368 ASTRONOMY 



the new facts, used as a test, differ in their special nature from those 

 relied on in the original construction of the theory. 



Thus the theory of gravitation not only accounted for known in- 

 equalities in the motion of the moon, but it was also the means of 

 pointing out the probable existence of other inequalities that were 

 subsequently verified by observation. Not only were the perturba- 

 tions of existing planets explained, but a hitherto unknown planet 

 was detected through its perturbations exerted on a known planet, 

 and was subsequently found to be actually visible in the sky. The 

 polar compression of the earth was predicted as a consequence of 

 gravitation before measurement was employed to verify the fact. 

 Not only was the precession of the equinox shown to be a consequence 

 of this theory, but the analogous nutation of the earth's axis was dis- 

 covered and formulated from that theory in advance of the ability of 

 observers to detect the minute apparent motions of the stars which 

 are traceable to this cause. This list might be indefinitely extended, 

 but it seems already sufficient to point out the most valuable element 

 in verification of natural law. 



The astronomy of the stars is now in much the same relative 

 situation as that occupied by planetary astronomy two thousand 

 years ago. The problems that confront it must be worked out by the 

 historic method of temporarily assuming simple geometrical con- 

 ceptions that may be suspected to underlie and connect the diverse 

 facts of observation. The Greek school of astronomy was preceded 

 by a long period in which classification of observed phenomena 

 marked the limit of attainment. The circles of reference for the sphere 

 were invented and the facts of diurnal rotation noted ; the path of the 

 zodiac was marked out; the recurrence of eclipses was studied; and 

 the length of the year was approximately determined. Thus a large 

 stock of conventional ideas was accumulated; and these proved use- 

 ful to the more exact and ingoing research upon which the Greek 

 school of astronomers entered. 



A similar accumulation of classified facts and conventional ideas 

 is going on now in the interest of stellar astronomy. First, and most 

 important of all, we are in possession of the results of a very large 

 expenditure of skill and energy in accumulating observed positions 

 of stars at various epochs from 1755 to the present time. As the time 

 seems to be drawing near when parts of the stellar problem may be 

 accessible to actual research with a good hope of results, both the 

 skill and energy devoted to observations of stellar positions, from 

 which the facts of stellar motion can be derived, is on the increase. The 

 effects of certain stellar motions which are merely apparent have been 

 formulated, and the greater part of these effects can now be disen- 

 tangled from the observations. Statistical researches concerning the 

 distribution of stars suggest valuable hypotheses to be subsequently 



