Mat 10, 1907] 



SCIENCE 



751 



TABLE V 



Results arranged according to Stellar Spectra 



Table I. shows that there is, as might be 

 expected, a distinct relation between parallax 

 and proper motion. Not only are there 

 striking individual exceptions to this law, 

 however, but the group having a mean proper 

 motion of 0".77, with a mean parallax of 

 + 0".039, destroys the continuity of the 

 series. 



In Table II. may be traced some relation 

 between magnitude and parallax. This com- 

 parison would have great interest, had the 

 selection of the stars been differently made. 

 As it is, only the first group of ten stars were 

 chosen with reference to their brightness, 

 while all the rest were selected because of 

 large proper motion, that is, in a general way, 

 because of their nearness. Only a hint can 

 therefore be obtained as to the real relation 

 between the magnitudes and parallaxes' of the 

 stars as a whole. The table shows that in 

 general bright stars are nearer than faint 

 ones, though even this obvious truth is ap- 

 parently refuted by the last two groups, 

 which make stars of mean magnitude 8.3 

 much nearer than those of magnitude Y.6. It 

 must not be inferred that the actual selection 

 was unwise. Any other selection than that 

 employed would probably have led for the 

 most part to negative results. The authors 

 made the best of an extremely difficult prob- 

 lem, perhaps the most difficult in the whole 

 realm of observational astronomy. 



Of Table III. the authors say : " This table 

 may also serve to indicate the number of 

 spurious parallaxes belonging to the work. 

 If, according to Newcomb, we regard all the 

 negative results as due to errors of observa- 



tion, and likewise an equal number of posi- 

 tive values to balance these, it would seem 

 that all seventeen of the group with parallaxes 

 between + 0".14 and +0".20 are real, 38 of 

 those from +0".07 to +0".13 and 35 of 

 those under +0"-06. Thus there are 90 stars 

 of the entire list of 163, for which there is 

 considerable presumption that the parallax 

 values found are actual." 



Tables IV. and V. appear to lead to results 

 of small importance so far as distribution is 

 concerned. 



Finally, a summary is given for the differ- 

 ent groups, except for Table III., of the aver- 

 age total stellar velocity relative to the sun, 

 and of the luminosity relative to the sun. 

 In this summary the greater luminosity of 

 the brighter stars is strikingly shown. 



Too high praise can hardly be given to these 

 parallax investigations, carried on during so 

 many years. Yet the results, though of great 

 value in themselves, do not encourage the 

 hope that by similar heliometer observations 

 we shall ever gain a knowledge of the dis- 

 tances of any large number of stars, especially 

 of those most distant. It is doubtful if we 

 have, at the present time, any mode of re- 

 search sufficiently refined to determine the 

 parallaxes of the most distant members of 

 our sidereal system. The quantities involved 

 are too small. By more powerful instruments, 

 especially by photographic telescopes of great 

 focal length, it may be possible to determine 

 smaller values than those yet found. The 

 relation between proper motion and parallax 

 offers a hopeful means for the determination 

 of mean values, but this method has limita- 

 tions. The relation between magnitude and 

 distance is as yet uncertain. Indeed, the solu- 

 tion of the most difficult parts of the problem 

 calls for some new means of research far 

 more powerful than any known at th« present 

 time. S. I. Bailey 



Haevaed College Observatoet 



TEE 



THE 



AMERICAN ASSOCIATION FOB 



ADVANCEMENT OF SCIENCE 



THE PLATTSBUEG MEETING OF THE SECTION OF 



GEOLOGY AND GEOGRAPHY 



Section E will hold a summer field meeting, 

 July 3-10, 1907, in New York State in the 



