208 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1921. 



The results, however, are not difficult to understand. The principal 

 ones are these: 



1. A ray of light passing near a gravitating body like the sun will 

 not travel in a straight line, but will be deflected slightly downward 

 toward the gravitating body, much as a very rapidly moving projec- 

 tile would be deviated. 



Calculation shows that the amount of deviation would be quite too 

 small to measure for a ray of light that has passed near the moon or 

 planets, but that for light that has passed near the sun the deviation 

 reaches nearly tAvo seconds of arc, which the modern astronomer, ac- 

 customed to accurate measurements, considers a large and very easily 

 measurable quantity. 



2. Newton's law of gravitation, on Einstein's principle, appears to 

 be only an approximation to the true law, but an exceedingly good 

 approximation — so much so that among all the intricate motions of 

 the planets there is but a single case in which the introduction of the 

 new law instead of Newton's principle produces perceptibly different 

 consequences. 



We all know the planets are moving in elliptical orbits about the 

 sun and that the line joining the sun to the nearest point of the orbit 

 has a certain definite position. 



On Newton's theory this line would remain permanently fixed in 

 S p ace — always in the same direction — if it were not for the fact that 

 the orbits of the planets are slightly but continually modified by their 

 mutual attraction. These influences, or so-called perturbations, can, 

 however, be accurately calculated and allowed for, so that they need 

 not worry us here. 



On the Einstein hypothesis this line to the nearest point in the 

 orbit, or the perihelion, should not remain fixed, but should move 

 slowly forward in the direction in which the planet is moving around 

 the sun. The rate of its motion can be calculated from the theory 

 when the distance and period of the planet are known. To this effect 

 are added the influences of the attraction of the other planets as 

 before. 



It has been known for some 30 or 40 years that the perihelion of 

 the planet Mercury, after allowance had been made for the per- 

 turbations due to the attraction of the other planets, was actually 

 moving slowly forward in a manner which was very difficult to ex- 

 plain. Attempts to account for it have failed. 



For example, the attraction of an unknown planet between Mer- 

 cury and the sun would do the trick, but observations made during 

 eclipses of the sun show that there is no planet there. Nor can there 

 be a great number of small bodies whose combined attraction would 

 do it, for these would reflect so much sunlight as to produce a bright 



