Apbii. 1, 1898.] 



KNOWLEDGE 



67 



name of " fixed " stars to the crude early notion of their 

 being rivetted — inri.vir — to the "palace-roof" they very 

 effectually serve to adorn. Yet in reality this starry vault 

 integrates, so to speak, immeasurable abysses of star-strewn 

 space. There, as elsewhere, "things are not what they 

 seem," and we are confronted with the inevitable question 

 as to the relation between the thing that is and the thing 

 that seems. From the aspect of the Milky Way came the 

 tirst suggestion of an answer. Tiie thought could not but 

 present itself to intelligent inquirers that the dimly-shining 

 girdle of the sphere might exercise a governing influence 

 over its contents. Varied investigations of this obvious 

 possibility were accordingly carried out, and the fact was 

 established of a general increase of stellar density, gaining 

 intensity with descent in magnitude, towards the galactic 

 plane. The combination into a system was thus rendered 

 probable of the sprinkled stars of the constellations and the 

 streaming stars of the nebulous zone running through 

 them ; but no hint could be gathered as to the nature of 

 the combination. Only the bare existence of some principle 

 of arrangement was perceived by unsatisfactory glimpses. 



Until lately the onlj' practically available means of 

 searching out the plan of stellar structure was by institutuig 

 comparisons between the numbers and the magnitudes of 

 the stars. Average brightness, it was reasonably supposed, 

 gave a measure of average relative distance, and from 

 distance relative abimdance per unit of space could be 

 inferred by simply counting the successive photometric 

 ranks. But the value of this method has been impaired 

 by an advance of knowledge in two directions. In the first 

 place it became evident, on a wider acquaintance with 

 stellar proper motions, that their assorted amounts afforded 

 a much surer test of distance than could be derived from 

 the consideration of magnitude alone. Next, it came to be 

 recognized that the intrinsic brilliancy of stars is very dif- 

 ferent for different spectral types, so that stars of the Vega 

 type must be nearly three times more remote than stars 

 like Capella or the sun, equally massive, and possessing the 

 same visual brilliancy. Diversityof stellar type implies then, 

 diversity ia laws of distribution — diversity, not only of the 

 photometrical, but also of the physical kind, and hence 

 constituting an essential feature of sidereal organization. 

 The preponderance of Sirian stars in the Milky Way, 

 simultaneously ascertained during the progress of Prof. 

 Pickering's spectographic survey and of Dr. Gill's photo- 

 graphic Durchmusterung, appears to be a combined result. 

 In part, it must depend, as pointed out by Mr. Monck, upon 

 the crowding in of Sirian stars lying far beyond the limiting 

 distance of the included solar stars ; but there is reason to 

 believe that it is also in part produced by a genuine 

 relationship. What is certain is that it fits in with much 

 that was already known, and provides a fresh platform for 

 further inquiry. 



It was in great measure through the labours of Prof. 

 •J. C. Kapteyn, of Grouingen, in measuring and reducing 

 the Cape Durehmuslerung plates, that the systematic dif- 

 ference in question was detected ; and their conclusion has 

 allowed him leisure to follow up the clue thus placed in his 

 hands. He has done so by a research into the distribution 

 of the stars, with special reference to their spectral types, 

 communicated to the Amsterdam Academy of Sciences in 

 two important papers, read AprU 29th, 1892, and January 

 28th, 1893, respectively. The first, indeed, gave only pre- 

 liminary inferences, but they were ratified and extended 

 m the second by the discussion of a greatly enlarged stock 

 of data. These related to 2357 stars, of which 1189 are 

 classed in the Draper Catalogue of Stellar Spectra as of 

 the first, 1100 as of the second, and 1)2 as of the third 

 type. The proper motions of 476 of them were taken from 



the list prepared by Herr Stumpe for his determination of 

 the solar translation ; those of the rest from the Bradley- 

 Auwers catalogue. The treatment of this material was on 

 the principle that stars are, on a fairly wide average, 

 distant from us in the inverse ratio of their apparent 

 movements. So far as concerns the perspective element 

 contained in them, this is of course strictly true, allowance 

 being made for differences of angular position with regard 

 to the apex of the sun's way. That is to say, the line 

 traversed in a given time by the sun, if seen square from 

 any star, would be of a length proportionate to that star's 

 remoteness ; and the transferred displacement of the star, 

 as viewed from the sun, would be equal and opposite to 

 the proper displacement of the sun as viewed from the 

 star. Thus, assuming the direction of the solar translation 

 known, and the '' peculiar " movements of the stars to be 

 so irregular as to give a zero efl'ect when numerously thrown 

 together, a secure measm-e is afforded of comparative 

 stellar distance. Prof. Kapteyn accordingly resolved his 

 proper motions into two components, one directed along a 

 great circle passing through an apex in R.A. 276 , Dec.+ 

 34°, the other at right angles to it. The first, reduced to 

 a position 90° from the apex, was treated as of wholly 

 perspective origin ; the second could not but be whoUy 

 original. And it was reassuring to find that their values 

 (abstraction being made of a few exceptionally swift objects) 

 varied in groups of stars arranged according to proper 

 motion, pretty nearly in the same proportion. Either com- 

 ponent, then, of stellar motion — the parallactic, or the 

 peculiar — appears to supply a valuable criterion of stellar 

 distance. 



The initial difficulty connected with space - sounding 

 operations being thus removed, a number of interesting 

 conclusions lay, comparatively speaking, close at hand. 

 Some of these had been anticipated by Mr. Monck, and 

 their independent deduction, through a far more elaborate 

 investigation, shows them to be deserving of no small credit. 

 To begin with, the remarkable circumstance seems fully 

 established that stars with weU-accentuated proper motions 

 show predominantly spectra of the solar description.* 

 Nor can we hesitate to agree that these objects owe their 

 mobUe character to their relative vicinity. They consti- 

 tute, accordingly, a group which surrounds and includes 

 the sun. It is most likely roughly spherical in shape, and 

 is so strongly condensed interiorly that a volume of space 

 near its centre contains 98 times as many stars as an equal 

 volume near its circumference. The maximum com- 

 pression appears to be round a pointf lying away fi-om 

 the sun, towards the north-western section of Andromeda, 

 supposed by Prof. Kapteyn to coincide with the centre of 

 the Milky Way. l!ut this identityis highly questionable. It 

 depends — since the centre found for the cluster has a southern 

 galactic latitude of about 20^ — upon the truth of Sir .John 

 Herschel's and Struve's respective inferences of a position 

 for the sun, eccentric as regards the round of the Galaxy, 

 and to the north of its medial plane. The brilliancy of tlie 

 Milky Way, however, in the southern hemisphere afl'ords 

 in reality not the slightest presumption of nearness J ; 

 and Gould was unable, from a much more complete inquiry 

 than Struve's could possibly be, to assign to the sun either 

 north or south galactic latitude. His situation, for any- 

 thing that can be proved to the contrary, may be perfectly 

 symmetrical within the great cosmical annulus. It cannot 

 then be admitted, at least on the present showing, that the 

 latter is concentric with the newly-constituted solar star- 



* See Ranvard, Knowledge, Vol. XIV., p. 50 ; " Old and Xew 

 Astronomy," p. 798. 



t The co-ordinates of which are R.A. Oh. Oin., Dec. + 42°. 

 X Suttou, Knowledge, Vol. XIV., page 42. 



