I04 



NA TURE 



[July 21, 1923 



theory of aberration not only holds for the point P,. 

 but also for the point where the ray reaches the surface 

 of the earth. 



Stokes's theory cannot, however, be maintained, 

 because the two assumptions that there is a velocity 

 potential and that all over the surface the ether has 

 the velocity of the earth contradict each other, 



4. Ether Whirls. — It can be imagined that a 

 rotating planet is surrounded by a whirl in the ether. 

 If definite assumptions are made concerning the 

 distribution of velocity in this whirl, one can determine 

 the deviation of a ray passing through it and the 

 amount of diurnal aberration thus produced. 



There is another phenomenon still by which one 

 could detect an ether whirl. If the planet Jupiter 



were surrounded 

 by a whirl, there 

 would be a devia- 

 tion that could be 

 observed in the 

 case of the occul- 

 tation of a star, 

 and that is found 

 to be quite ap- 

 preciable if plaus- 

 ible assumptions 

 concerning the ex- 

 tension of the whirl 

 be made. Nothing 

 of the kind has 

 ever been observed. This speaks in favour of the 

 hypothesis that the ether is not set in motion by the 

 planet. 



5. Fresnel's Theory of Aberration. — Fresnel 

 assumed that the earth is absolutely permeable to the 

 ether, so that it can pass through' the ether without 

 in the least setting it in motion. So far as the subject 

 of this lecture is concerned, this assumption leads to 

 the same results as the theory of relativity. The 

 annual aberration is now immediately explained by 

 what was said in § 3. As to the daily aberration, it 

 is determined in the same way as the annual aberration 

 by the motion of the observer. If an astronomer 

 placed at a point on the equator observes a star 

 situated in the plane of that circle, at an altitude h, 

 the aberration amounts to 0-32" x sin /;.. It would be 

 possible to observe it if the distance of two stars far 

 apart could be measured to within o'l". If, for example, 

 two stars A and B in the plane of the equator follow 

 each other in their daily motion at a distance of 60°, 

 their distance will be diminished by o-i6" when A has 

 reached the zenith, and increased by the same amount 

 when B has reached that point. In reality, however. 



Fig. 



the existence of the diurnal aberration has not been 

 proved, though astronomers correct their observations 

 for it. 



6. Michelson's Interference Experiment. — Prof. 

 Michelson has devi.sed an experiment in which t^o \ 

 rays of light propagated in opposite directions along the ' 

 sides of a great triangle ABC (Fig. 3) in a horizontal 

 I)lane are made to interfere ; Mj, Mj are mirrors, 



P a dividing plate of glass ; the course of one ray is 

 LACBAQ, and of the other LABCAQ. 



In general, let L and Q be any two points having 

 a fixed position in the figure, which is attached to the 

 earth. The ether (supposed not to share the earth's 

 rotation) has a motion through the diagram, consisting 

 in a rotation about the axis of the earth. From what 

 has been said in § 2, one can deduce the time that is 

 required for the passage from L to Q. Let s be the 

 course of the ray if there were no rotation, s' the actual 

 course. One has to calculate the value of (2) for s , 

 but if one neglects terms of the second order, one 

 can substitute for it the value for the path s, because 

 the integral is a minimum for s' . The influence of 

 the earth's rotation is given by the integral of the last 

 term in (3), and is found in the case of Fig. 3 to be 

 proportional to the area of the triangle ABC, to the 

 angular velocity of the earth, and to the sine of the 

 geographical latitude. The effect would be of equal 

 magnitude but of opposite sign for the two interfering 

 rays, and so the position of the interference fringes 

 will be slightly changed by the rotation of the earth. 

 It may be hoped that it will be possible to observe 

 the effect by a suitable method of observation. 



7. An Imaginary Experiment. — Suppose two 

 parallel metallic wires (perfect conductors), such as 

 are used in Lecher's experiment, to be placed round 

 the equator of the earth, each forming a closed circle. 

 Let standing electromagnetic waves be produced 

 between these wires. One may confidently expect 

 that the loops and nodes will travel around the earth 

 from east to west in 24 hours, and this can be con- 

 sidered to be a proof of the earth's rotation. 



If the statement that the earth rotates is to have 

 any meaning one must assign some system relatively 

 to which the rotation takes place. If the imaginar}- 

 experiment were performed with the result just 

 mentioned, one could say that the earth rotates 

 (i) relatively to the loops and nodes, (2) relatively 

 to a system of co-ordinates in w hich light is propagated 

 in straight lines with the speed c, (3) with respect to 

 the stationary ether in which the loops and nodes 

 have their seats, or (4) with respect to the fixed stars 

 by the influence of which the position of the loops 

 and nodes is determined. (Mach, Einstein.) 



A Large Refractor for Johannesburg. 



By Frank 



'T*WENTY years ago, and soon after the close of 

 -*- the Boer War, the South African Association 

 for the Advancement of Science petitioned the Trans- 

 vaal Government for the establishment of an obser\a- 

 tory for the sciences of meteorology and astronomy. 

 The reply was immediately favourable as regards 

 meteorology, but it was not found possible to organise 



NO. 2803, VOL. I 12] 



ROBBINS. 



an astronomical department for some half-dozen years 

 or so. In the meantime, by loan or by gift, a few 

 instruments were obtained, and quite soon attention 

 was forcibly directed to the very exceptional climate 

 and sky of Johannesburg by means of work actually 

 done there by the aid of a modest 9-inch refractor. 

 In consequence, early in 1909 the Minister for Lands 



