June 1, 1889.] 



♦ KNOW^LEDGE ♦ 



171 



Such an arrangement as this cannot be the result of chance, 

 and unless all very distant stars are green or blue, it must 

 be the result of physical connection. W. H. S. M. 



As to the nearer stars being mostly double, Mr. Monck 

 goes even further than I do in rejecting small parallaxes as 

 doubtful. There are fourteen stars in Professor Young's 

 list with parallaxes all considerably over a tenth of a 

 second, which Mr. Monck does not take into account 

 because their parallax has only been determined by one 

 observer. Of these fourteen, eleven are single and only three 

 are double stars. Facts such as these are best illustrated 

 by presenting them to the eye, and I have therefore made 

 a diagram showing the distance of twenty stars with large 

 parallaxes. The parallaxes adopted for this purpose are 

 those given in Professor Young's list, with the addition of 

 some of Dr. Elkin's determinations. The diagram shows 

 six of the seven stars referred to by Mr. Monck, but 



have been equally weighted, and the probable errors adopted 

 by the observers are not taken into account. It will be 

 seen that out of the twenty stars in the diagram, lying 

 within a distance corresponding to a parallax of an eighth 

 of a second, seven are double, one is a triple, and twelve are 

 single. If we count the sun as a single star, the proportion 

 which double and triple stars together bear to single stars is 

 as two to three, in this small region of space. I admit that 

 our list of doubles is probably still very imperfect, but our 

 list of near stars is probably still more imperfect. If the 

 stars are evenly distributed in space, we should expect to 

 find seven times as many stars between the first and 

 second concentric circles in the diagram as within the first 

 circle, for the two circles correspond to spheres whose 

 volumes are as one and eight, and there should be fifty-six 

 times as many stars between the second and third circle as 

 within the first circle, or seven times as many as within the 

 .sphere correspjnding to the second circle. There is evidently 



|S.r.., ~ 





► /iC.„../..;„ 



I 







• <rCr^„,,-s 



DiAliUA.VI SlI0WlN(; THE DISTANCES OF TWENrV STAUS FOR WHICU LAH';K PaKALLAXES HAVE HEEX FOIT.ND. 



Aldebaran, with a parallax of 0"'11G, would on the .scale 

 of the diagram have fallen outside the page. Mr. Monck's 

 contention will not be affected by this, for I amnot count 

 Aldebaran as a binary ; there is no known physiwvl con- 

 nection between it .and its niinuto companion 108'' off. The 

 scale adopted is ton uiilliinctors for (he distance of a star 

 with a parallax of one second of arc, or one millimeter for 

 20,620 radii of the earth's orbit. 



Where there have been several determinations of the 

 parallax of a star I have .adopted a mean value, and have 

 drawn a ra<lial Hue indicating the range in distance corre 

 spending to the various determinations.* All observations 



* The short line through the place of a Ceotauri corresponds to 

 the determination.s of Henderson in 1842, 0"-913; Maclear in 1851, 

 0"!)19; Moesta at Santiago in 1804, 0"SSO; and Gill and Elkin in 



a greit range in the real magnitudes of stai-s, and probably 

 many small stars having large parallaxes have not yet been 

 detected. 



As to the colours of double stars, Mr. Monck rejects my 

 facts and ai)peals to authorities; but in this instjince I 

 venture to differ with Sir John Her.schel, wlio has, I think, 

 been a little too sweeping in the statement that " no green 

 or blue star of a decided hue has hitherto been found 

 unassociated with a companion brighter than itself." The 

 stars referred to in my note are all described as decidedly 



1882, 0"-7.'>0. The sliort line thr.mgh Lalande 21, 2oS, corresponds 

 to the determination of Auwers in 1863, 0"-262, and Kniger in 

 1864, 0"-2o0. Tlie line through a Lyra should extend outside the 

 diagram. Dr. Elkin has recently determined its parallax to be 

 only 0"034 ± -045 ; Peter's determination in 1840 was 0"103 : and 

 Halls in 1881, 0"l;M. 



