370 SCIENCE PROGRESS 



helion of Mercury ; this was a distinct initial success which at 

 once directed attention to the theory. He further pointed 

 out two other results predicted by the theory, which should 

 be capable of verification or disproof. The first of these was 

 a displacement to the red of the Fraunhofer lines of the solar 

 spectrum of about o-oo8 angstroms. The careful measure- 

 ments of Dr. St. John at Mount Wilson do not appear to con- 

 firm this prediction, although the smallness of the quantity 

 in question and the possibility of its being concealed by other 

 disturbing factors may cause doubt in some minds as to the 

 validity of Dr. St. John's conclusion. The second prediction 

 of the theory is that the path of a ray of light should be slightly 

 deflected in passing through a gravitational field. The only 

 means of testing this conclusion is to ascertain whether there 

 is any displacement in the apparent position of a star seen 

 near the edge of the sun. Naturally this can only be attempted 

 during a solar eclipse. The displacement predicted should 

 vary inversely as the distance from the sun's centre and 

 amount to i " - 7 5 at the sun's edge. In this connection, it 

 must be pointed out that if light is assumed to possess inertia 

 — -which is a reasonable supposition, since a beam of light con- 

 tains energy and can exert a pressure — then it necessarily 

 follows that it must be deflected in any field of force. On 

 Maxwell's theory, therefore, we should expect a deflection 

 which calculation shows would amount to o"-87 at the edge 

 of the sun, or to exactly half the amount predicted by Einstein. 

 If, on the other hand, light does not possess inertia, there 

 should be no deflection. There were thus three possible 

 alternatives — no deflection, or deflections of o"-Sy, or 1 "-75 

 respectively, any one of which might not unreasonably be 

 expected to be obtained when the test should be made. Of 

 these the nil deflection would probably be most unexpected. 



As already pointed out in these notes, the eclipse of 

 May 28-29 last provided a very favourable opportunity for 

 testing these alternatives, owing to the presence of several 

 bright stars in the neighbourhood of the sun at the time of 

 totality. As soon as it appeared probable, after the Armis- 

 tice, that it would be possible to send expeditions to observe 

 the eclipse, preparations were commenced for two expeditions 

 to set out under the auspices of the Joint Permanent Eclipse 

 Committee of the Royal and Royal Astronomical Societies. 

 One expedition went to Sobral, in North Brazil, the observers 

 being Dr. A. C. D. Crommelin and Mr. C. Davidson ; the other 

 went to the Island of Principe, off the West Coast of Africa, 

 the observers being Prof. A. S. Eddington and Mr. Cottingham. 

 Both expeditions were fortunate in securing observations dur- 

 ing totality, although there was much cloud in the sky at 



