THEORY OF RELATIVITY 
285 
purpose a wide-angle lens (of short focal length) was chosen 
giving a field of great range, recording stars at from 5° to 10° 
away from the Sun, by means which that law was also confirmed, 
the residuals, observation minus theory, coming out extremely 
small. 
Third : We have pointed out in connection with a field of cen- 
trifugal force that a clock placed in such a field loses time. The 
same would be true if a clock were placed in a gravitational field. 
Now an atom is a periodic mechanism comparable to a clock. 
Its period would be greater in the gravitational field of the Sun 
than in that of the earth. As a consequence of this the spectral 
lines of a substance in the vapor envelope of the Sun would be 
shifted to the red as compared with the lines of the same sub- 
stance produced on the earth. The relative displacement is about 
the same as that produced by a radial velocity of recession of 
0.63 kilometers per second, being of the order of 0.008 Angstrom 
o 
units measured in the region of A=3883 A lying almost within 
the errors of observations. Nevertheless the results of Evershed 
in 1918 and those of Grebe and Bachem in 1919 have offered con- 
siderable evidence in corroboration of the prediction; while at 
that time the findings of Prof. St. John of the Mt. Wilson solar 
laboratories were not confirmatory. In September 1923, how- 
ever, in a paper read before the Astronomical Society of the 
Pacific he expressed himself as having come to the conclusion 
that he had confirmed the Einstein effect relating to the above 
shift of spectral lines ; and that in the future a correction would 
have to be applied to spectroscopic measures. 
At the last meeting of the National Academy of Science, 
(Science, May 16, ’24) Prof. H. D. Curtis of Allegheny observa- 
tory reported recent measures of Drs. Burns and Meggers: 
“showing shift of lines very different from the simple and uni- 
form amounts predicted by the relativity theory. Instead of all 
the lines being shifted by an equal amount to the red and that 
amount predicted by the Einstein theory, a very marked line- 
intensity factor is found. That is, for very faint solar lines there 
is little if any shift, and the amount of this shift increases as 
the wider and stronger lines are used.” Then follow numerical 
results showing the above; and finally, after some further re- 
marks along the same line, the following conclusion: “Accord- 
ingly the authors regard these results as a negation of one of 
the so-called proofs of the theory of relativity.” 
