= 58 



NA TURE 



[J ULY II, I 90 I 



this is to be understood in a limited sense. The dis- 

 cussion of the catalogue is about to begin. Such dis- 

 cussion will include the e.Kamination and detection of 

 errors in the " Precision Catalogues," the search for, and 

 discovery of, stars with large and unsuspected proper 

 motions, and the formation of a catalogue of variable 

 stars for the southern hemisphere. Further, the course 

 of the work has disclosed the existence of a possible sys- 

 tematic difference of colour in stars, depending on the 

 galactic latitude, and intimately connected with this in- 

 quiry is the investigation of the systematic corrections 

 which should be applied to the magnitudes derived from 

 the Cape plates to ensure one uniform system, photo- 

 graphically considered, or to connect the photographic 

 and optical magnitudes. A revision conducted on such 

 ample lines is a task of only less magnitude than that of 

 the construction of the catalogue itself, while the im- 

 portance and interest are even greater. That the same 

 competent hands will carry such a discussion to a final 

 issue will be the liope and the expectation of all 

 astronomers. 



The introduction supplied by Prof. Kapteyn to the 

 third volume cannot possess the novelty and the interest 

 which naturally attaches to that accompanying the first 

 volume of the annals. In that it was necessary to detail 

 his scheme of measuring the plates and effecting the 

 reduction ; he also sketched the results at which he had 

 arri\ed by comparison with the work made on the 

 meridian by other astronomers, pursuing similar but less 

 extensive methods. The subsequent volumes have had 

 to record the mechanical accumulation of the places and 

 magnitudes of stars observed by the same method, till 

 we have piled up for us the enormous total of 454,384 

 stars, catalogued within an area of 13,911 square degrees, 

 •embraced in the district between the South Pole and the 

 parallel of 19' south declination. The greater richness 

 of the southern skies is shown by the fact that this 

 number is only slightly less than that contained in the 

 joint Durchmusterung of Argelander and Schonfeld, 

 extending from the North Pole to zf south declina- 

 tion. Further, this richness has increased as the ob- 

 servations have been carried polewards, the maximum 

 being reached in the zone 48^-58" S. decl. In the first 

 section, comprising the zone iS"'-37" S. decl., the average 

 number of stars to a square degree was 2 5 '43 

 (Nature, loc. cit.) ; for the whole the average number 

 is 32'66, or more than double that of the Northern Bonn 

 Durchmusterung. But in a discussion involving relative 

 density of aggregation the vexed question of the maui- 

 tenance of a uniform standard of magnitude throughout 

 the whole sur\ey enters with perplexing uncertainty. C>n 

 this point Prof Kapteyn entertains views into which we 

 do not care to enter too minutely, for we are yet await- 

 ing his complete answer to sundry criticisms which have 

 been advanced. We are certainly inclined to follow him 

 in his assertion that if two or three tenths of a magnitude 

 be deducted from the estimates we shall obtain the 

 limit of photographic magnitude to which the stars of 

 the catalogue are certainly practically complete, and 

 that consequently we may assume the whole catalogue 

 to embrace all stars down to 9'2 mag. 



There is, however, the other and more thorny question, 

 which touches on the relative chemical activity of stars 

 in different parts of the sky, which it is not so easy to 

 answer. Prof. Kapteyn puts the question thus. To what 

 limit of magnitude would the plates be found complete 

 were the magnitudes of the Cape Durchmusterung re- 

 duced to a homogeneous set of photographic magnitudes 

 for the whole sky 'i The answer which he offers is that 

 the Durchmusterung will be found practically complete 

 in or near the Milky Way, to stars which in the scales of 

 Schonfeld, of Gould and of Thome are of the magnitude 

 9'5, and for the rest of the sky to stars actinically equiva- 

 lent to these. The vagueness of this reply is due to 



NO. 1654, VOL. 64] 



the peculiar feature that Prof. Kapteyn's discussions have 

 disclosed, and whose complete explanation is not yet 

 forthcoming. The measurement of a great number of 

 plates has satisfactorily shown that the law expressing 

 the growth of star density depending on proximity to 

 the Milky Way differs essentially from that exhibited in 

 the optical observations of Schonfeld and Gould. For 

 this fact two explanations, equally plausible, can be 

 offered. Prof. Kapteyn's contention is that there is a real 

 difference in the colours of the stars as the Milky Way 

 is approached, and that the increase of blueness leads to 

 increasing discrepancies in the differences between photo- 

 graphic and \isual magnitude, amounting roughly to coi 

 mag. for each degree of galactic latitude. On the 

 other hand, the ground for the observed inequality may 

 be due to systematic errors in assigning the optical 

 magnitudes to stars under the different condition in the 

 method of selection of the stars, when they pass slowly in 

 the sparser regions of the sky, and when the richer 

 regions are being observed. The tendency might very 

 well be, from the greater time at the disposal of the 

 observer in the first case, to observe fainter stars than 

 when he finds his field crowded with passing objects. 

 Of course, both views were fully admitted by Prof. Kap- 

 teyn, but he considered he had sufficient evidence to 

 establish his case, and though he acknowledges the force 

 of the arguments which have been brought against him, 

 he is still inclined to maintain his view. The fuller dis- 

 cussion is one of those points which have been left for 

 future investigation so as not to delay the completion of 

 the main work, but if the Groningen astronomer can 

 obtain support for his theory it may have an important 

 bearing on our views concerning the cosmical arrange- 

 ment of the stars. 



The accuracy of the stars' coordinates remains practi- 

 cally the same as in the earlier volumes, and from a 

 comparison with Gould's meridian places may be given 

 as follows ; — 



Prob. Error in Prob. Error in 



Declination. Right Ascension. Declination. 



- 38 to - 58'" + 0='2SS ± o'-0444 



- 58° to - 86' ± (o»-i57 -t- 0^-0764 sec S) + o''0559 



But the method of measurement of stars on the polar 

 plate affords a better means of determining the degree 

 of accuracy attainable on these plates, and of the possible 

 service they are likely to render in settling questions 

 of identity or of proper motion. On this plate the 

 rectangular coordinates have been measured with the 

 Repsold apparatus acquired by the astronomical labora- 

 tory at Groningen, and the measures reduced by com- 

 parison with all the stars (save one) common to the plate 

 and to Gould. Making due allowance for proper motion 

 and error of observation in the Cordoba places, the 

 probable errors of the Cape positions are found to be in 

 R..A.. ± o"'53 (arc of great circle) ; in decl. ± o"'76. 



Considering the shortness of the focal length (54 inches) 

 such a result is extremely gratifying, and it seems likely, 

 as anticipated by the authors, that many questions con- 

 nected with the proper motion of the southern stars 

 can be at once set at rest by an appeal to the original 

 plates. Some such work seems to have been already begun, 

 judging by the tables added to the volume, in which are 

 shown instances of stars missing on the plates and yet 

 given elsewhere, and of stars found on the plates not 

 recorded in known catalogues. Such work requires 

 infinite patience and care, and we can only once more 

 congratulate the ioint authors on the success that has 

 attended their unwearied efforts to secure uniformity and 

 accuracy. 



The second work under notice also e.xhibits the results 

 of continued labour pursued with persistence and success. 

 Photometry has so long been a feature in the researches 

 at Harvard College Observatory, and so many successive 

 volumes have detailed the method of observing, that on 



