148 



NA TURE 



[July 28, 1923 



There seems to be evidence ' of sudden as well as 

 slow and continuous changes, which, if they are real, 

 may be due either to shifting of parts of the crust of 

 the earth relatively to each other, or to a slow sliding 

 of the whole of the crust over the core. 



All these problems, which evidently are of the 

 greatest importance not for astronomy alone, depend 

 for their solution on very small quantities which even 

 now only begin to come within the reach of our most 

 accurate measures and most refined discussions. 



The third set of problems of fundamental astron- 

 omy concern questions relating to the positions and 

 motions of the fixed stars. Bessel's great work called 

 " Fundamenta astronomiae " consists of a careful dis- 

 cussion and synthesis of the observations made by 

 James Bradley as Astronomer Royal at Greenwich 

 upon the positions of the stars. 



Indeed, the positions and motions of the " fixed " 

 stars are the basis on which the whole structure of 

 astronomy rests. The manner in which these posi- 

 tions are determined is forced upon us by our location 

 on the moving earth. The accumulated labours of 

 astronomers since the commencement of accurate 

 observing by Bradley have resulted in a system (or 

 rather three systems, differing by small, but not 

 negligible, quantities) of positions and motions of the 

 stars. These are referred to a frame of reference, 

 which is defined by the motion of the earth, and 

 consisting of the equator, and a zero point on it. 

 Both the equator and the zero point are moving. It 

 need scarcely be stated that the formation "of such a 

 system of positions and motions of stars is a most 

 intricate and difficult problem, and we must confess 

 that it has not, so far, been solved in a manner which 

 satisfies the demands of statistical astronomy and 

 cosmogony. 



The system which is generally considered the best 

 of those now in use, that of Boss, is by no means 

 perfect : large errors in it are not at all improbable.* 

 These errors are errors of the system, not of the 

 individual star-positions, and the question naturally 

 arises : Is g,n " absolute " system at all necessary ? 

 Strictly absolute, of course, it is not : all systems of 

 reference are relative. By "absolute" we mean 

 relative to the inertial frame defined by the motion of 

 the earth in the solar system. But is it necessary to 

 base our system of star-positions on this motion of 

 the earth ? Would it not be much more natural, 

 and much more simple as well, to have relative posi- 

 tions and motions of the stars with regard to one 

 another, or to the general average of them? 



Many astronomers are inclined to answer this 

 question in the affirmative, and to consider the 

 absolute system more as a time-honoured institution 

 of our predecessors, a venerable relic from the pre- 

 photographic days, than as a useful and necessary 

 adjunct of modern stellar astronomy. In fact, by 

 the application of photography, we can easily derive 

 relative motions, or motions of individual stars rela- 

 tively to the " background," with an accuracy which 

 many times exceeds that attainable by fundamental 

 methods. 



By the blink-microscope w-e find, with compara- 

 tively very small labour, proper motions of ver^v^ 

 satisfactory accuracy referred to the background of 

 faint stars in the area examined. Of course this 

 " background " is a rather loosely defined frame of 

 reference, and we have no guarantee that the motions 

 of stars in different areas of the sky are really referred 

 to the same frame. A more elaborate method of j 

 referring the relative motions determined photo- 

 graphically to a quEisi-absolute system is proposed by 



' Lambert, U.S. Coast and Geodetic Survey, Serial No. 183, giving many 

 references to other papers. 



• See i.g. Kapteyn, B.A.N. 14. 



NO. 2804, VOL. 112] 



Kapteyn.* This method, however, depends on the 

 hypothesis that the sun's motion relative to faint 

 stars is the same as that relative to "bright stars. 

 This is why I call it a " quasi-absolute " system. 

 Are not the proper motions derived by these and 

 similar methods quite as valuable as those found by 

 fundamental methods ? 



My answer is decidedly in the negative. We cannot 

 do without the " absolute " system of fundamental 

 astronomy. The value of that system is not that it 

 is attached to the earth, but that we know exactlv 

 what the frame of reference is and that it is a rigorous 

 system, giving certainty that all motions are really 

 referred to the same frame. To see the importance 

 of this, I will put some questions, which cannot 

 be answered until we have a fundamental system 

 including the faint stars. 



Is star-streaming a universal phenomenon, or is it 

 local, and in the latter case, how far from the sun 

 does it extend ? Do the Orion-stars take part in the 

 star-streaming or not ? Is there a systematic motion 

 of faint stars relatively to bright stars ; or in other 

 words, is the average motion in space of the stars inde- 

 pendent of their brightness ? " Is there a rotation of 

 the system of stars as a whole ? 



These and similar questions are again examples of 

 great problems the solution of which depends on very 

 small quantities. These small quantities, the proper 

 motions of faint stars, cannot be profitably discussed 

 unless we have the certainty that these motions are 

 referred to a rigorous system. 



The necessity of a fundamental system being 

 granted, we must next ask : how are we to improve 

 and extend our present system ? Must we, in order 

 to establish an absolute system, necessarily retain 

 the old methods, or can we find other means ? Is 

 the meridian instrument to remain the only one by 

 which star places are to be determined ? To this 

 question I wish, as emphatically as to the former 

 one, to answer in the negative. We must look for 

 other methods, if it be only to verify the results 

 from the meridian work. 



Here I think is the greatest problem, and the most 

 urgent problem, of fundamental astronomy. It is 

 twofold : the determination of the positions of the 

 stars, and that of their motions. We must thus not 

 only establish a rigorous and faultless system of star 

 positions for the present day, but also strengthen as 

 much as possible our knowledge of the positions in 

 the past. These latter as now used depend prac- 

 tically exclusively on Bradley's observations. But 

 there are other data available, though not yet, or 

 not yet sufticiently, reduced. Among these the 

 most important are the rich mine of material still 

 lying unused in the observations made in the last 

 quarter of the eighteenth century and the first quarter 

 of the nineteenth by Hornsby and Robertson at 

 Oxford." I think the careful reduction of these 

 observations, which are of the same excellent quahty 

 as those of Bradley, is one of the most urgent demands 

 of fundamental astronomy. 



As to the means by which the modem positions 

 must be determined, I will not attempt now to enter 

 into details regarding the methods which have been, 

 or. may be, proposed to supplement the classical 

 meridian methods. All I wish is to convey an idea 

 of the meaning and the importance of the problems 

 of fundamental astronomy, and to show that far from 

 being uninteresting remains of a past period, their solu- 

 tion has become even more urgent by the newest de- 

 velopments of several branches of modern astronomy. 



• Groningen Publications, 28. 



" A considerable difference in average velocity would arise if the percentage 

 of high-velocity stars (cf. Oort, B.A.N. 23) were not the same for all 

 magnitudes. 



»» See Rambaut, Men. Not. R.A.S., 1\. 265. 



\ 



