624 SCIENCE PROGRESS 



It is one of the consequences of Einstein's theory that since 

 electromagnetic energy possesses inertia, it must be subject 

 to and must exert gravitational attraction in a similar manner 

 to the inertia possessed by matter. Since a ray of light 

 possesses electromagnetic inertia, it follows that when a ray of 

 light passes through a gravitational field, its path will in general 

 be curved owing to the gravitational action of the field. The 

 curvature produced is extremely small ; in fact, if we consider 

 a ray of light coming from an infinite distance and passing at 

 a distance r from a body of mass M, the ultimate deflection 

 produced in the direction of the path of the ray is wMftfr 

 where 7 is the constant of gravitation, 67 x io~ 8 in C.G.S. 

 units and c is the velocity of light, 3X io 10 cms. Thus M 

 must be enormously great in order that a measurable deflection 

 may be obtained. The only means of testing this result of the 

 theory is in the case of a star seen close to the limb of the Sun, 

 when the apparent displacement due to the sun's gravitational 

 action should be 175 seconds of arc. This is a very small 

 quantity when the difficulty of observing a star close to the 

 limb of the sun is considered, and the deflection produced 

 decreases proportionately to the increase in the distance of the 

 star from the sun. A. F. and F. A. Lindemann (M.N., R.A.S., 

 vol. lxxvii. p. 140) have discussed the possibility of testing the 

 hypothesis by photography of stars in daytime. They find 

 that by using red colour filters and plates specially sensitive 

 to the red it is possible to photograph bright stars at some 

 distance from the sun in broad daylight ; further experiments 

 carried out on similar lines at an observatory enjoying a fine 

 climate and a clear atmosphere might result in stars being 

 photographed sufficiently near to the sun for the theory to be 

 tested. The only other and the better alternative is to use 

 the favourable opportunity provided by a total solar eclipse ; 

 unfortunately total eclipses are somewhat rare phenomena 

 and bright stars are not always sufficiently near the sun at 

 the time of eclipse for the test to be made. A unique oppor- 

 tunity will occur, however, in the eclipse of 19 19, May 29, 

 visible in Peru, Brazil, and Central Africa, when the sun will 

 be in the crowded region of the Hyades and many fairly bright 

 stars will be available for measurement. With a telescope of 

 sufficient power the relative displacement of about three 

 seconds between two stars one on either side of the sun will 



