38 CARNEGIE INSTITUTION 



possible to obtain in this manner an absolutely independent check 

 upon the direction of solar motion which, in a problem so impor- 

 tant, will possess the highest philosophical value and will become a 

 valuable test of fundamental hypotheses as to structure and motion 

 in the sidereal S5^stem. 



Whether the stars are distributed with approximate uniformity in 

 volume, and whether the motions are at random in every conceivable 

 direction, are questions which require for a definite decision the 

 added information which can be obtained by the measurement of 

 radial velocities of large numbers of stars extending, if possible, 

 somewhat below those visible to the unassisted eye, in combination 

 with discussions founded on proper motions which result from me- 

 ridian observations. What we now most need is such measurements 

 for stars not visible from the observatories of the northern hemi- 

 sphere. Scarcely anything in this line has been accomplished for the 

 southern hemisphere. Recently the Lick Observatory has dispatched 

 an expedition to Chile with the Mills three foot reflector to make 

 such measurements, and the maintenance of this expedition has 

 been provided for by Mr. D. O. Mills for three years. It may be 

 expected to produce results of the highest importance — of impor- 

 tance relatively several times as great as would attach to like efforts 

 with the same instrument in the northern hemisphere. 



Thus we may hope to have at our disposal within a few years the 

 measured radial velocities of practically all the stars in the whole 

 heavens that are brighter than the fifth magnitude. This will be 

 an extremely valuable result ; but it would be made far more valu- 

 able if such measurements could be secured for a greatly increased 

 number of stars over a greater range of magnitude. 



Are the velocities of the more distant stars the same on an aver- 

 age as those of the nearer stars ? Are the peculiar motions of the 

 stars, after abstraction of parallactic motions, the same in the direc- 

 tions to and from the observer at different distances from the Milky 

 Way ? In order to answer these and similar questions with sufficient 

 weight of evidence, the objects for which velocities in the line of 

 sight have been measured should number 2,000, if possible ; and in 

 order to accelerate the rate at which resvilts can be reached, we need 

 telescopes of the largest possible light-grasp. There is no apparent 

 obstacle, except cost, in the way of employing a telescope with a 

 five foot mirror for this investigation. 



That an increase in the resources for the measurement of radial 

 velocities of stars in the far southern skies is desirable, appears from 



