XIX .
On the parallax of the fixed stars .
By J. F. W. HERSCHEL , Esq. M. A. Sec. R. S. Communicated January 19 , 1826 .
Read March 9 and 16 , 1826 .
T HE determination of the existence and amount of annual parallax in the fixed stars , was the object which originally drew the attention of astronomers to the examination and measurement of double stars , upon a principle suggested originally by GALILEO , and improved on and more fully developed by my Father in one of his earliest communications to the Royal Society , Phil. Trans. 1782 ; according to which , two stars placed very nearly in a line with each other and with our system , ought , if situated at very different distances , in the line of sight , to be subject to periodical variations in their apparent distance from each other , according as the earth in its annual motion approaches to or recedes from that point in the plane of the ecliptic , where the line joining them , prolonged , would meet it .
The difficulty of determining the distance of two stars of a double star with the necessary degree of exactness for this delicate purpose , has however hitherto put a stop to this enquiry ; and , as it is now rendered extremely probable that the parallax at least of the generality of stars is much below i " , must continue to do so , until some very great improvement of micrometers shall'enable us to measure tenths of seconds with as : great certainty as we at present can units .
I do not find that it has been noticed , however , that parallax must occasion a periodical change in the angle of position , as well as in the distance of the two stars composing a double star , and that this variation is much more susceptible of ready and exact appreciation with our present micrometers than that of their distance .
To render this sensible , we need only remark , that the effect of parallax is to cause each star to describe ( apparently ) a small ellipse in the heavens , the major axis of which is parallel to the ecliptic , and the minor in the direction of a secondary to that great circle , the true place of the stars being in the centre .
Were these ellipses of the .
same magnitude , for each of two contiguous stars , the line joining their apparent places ( which are necessarily homologous points in the circumferences of each ) would preserve its parallelism at all times ; but as the axes of the ellipses are reciprocally as the distances of the stars , that parallelism cannot obtain when the stars are situated at very unequal distances from the earth , and an alternate increase and de- , crease of the angle of position made by this line with any .
fixed direction must be the necessary consequence .
To estimate the extent of this variation , let us conceive two stars so situated as to have their apparent line of j"unctionr in the direction of a secondary to the ecliptic , and therefore , at right angles to the major axes of their parallactic ' ellipses._ .
let their distances from us be such that the nearer one shall have a parallax of I " , and the farther no appreciable amount of it .
Also , let their apparent angular distance from each other be g ' .
It is evident that the variation alluded to will equal the angle subtended by a line of 1 " in length , at a point 5 " distant from its middle , that is , to 10 25 ' .
Now this is a quantity which is quite beyond all conceivable ' limits of error of observation in the measurement of double and for stars nearer than 5 ' the amount is of courseproportionally greater .
Thus for two stars , at only i ' distance from ' each other , of which the one is affected by parallax to the -amount of i " , and the other not all , the annual variation in position will amount to upwards of 530* When the distance between two stars amounts to no more than s " , and in many cases even when they are still nearer , Mr. SOUT 's and my own experience in the use of the position micrometer , as well as Mr. STRUVE'S numerous and excellent measures recorded in the Dorpat Observations , lead me to believe that a single degree in the angle of position is a quantity distinctly appreciable in the mean of several sets of measures carefully taken on different nights of observation , even when a considerable inequality in the stars , or other unfavourablecircumstances exist ; and were observations continued for a series of years at the proper times , and made with the care so delicate a subject of research would peculiarly call for , there can be no doubt that a greater degree of precision might be obtained ; and it certainly seems not too much to assume , that half that quantity , or 3o ' of variation in the angle of position , zf regularly periodical , must ultimately be detected .
This conclusion appears warranted by the interesting re-examination by Mr. SOUTH of stars pointed out in our joint paper on Double Stars ( Ph. Tr. 1824 .
iii .
) as having probably a relative angular motion , or as otherwise remarkable , lately communicated by him to the Royal Society , ( See Part i. of the present vol. ) in which we find instances of coincidence between calculated and computed , motions falling within the limit assumed at the outset-too frequent , and too remarkable to be the effect of mere accident .
Now , the tangent of 30 ' to a radius of A " corresponds to a subtense of 0o".o6 , or Tooth of a second , so that a difference of parallaxes to the amount of a 40th of a second , existing in the two stars of a double star so circumstanced , could scarcely escape detection ; and that even much less quantities than this , under favourable circumstances , might be rendered sensible , I think may fairly be concluded , when we consider that in this estimate , the data are certainly assumed within bounds .
