236 



SCIENCE. 



[Vol. II., No. 29. 



form a stable system in the sense in which we say 

 that the solar system is stable, — that the stars of this 

 system do not revolve around definite attractive cen- 

 tres ? Admitting that the solar system is moving 

 through space, can we at the present moment even de- 

 termine whether that motion is rectilinear, or curved, 

 to say nothing of the laws which govern that motion ? 

 How much of truth is there in the conjectures of 

 Wright, Kant, Lambert, and Mitchel, or even in the 

 more serious conclusions of Moedler, tliat the Alcyone 

 of the Pleiades is the central sun about which the 

 solar system revolves ? 



These are questions which, if solved at all, must be 

 solved by a critical study of observations of precis- 

 ion accumulated at widely separated epochs of time. 

 The first step in the solution has been taken in the 

 systematic survey of the northern heavens undertaken 

 by the gesellschaft, and in the survey of the south- 

 ern heavens at Cordoba by Dr. Gould. The year 

 1875 is the epoch about which are grouped the data 

 which, combined with similar data for an epoch not 

 earlier than 1950, will go far towards clearing up the 

 doubts which now rest upon the question of the di- 

 rection and the amount of the solar motion in space; 

 and it cannot be doubted that our knowledge of the 

 laws which connect the sidereal with the solar system 

 will be largely increased through this investigation. 

 The basis of this knowledge must be the observed 

 proper motions of a selected list of stars, so exactly 

 determined that the residual mean error shall not 

 affect the results derived ; or, failing in this, of groups 

 of stars symmetrically distributed over the visible 

 heavens, sufficient in number to affect an elimina- 

 tion of the accidental errors of observation, without 

 disturbing the equilibrium of the general system. 



For an investigation of this kind, a complete sys- 

 tem of zone observations, at widely separated inter- 

 vals, will afford the necessary data, if the following 

 conditions are fulfilled. 



■First: The proper motions must be derived by a 

 method which does not involve an exact knowledge 

 of the constants of precession. In every investiga- 

 tion with which I am acquainted, the derived prop- 

 er motions are functions of this element. 



Second : The general system of proper motions 

 derived must be free from systematic errors. Errors 

 of this class may be introduced either through the 

 periodic errors inherent in the system of fundamental 

 stars employed in the reduction of the zone observa- 

 tions, or in a change in the constants of precession. 

 It is in this respect that the utmost precaution will 

 be required. If from any cause errors of even small 

 magnitude are introduced into the general system of 

 proper motions at any point, the effect of these errors 

 upon the values of the co-ordinates at any future 

 epoch will be directly proportional to the interval 

 elapsed. We can, therefore, compute the exact 

 amount of the accumulated error for any given 

 time. 



When this test is applied to the fundamental stel- 

 lar systems independently determined by Auwers, 

 Safford, Boss, and Newcomb, we find the following 

 deviations inter se at the end of a century. 



It is the common impression, that both the direction 

 and the amount of the motion of the solar system in 

 space are now well established. The conclusions of 

 Struve upon this point are stated in such explicit 

 language that it is not surprising that this impres- 

 sion exists. He says, " The motion of the solar sys- 

 tem in space is directed to a point in the celestial 

 sphere situated on the right line which joins the two 

 stars measured from ir and u Herculis. The velocity 

 of this motion is such that the sun, with the whole 

 cortege of bodies depending on hira, advances annu- 

 ally in the direction indicated, through a space equal 

 to one hundred and fifty-four million miles." 



It must be admitted that there is a general agree- 

 ment in the assignment by different investigators of 

 the co-ordinates of the solar apex. This will be seen 

 from the following tabular values. 



Hci-acliel, 1783 

 Prevost . . 

 Klugel, 1789 . 

 Hcrschel, 1805 

 Argelander, 18 

 Lundahl . . 

 Struve . , 

 Galloway 

 Miidler . . 

 Airy . . . 



Dunkin . . 



257° 00' 



230 00 



260 00 

 245 52 

 257 49 

 252 24 



261 22 



260 01 



261 38 

 (256 54 

 1 261 29 



261 14 



263 44 



+25° 00' 



+25 DO 



+27 00 



+49 38 



+28 60 



+14 26 



+37 36 



+34 2S 



+39 54 



+34 29 



+26 44 



+32 55 



+25 00 



In estimating the value which should be attached 

 to these results, several considerations must be taken 

 into account. 



(a) All of the results except those of Galloway de- 

 pend practically upon the same authorities at one 

 epoch, viz., upon Brodley. 



(6) The deviations inter se probably result, in ai 

 large measure, from the systematic errors inherent in 

 one or both of the fundamental systems from which, 

 the proper motions were derived. For example, 

 Lundahl employed Pond as one of his authorities, 

 and it is in Pond's catalogue that the most decided 

 periodic errors exist. 



(c) Brot in 1812, Bessel in 1818, anid Airy in 1860, 

 reached the conclusion that the certainty of the move- 

 ment of the solar system towards a given point in the 

 heavens could not be affirmed. 



(d) The problem is indirect. In the case of a mem- 



