[Read 24 January 1921.] 



The present coi^munication contains a summary of results on 

 three subjects connected with the dynamics of rotating fluids. 



1. Experiments to illustrate the difference between two and three 



dimensional fluid motion. 

 The first experiment to be shown consists in towing a sohd 

 circular cyhnder of the same density as water through a tank 

 contammg water, the whole system rotating at a uniform speTd 

 iHe tank IS a circular glass crystalhsing dish which is filled three- 

 quarters full with water and floats in another dish of shghtly larger 

 diameter The inner dish is rotated by means of a ift of water 

 ottafned " '' *-^g-tially. In this way a unifoJm speldls 



The sohd cyhnder made of box- wood or wax is placed mth its 

 generators vertical, t.e. parallel to the axis of rotation. It is held 

 ma position close to the side of the dish by means of a device 

 which releases it at an appropriate moment. It is towed hori- 

 zontally through the tank by means of two threads which pass 

 through two small rings fixed to the dish on the opposite side to 

 the point to which the cyhnder is initially attached^ It is found 

 Stl .' ^,f ^df /?^«^es straight through the hquid, moving rela- 

 tively to the rotating system m the direction in which the thread 

 IS pu hng it so that it passes through the centre of the dish 

 _ When the same experiment is performed with a sohd sphere 

 instead of a sohd cyhnder, it is found that the motion is very 

 different The sphere is deflected and moves through the tank n 

 a curved path, leaving the centre of the dish well on its rfght 5 



by tie author* ^^''"'''"'' ''"^™ " theoretical prediction given 



is flfr^!^'' prediction which is verified with remarkable accuracy 

 IS that all small steady motions of a rotating fluid relative to the 



L thatffT*'^ T'*!^' 'T ^^-----1- A consequence of this 

 IS that If a spot of coloured water is placed in the rotating water 

 and If any small motion is communicated to it, the colouring 



Txi of rot.T"" 'k.T*' '\''' ^^'^ ^'' always' parallel ?o hf 

 rnffffn T ■ ^^ *^''^ '^""*' ^'^ «b«^^^ed along the axis of 

 rotation they appear as thin fines. This property is so strikingly 



* Proc. Roij. Soc. A. 1917, p. 99. 