No account is here taken of the improvements in the position micrometer , which may reasonably be expected because the object at present is only to appreciate the degree of delicacy which the method now proposed may lay claim to , with our present instruments and habits of observing , and to compare it with those which have hitherto been resorted to in the investigation of parallax .
In selecting stars for examination , it appears to me that we ought by no means to confine ourselves , by assuming it as a universal law that the brightest stars are the nearest to us .
From what we know of the variety of nature and the enormous differences in point of magnitude between the bodies of our own system , it seems improbable that the real magnitude and brightness of the stars should be confined within narrow limits .
Their distances are equally undeter .
mined ; nor have we any reason whatever to conceive these two elements related to each other .
There is not therefore the slightest a priori improbability , in supposing that among stars of apparently equal lustre , the greatest diversity of distance may exist , or that innumerable of the minutest stars visible in telescopes may be nearer to us than any of those of the first magnitude ; and consequently , that that delicate element in search of which astronomers have exhausted refinement , may with nearly or quite equal probability of success be sought among stars of far inferior magnitudes .
The proper motions of the stars afford an argument from analogy they bear no relation to their apparent lustre , and by far the greatest proper motions known belong to stars low in the scale of magnitudes .
If these remarks have any foundation , it must be obvious that we ought not to be deterred from the research of parallax by the smallness of the stars composing a double star , or their approach to equality .
However , we should except here stars in , or very near the milky way , below the 7th or 8th magnitudes , because the probable laminar form of this great sidereal stratum affords a presumption almost amounting to certainty , that minuteness is here , on the average , an effect of distance .
But , on the other hand , such large stars as i8 Orionis which are situated in , or near the borders of the milky way , with small ones near enough to constitute them double stars , have an additional claim to examination , from the additional probability thus afforded of being favourably placed for the detection of parallax : and moreover , this consideration affords plausible grounds for a belief , that , in situations remote from the milky way , minuteness , on the average , is not the effect of distance .
It is hardly necessary to insist on the great advantages presented by the method here proposed , in its complete exemption from those instrumental errors depending on un-steadiness , erroneous graduation , and expansion , and from all that uncertainty on the score of refraction , and any doubts still remaining as to the magnitudes of the constants of aber .
ration , nutation , &amp;c. , which so much embarrass astronomers .
A good telescope and a good micrometer are all the instruments required .
This advantage is really incalculable .
Instead of confining our attention to one or two principal stars it puts an almost unlimited range of objects in our power , by enabling us to employ in this research the largest telescopes , and thus easily obtain measures of those stars , which , from their faintness , must present insuperable difficulties with instruments of ordinary apertures .
In selecting objects for examination by this method , we must be chiefly guided by their angles of position and distances .
Taking such whose distances are below 15 " , ( which on the supposition of 2 ?
periodical variation of position , correspondds to 8 of a second of relative parallax in stars properly situated ) , the angle of position ought to be such , that the line joining the two stars shall point as nearly as may be , to the pole of the ecliptic .
Ten , twenty , or even thirty degrees of deviation either way from this direction , will however not materially vitiate the application of this method to stars near the ecliptic , while , for such as have considerable latitudes , proportionally greater deviations may be allowed , and within thirty degrees of the pole of the ecliptic this element is of comparatively small moment .
In general , to ascertain whether any double star is or is not favourably situated for the application of this method , we must ( if we would take up the problem on strict mathematical grounds ) , proceed as follows : Let I represent the longitude , and +x the N. latitude of the star , aits angle of situation , * or the angle included at the star by great circles joining it and the north poles of the equinoctial and ecliptic respectively , C being considered positive for stars in the western hemisphere ( or that whose pole is the point cr ) , and negative for stars in the eastern , whose pole is &amp; .
Let also o represent the sun 's longitude at any time , and call a the maximum semi-annual parallax or the angle ( expressed in seconds ) , which the radius of the earth 's orbit would subtend if perpendicularly presented to an eye at the distance of the star .
It is obvious then that a will represent the major semi-axis of the star 's parallactic ellipse , and a. sin. x the minor , so that its excentricity = Va2 a2 .
sin. A !
=a .
cos. X , which if we call a e , we have e= cos. X. The star will appear to describe this ellipse in the direction npsf .
Its motion in it will however not be uniform , but equal areas will be described in equal times about its centre ( or the star 's mean place ) : this is evident , because the area described by the star in the parallactic ellipse round its centre is the orthographic projection on the surface of the heavens of that described by the earth round the sun in its orbit .
This consideration gives us at once , the equation tan .
a-sin .
I. cotan .
( e- ) ; ( I ) where 0 represents the elongation of the star in its ellipse from the eastern extremity of its major axis reckoned in the direction npsf .
