122 PROCEEDINGS OF THE AMERICAN ACADEMY 



of light) must be diminished by 3 per cent, to suit Foucauh's experi- 

 ment, then we must at the same time diminish the other term (the 

 velocity of the earth) proportionally ; and the old ratio will be pre- 

 served and the value of aberration will be left unchanged. Is it possi- 

 ble, therefore, that there can be an uncertainty to the extent of 3 per 

 cent in the velocity of the earth ? If so, the tables are turned ; and 

 instead of employing the ratio which aberration supplies to calculate 

 tlie velocity of light from the velocity of the earth, as the best known 

 of the two, we henceforth must calculate the velocity of the earth from 

 the velocity of light. For Foucault has found the latter by experi- 

 rtient more accurately than Astronomy gives the former. If there is an 

 error of 3 per cent in the velocity of the earth, it is an error in space, 

 and not in time. To diminish the velocity of the earth sufficiently by 

 a change of time would demand an increase in the length of the year 

 amounting to 11 days nearly. 



The only other way of reaching the velocity of the earth is by di- 

 minishing the circumference of the earth's orbit, and this, if diminished, 

 changes proportionally the mean radius of the orbit ; that is, the sun's 

 mean distance. The question, therefore, resolves itself into this. Can 

 the distance of the sun from the earth be considered uncertain to the 

 extent of 3 per cent of the whole distance. 



The answer to this question will lead me into a brief discussion of 

 the processes by which the sun's distance from the earth has been de- 

 termined, and the limits of accuracy which belong to the received 

 value. To see the distance of any body is an act of binocular vision. 

 When the body is near, the two eyes of the same individual converge 

 upon it. The interval between the eyes is the little base-line ; the 

 angle which the optic axes of the two eyes, when directed to the body, 

 make with each other, is the parallax ; and by this simple triangulation, 

 in which an instinctive geometrical sense supersedes the use of sines 

 and logarithms, the distance of an object is roughly calculated. As 

 the distance of the object increases, the base-line must be enlarged ; 

 but the geometrical method is the same, even when the object is a star, 

 and the base of the triangle the diameter of the earth's orbit. Substi- 

 tute, then, for the two eyes of the same observer the two telescopes of 

 different astronomers, planted at the opposite extremities of the earth's 

 diameter, and any one will understand the principle upon which the 

 binocular eye of science takes its stereoscopic view of the universe, 

 and plunges into the depths of space. In this way it is that the dis- 



