NATURE 



[A/ay 26, 188] 



On a review of the whole question there seems no doubt that 

 tlie annual parallax of 61 Cygni is nearer to the half second 

 found by Strnve, than to the third of a second found by Bessel. 



To exhibit the nature of the evidence which is available for 

 the solution of such a problem, a diagram showing the second 

 series of observations has been prepared (Fig. 2). The ab- 

 scissae are the dates of the second series of observations made 

 at Dunsink. The ordinates indicate the observed effect of 

 parallax on the difference of declinations between 61 (B) Cygni 

 and the comparison star. Each dot represents the result of the 

 observations made on the corresponding niglit. The curve 

 indicates where the observations should have been with a paral- 

 lax of o"'47, the effect of the parallax in declination being 

 only o"'40. The discordances are not so great as might per- 

 haps be at first thought. The distance from the top of the 

 curve to the horizontal line represents an angle of four-tenths of 

 a second. This is about the apparent diameter of a pennypiece 

 at the distance of ten miles. The discordance between the 

 observations and the curve is in no case much more than half so 

 great. It therefore appears that the greatest error we have 

 made in these observations amounts to but two or three tenths 

 of a second. This is equivalent to the error of pointing the 

 telescope to the top edge of a penny-piece instead of to the 

 bottom edge when the penny-piece was fifteen or twenty 

 miles off. 



The entire quantity to be measured is so small 

 that the errors, minute as they are, bear a large 

 proportion to the parallax. In this lies the 

 weakne-s of such work. By sufficiently increasing 

 the number of the observations, and by discussing 

 them w ith the aid of the method of least squares, 

 considerable confidence may be attached to the 

 results. 



Gtvombridgc 1830. — This star has been the 

 subject of much parallax work. It has a tele- 

 scopic proper motion of seven seconds annually. 

 Mr. Huggins or Mr. Christie could perhaps 

 ascertain by the spectroscope what its motion 

 may be in the line of sight. From the theory of 

 probabilities the total proper motion may not 

 improbably be nine seconds. We shall however 

 take it at seven seconds. The parallax has been 

 determined by Struve and by Briinnow. It is 

 very small, being one-tenth of a second. The 

 actual velocity of 1830 Groombridge nnist there- 

 fore be at least 70 radii of the earth's orbit per 

 annum, or 200 mdes per Fecond. 



Newcoiub has employed this result to throw 

 light on the question as to whether all our stars 

 form one system. If an isolated body in our 

 system is to remain there for ever, the theory 

 of gravitation imposes the imperative condition 

 that the velocity of the body mu-.t not exceed a 

 certain amount. Assuming that the stars are 

 100,000,000 in number, and that each star is 

 five times as large as the sun, assuming also 

 that they are spread out in a layer of such 

 dimensions that a ray of light takes 30,000 years to pass it, 

 Newcoinb shows that the critical velocity is twenty-five miles 

 per second. 



As this is only the eighth part of the velocity of Groombridge 

 1S30, we are thus led to the dilemma that either the masses of 

 the bodies in our system must be much greater than we have 

 supposed, or Groombridge 1830 is a runaway star, which can 

 never be controlled and brought back. 



Si-arch for Stars 'vith Parallax. — The lecturer has been en- 

 gaged for some years at Dunsink Observatory in a systematic 

 search for stars « hich have an appreciable parallax. Up to the 

 present about three hundred stars have been examined. In the 

 majority of cases each of these stars has been observed only 

 tw ice. The dates of the observations have been chosen so as to 

 render the elTects of parallax as manifest as possible. It is not 

 of course expected that a small parallax of a few tenths of a 

 second could be detected by this means. The errors of 

 the observations would mask any parallax of this kind. It 

 seems however certain that no parallax could have escaped 

 detection if it equalled that of o Centauri, i.e. one second 

 of arc. 



The stars examined have been chosen on various grounds. It 

 had J-cea supp^s-'d 'ha' some of the red stars were possibly 



among the sun's neighbours, and consequently many of the 

 principal red stars were included in our list. No conspicuous 

 jjarallax has however been detected in any of the red stars up to 

 the present. Many of the principal double stars are also included 

 in the list. Other stars have been added on very various grounds ; 

 among them may be mentioned the Nova, w hich some time ago 

 burst out in the constellation Cygnus, and dwindled down again 

 to a minute point. The earth's orbit however does not appear 

 any larger when seen from Nova Cygni than from any of the 

 other stars on our list. 



Groombridge 1 6 18. — We have however found one star which 

 seems to have some claim to attention as one of the sun's neigh- 

 bours. The star in question is Groombridge 1618. It lies in 

 the constellation Leo, and is 6'8 magnitude. Groombridge 1618 

 has a proper motion of i"'4 annually. From a series of measure- 

 ments of its distance made on fifty-five nights from a suitable 

 comparison star the parallax of Groombridge l6i8 appeared to 

 be about one-third of a second. As this seemed to be a result 

 of considerable interest, measures were renewed for a second 

 series of forty nights. The result of the second series con- 

 firms the first. Measurements of the position angle were 

 also made at the same time. Some difficulties not yet fully 

 explained have arisen, but on the whole the measurements 

 of the position angle seem to confirm the supposition that 

 the parallax of Groombridge 1618 is about one-third of a 



Parallax in declination of 61 (E) Cygni. Abscissse indicate dates 

 indicate parallax ; dots indicate observations. 



second. The orbit of the earth viewed from Groombridge 

 l6l8 is about the same size as a pannypiece at the distance of 

 seven miles. 



.'Secular Proper Motions. — Geologists have made us acquainted 

 wiih the enormous intervals of time which have elapsed 

 since the earth first became the abode of living animals. Re- 

 garding a period of 50,000,000 of years as comparable with 

 geologic time, some considerations were adduced as to the 

 effect of proper motions during such an interval. It was 

 pointed out that in all probability none of the stars now 

 visible to the unaided eye can have then been visible from the 

 earth. 



The Nature of Space. — The possible connection cif parallax 

 work with the problems of the nature of space was then alluded 

 to. It was shown that if space be hyperbolic the observed 

 parallax is smaller than the true parallax, while the converse 

 must be the case if space be elliptic. The largest triangle acces- 

 sible to our measurements has for base a diameter of the earth's 

 orbit, and for vertex a star. If the defect of the sum of the three 

 angles of such a triangle from two right angles be in any case a 

 measurable quantity, it would seem that it can only be elicited 

 by observations of the same kind as those which are made use of 

 in parallax investigations. 



