19Ji 



KNOWLEDGE. 



[October 1, 1891. 



know, a measurable parallax has not yet been found for 

 any variable star having a rcf/iilar period. Another fact 

 which may perhaps suggest (although it does not jirorc) 

 great distance, is that few of the long period variables rise, 

 even at maxiumm, above the range of naked eye visibility. 

 As in the case of most rules, however, there are exceptions 

 to this one ; Mira Ceti, R. Hydras, and y^ Cygni being the 

 most notable examples. 



The evidence in favour of great distance -is further 

 strengthened by the fact that none of the Variable Stars 

 have any considerable proper motion. The list of proper 

 motions greater than one second of arc per annum, given 

 in Miss Gierke's " System of the Stars," does not contain 

 a single known Variable Star. We seem, therefore, to 



have evidence that the Variable Stars lie at a vast 

 distance from the earth. How is this peculiarity to be 

 accounted for '? The only plausible explanation I can see 

 is that the Sun and Solar system do not he in a region 

 of Variable Stars, 'i'hc periodical increase and decrease of 

 sun-spots may possibly denote some s}inill fluctuation of 

 light in our Sun, but seen from the nearest fixed star, this 

 variation of light, if it has any real existence, would be 

 quite imperceptible, and the Solar light would probably 

 seem to be invariable. Our nearest neighbours iu the 

 Sidereal System, a Centauri, 61 Cygni, Lalandi 21,185, 

 Sirius, &c., appear constant in their light, a proof that in 

 a large region of space surrounding the Sun there is not a 

 single Variable Star. 



Map on t-lie I.sographic projection, showing the place of twenty Short Period Variable Stars with reference to the Milliv "Way and 

 Nebidie. The Short Period Variables are represented by crosses and the Nebulse by dots. 



Eeji.\eks by a. C. Eanyaed. 



In order to exhibit to the eye the grouping of the short 

 period variables mentioned by Mr. Gore, I have plotted 

 down the places of the 20 variables of short period referred 

 to on one of Mr. Proctor's pairs of maps, showing the 

 distribution of nebula) with respect to the Milky Way. It 

 will be seen that 19 out of 20 of these short period variables 

 lie on or near to the region thickly strewn with stars and 

 nebulous matter which we know as the Milky Way. This 

 same region is also rich in red stars and in stars exhibiting 

 bright lines in their spectra, as well as in large and 

 irregular nebute, and in star clusters — while the smaller 

 nebuljp seem to avoid it, and to cluster in the poles of the 

 Milky Way. W Virginis, the only one of the short period 

 variables which falls at a considerable distance from the 

 Milky Way, has a comparatively long period of 17'27 days, 

 and its spectrum seems to differ from the spectra of other 

 short period variables. 



The distribution of short period variables with respect to 

 the I\Iilky Way was pointed out some years ago by Prof. 

 E. C. Pickering. Miss \. JI. C'lerke, in her " System of 

 the Stars," p. 115, remarks that within the zone of the 

 Milky Way these short period variables display "an 



evident disposition towards clustering where the Milky 

 Way di^•ides in Cygnus ; the variables follow its southern 

 branch, and they are thickly sown over the whole region 

 from Lyra to Sagittarius." Indications indeed abound, that 

 the conditions of variability and even of particular kinds of 

 variability are localized iu space. Thus in Sagittarius no 

 less than four r'tars fluctuate in periods of six to seven 

 days. 



The absence of any appreciable parallax in variable 

 stars need not necessarily bo due to their great distance. 

 All modern determinations of parallax are based on 

 measures of the distance of the star whose parallax is 

 sought from small stars in its neighbourhood. If 

 variable stars occur in groups or are situated in clusters 

 of small stars, we should expect to find no relative 

 parallax compared w'ith small ' stars situated at about the 

 same distance from us as the principal star. We have so 

 few stars showing undoubted parallax that it would be 

 unsafe to base any important general conclusion on the fact 

 that no star of the variable class has yet been discovered 

 showing such parallax ; but it is more remarkable that no 

 variable star shows any considerable proper motion. I am 

 indebted to Mr. Gore for the following interesting table. 



