582 



SCIENCE 



[N. S. Vol. LI. No. 1328 



Scale 



Observed DeFLEcrtarc 



^^Hl 



~"&iEii) 



,-/. «T/ujai l««'llf, 



•,^5e«,g 



Fig. 1. Dr. Crommelin 's observed light deflections at Sobral, Brazil, plotted for each star according 

 to direction and a relative scale of magnitude. 



(Full line is observed vector; broken line is the Einstein vector. It will be observed that, in general 

 the observed vector departs from the Einstein vector in a direction away from a diameter of the sun 

 passing through the zenith for Sobral as projected on the photographic plate; ajbout this diameter, fur- 

 thermore, the angular departures, or non-radical effects, are found t o be symmetrical. ) 



expeditions of the Department of Terrestrial 

 Magnetism during the solar eclipse of May 29, 

 1919, at Sobral (D. M. Wise, in charge) and 

 at Cape Palmas, Liberia (L. A. Bauer, in 

 charge) upon the complete interpretation of 

 results of the astronomical observations. We 

 also received from Dr. H. Morize, director of 

 the Rio de Janeiro Observatory, meteorolog- 

 ical data pertaining to his eclipse station, 

 which was likewise Sobral, and desire to ac- 

 knowledge our indebtedness to him. It may 

 be recalled that the rays of light whose 

 deflections were measured during the solar 

 eclipse were subject chiefly: a to a gravita- 



tional action from the sun, h to optical re- 

 fraction in the sun's atmosphere, and c to 

 optical refraction in the earth's atmosphere. 

 The bearing of the geophysical observations 

 will be chiefly in relation to c. 



3. Let flu be the gravitational deflection of 

 a light ray grazing the sun's limb, ar, the 

 gravitational deflection of the ray at the dis- 

 tance p from the center of the sun expressed 

 in units of the sun's radius; then, according 

 to the Einstein law of gravitation, we have 



ao 1-74 



