Apkil 1, 1892.] 



KNOWLEDGE 



73 



velocity on the part of some stai's as compared ■with others. 

 This appears not only from the accurate resiilts of Vogel, 

 but from the more extensive but less accurate researches 

 made at Greenwich and elsewhere. I doubt if there is a 

 star observed half a dozen times at Greenwich which gives 

 an average spectroscopic velocity of 80 miles per second, 

 and the average results obtained at Greenwich are con- 

 fessedly too large. Stars no doubt diiJer in velocity as 

 well as in other respects, but not to such an extent as to 

 deprive large proper motion of its value as a test of near- 

 ness. And an examination of any catalogue of stars, with 

 large proper motion, leads to the same conclusion, by 

 showing us that their real motions are comparable with 

 the sun's motion in space. Thus, taking 'M. Bossert's 

 Catalogue of Stars, with proper motions of half a second 

 or upwards annually, and taking the right ascension of 

 the pomt towards which the sun is mo'\'ing as 270°, the 

 effect of the sun's motion will be to diminish the right 

 ascension of all stars in the second and third quadi'auts, 

 and to increase the right ascension of all stars in the first 

 and fourth quadrants. As a matter of fact I find that 

 three-fourths of the stars in the second and third quadrants 

 are moving in the direction of diminishing right ascension. 

 and three-fourths of those in the first and fourth quadrants 

 are moving in the direction of increasing right ascension. 

 This is just the result which would be obtained if the sun's 

 motion in right ascension was equal to the average motion 

 of the stars in question. The effect in declination is also 

 manifest, though uot quite so well marked, and the result 

 appears to be that the stars in M. Bossert's Catalogue are 

 not on the average moving with much greater velocity 

 than the sun. 



But when we examine the spectra of these fast-moving 

 stars, one important fact immediately occurs to us. Over 

 the whole sky the numbers of stars with spectra of the 

 first type (Siriau stars) and of the second type (Solar stars) 

 are about equal ; but when we examine M. Bossert's 

 Catalogue of Stars with large proper motion, we find that 

 there are about eight Solar stars to one Siriau. (Stars 

 with other tj-pes of spectrum are almost wholly absent.) 

 The Sim thus appears to be one of a class or cluster of 

 stars whose spectra are of a type similar to its own. And 

 we begin to see, too, why the brightness or magnitude of 

 a star affords so doubtful an index to its distance from us. 

 Ceteris parilnin, Sirian stars appear to be brighter than 

 Solar stars, and a classification according to brightness or 

 magnitude vnh prove to a large extent misleading so long 

 as we omit the consideration of the spectra of the stars. 

 The relative efl'ect of the light of two stars on the eye is 

 often widely dift'erent fi-om the relative effect on the photo- 

 graphic plate, and a mere change in the structure of the 

 eye might revolutionize the whole aspect of the heavens. 

 We must, therefore, note the kind of light as well as the 

 quantity before we found any inference as to distance on it. 



Another test of distance occurs in the case of binary 

 stars, which is worth examining. If two stars, which are 

 separable by the telescope, complete their revolution in a 

 moderate period, the inference is that they are comparatively 

 near us ; otherwise the mass of the system must be enormous. 

 On the other hand, if the angular distance between the 

 stars is not large and yet the revolution is very slow, the 

 pair is probably at a great distance ; otherwise the mass 

 must be very small. I accordingly compared Mr. Gore's 

 Catalogue of Orbits of Binary Stars with the Draper 

 Catalogue of Stellar Spectra, and I foimd thit the former 

 catalogue contained 32 Solar stars to 11 Sirians. That 

 this preponderance was due to the comparative nearness of 

 the Solar stars rather than to their large masses, appeared 

 from the fact that many of them had large proper motions. 



and that in more than one instance a sensible parallax 

 (with the result that the mass was not very large) appeared 

 to be determined by actual measurement. Such stars are 

 a. Centauri, 61 Cygni, 40 Eridani, -q Cassiopeia, and 70 

 Ophiucbi. It might be thought, perhaps, that Sirian stars 

 were, for some reason, less likely to enter into binary 

 combinations than Solar stars. But this does not seem to 

 be the case. On comparing the list in the " Handbook of 

 Double Stars" with the Draper Catalogue, I found that the 

 Sirian stars were quite as numerous as the Solar stars — 

 perhaps a httle more so. But when we turn to the double 

 stars, whose orbits have been computed, the proportion 

 drops at once from one-half to one-fourth. 



The exact limits of this Solar cluster, and the number of 

 stars comprised in it, remain to be determined, but one 

 thing, I think, may be affirmed of it — that it has a 

 northerly drift. Thus, when Professors Boss and Stumpe 

 deduced the position of the apex of the sun's way from stars 

 with large proper motions, they obtained a less northerly 

 position than when they determined it from stars with 

 smaller proper motions.'^-' It might be supposed that this 

 was due to the greater actual velocity of the stars with 

 large proper motions which had the effect of disguising the 

 sun's motion in space, but a comparison of M. Bossert's 

 Catalogue with the more extensive catalogue of Mr. Main 

 does not lead to that conclusion. The ett'ect of the sun's 

 motion in R. A. is more marked in the case of the stars 

 with large proper motions, though that in N. P. D. is less 

 so. The stars with large proper motions are not perhaps 

 faster movers than their compeers, but they have a general 

 drift, which others have not ; though whether the more 

 distant stars are moving indifi'erently in all directions, or 

 have a general drift dift'ering from that of the solar cluster, 

 remains to be determined. 



I do not think, however, that all the Solar stars, or even 

 all the bright Solar stars, belong to the cluster or system 

 of which I am speaking. Betelgeuset and Eigel, for 

 instance, have not only very small proper motions, but 

 judging from the attempts which have been made to 

 determine their parallaxes are nearly as remote as the 

 fainter Sirian stars which abound in the region of Orion. 

 Procyon, Capella, Aldebaran, and Arcturus (in spite of Dr. 

 Elkin's insensible parallax of the latter, m which I have 

 not much faith), on the other hand, probably belong to the 

 Solar cluster. The motions of the stars in this system 

 may be illustrated by supposing that all the elliptic comets 

 which move round the sun were simultaneously visible. 

 They would appear to be moving almost indifl'erently in 

 every possible direction, yet, taken all together, there 

 would be a general drift in the direction of the sun's 

 motion. Apparent indift'erence of motion may in this way 

 be combined with a general northerly drift. 



Further light, however, may, I think, be' thrown upon 

 the structure of the Solar cluster by sub-dividing the Solar 

 stars into two classes, which I propose to designate 

 Capellan and Arcturian respectively, after their most 

 brilliant representatives. The former class includes stars 

 with the spectra denoted by E and F, but more especially 

 F, in the Draper Catalogue. They are, on the whole, 

 somewhat less numerous than the Arcturian stars, but of 

 the 32 Solar binaries for which orbits have been computed, 

 no less than 23 have spectra of the CapeUan type, the 

 spectrum in 20 cases being F. They actually outnumber 

 all other binaries put together. Taking Mr. Main's 

 Catalogue of Proper Motions and comparing it with the 

 Draper Catalogue, I found that, roughly speaking, only 24 



•See Old and New Astronomy, Part XII. 

 t But the spectrum of Betelgeuse is probably of the third tjpe. 



