86 Dr. C. V. Burton on a 



There is neither attraction nor repulsion when the common 

 velocity of the two bodies is inclined at an angle tan~ 1 - v /2, 

 or approximately 54° 44' to the line of centres. But in 

 addition to the forces in the line of centres, there are in 

 general forces perpendicular to this line, and if we imagine 

 the two bodies connected by a rigid immaterial (or material) 

 bar, the tendency will be to set the line of centres perpen- 

 dicular to the direction of motion ; the couple acting on the 

 system only vanishing when the direction of motion is parallel 

 or perpendicular to the line of centres. When the direction 

 of motion is inclined 45° to the line of centres, the couple 

 has its maximum value, namely, 



3 mm'Wp 9 Tr mm' „ /OQN 



— . r-^v 2 or K— k-v 2 . . . . (33) 



477 r 6 r 6 



23. If we consider a bar made of given material and of 

 definite proportions, and so suspended as to have a given 

 period of oscillation about an axis through its centre of figure, 

 then the angular deviation of the bar arising from a given 

 velocity of translation through the aether will, at its maximum, 

 be inversely proportional to the square of the linear dimen- 

 sions of the bar. Though no experiments may have been 

 specially made with the object of detecting motional forces 

 of this kind, it seems certain that if more than a very slight 

 residual effect existed, it must have become apparent in other 

 observations. None the less it would be interesting to make 

 some definite experiments to put the question to the test, the 

 apparatus required being simple. Incidentally, if the couples 

 proved to be of measurable magnitude, we could readily 

 obtain the data necessary for a determination of our motion 

 relatively to the aether. This would involve no contradiction 

 of the principle of relativity in electromagnetism ; the phe- 

 nomena with which we are here concerned not being electro- 

 magnetic in character. 



24. Though we do not know what may be the value of the 

 constant K defined by (32), or even whether it be different 

 from zero, we can make some attempt to assign an upper 

 limit to its value. Imagine two compact masses of 1 gram 

 each, connected by a bar of negligible mass, the distance 

 between their centres being 2 centimetres. Let this system 

 be suspended by a quartz fibre, so that its axis is horizontal, 

 and the period of its oscillations 30 seconds. Let us further 

 suppose that we are able to detect any deviation as great as 

 one second of arc on either side of a mean position, when 