Let ir be the angle of position of the small star , ir being reckoned from a parallel to the equinoctial , in the same direction npsf and from the east , so that the nf quadrant shall line joining the two stars with the parallel , by this name , it becomes necessary to distinguish them , and the expression used in the text is perhaps also more correct than that in common use .
correspond to values of Xi between o and , go9 .
the np to values between 9g0 and 1800 and so on .
It is evident , then , that when the large star is in the point of its apparent orbit immediately in the line joining the mean places of both , its elongation , or the value of 8 corresponding to this situation will be -fa .
Since it is only the difference of parallaxes which this method can render sensible , we may suppose the small star a fixed point , and since the dimensions of the parallactic ellipse may be supposed small in comparison with the distance of the small star , two tangents drawn from the latter to the circumference of the former , and which , of course , mark the situation of the large one in its apparent orbit where parallax has the greatest effects in opposite senses on the angle of position , will nearly meet it in the two extremities of a diameter conjugate to that passing through the small star .
To determine the direction of the conjugate diameter , we must have recourse to the general equations of the ellipse from its centre with polar co-ordinates .
Thus r being the length of any semi-diameter CP , whose angle D of elongation from CA , is ACP =cP and r ' , that HA of its semiconjugate CD , the angle ACD being called q ' , we have , H and S being the foci , and calling CN the abscissa from the centre , x , by a property of the conic sections , SPa-exa.ercos.cP ; HP=a+exa+ er.cos.q ; whence , SP .
HPCD2=rJ--a2 _.e2 r. cos. 2 .
SP .
HP have also1 , I ' Cos._ But we have also , r2= r %(I11 Eliminating from these three equations , r and r ' , we get the relation required between e and qi , viz. CO. 2 Sill.9 C05* m= t--(z-e ' ) e0 .
cos. P This equation will however be reduced to a much more convenient form for our present purpose by a transformation , viz. sin. 9 " I -Cos.2 tan .
Pc'2 COS. 912 ( I -e2)2 .
Cos. sin. 9 that is simply tan .
p'= ( i e2I ) .
cot .
P the sign being prefixed in extracting the root , became in the ellipse CP and CD lie in different quadrants .
Now in the case before us we have m-C7r andp'O ; so that tan .
6=_ we ) .
cotan .
( ra But ( i-e2 ) = sin. xA , and by equation ( i ) -tan .
6= sin. I. cotan .
( -1 ) .
Hence substituting , we get sin. .
tan.(o-l)=_ sin. 1 .
tan .
( X_ a ) ; ( 2 ) This equation gives at once the value of o the sun 's longitude , and therefore ( by consulting an ephemeris ) the time of year sought , by an exceedingly simple process adapted to logarithmic computation .
The actual advantage or disadvantage in point of situation in the case of any particular star , is expressed by the magnitude of the whole change in the angle of position produced by parallax , i. e. by the angle subtended by the ellipse at the small star , or , ( calling the distance of the latter from its centre A ) on the original supposition of the whole effect being small , by the expression P= 2- .
sin. ( p'q ) .
Now if we substitute for r ' and sin. ( Ip'f p ) their values in terms of p , or --deduced from the equations already stated , by proper eliminations , we find P/ -cos .
X V'I 2 cos. whence the maximum effect of parallax on the angle of position of any double star is readily computed ; the expression is inconvenient for logarithmic computation , but this inconvenience is obviated by the very obvious substitution Cos. X. Cos.(i-o)=COs .
M ; P sin. M. ( 4 We also have P sin. P si .
M r. ( 5 ) 2a-S .
sin. M sin. M from which we may deduce the value of 2 , a the maximum of parallax when the total effect ( P ) on the angle of position is-known .
sin. M3 If we take -P -so ' we have 9 , a=R siin .
M ( 6 ) which expresses the amount of annual parallax in any proposed star which will be indicated by a periodical variation of so ' in the angle of position .
The smallness of this amount is a criterion ( mathematically speaking ) of the favourable or unfavourable nature of the individual star for researches of this kind .
By these theorems the proper times of year for observa. .
tion , and the amount of difference of parallax , appreciable by a variation of so ' in the angle of position , have been computed for such stars among those observed by Mr. SOUTH arid myself , in the paper above alluded to , as appear favourable to the application of the common wire micrometer to this enquiry , in the method now proposed .
As the number of known double stars increases , others .
may easily be added to the list .
The Catalogue of 460 double Stars observed by Mr. SouTu , and lately communicated to the ' Royal Society , will furnish a great many ; and Mr. STRUVE'S immense cIlection , amassed in his reviews with the large refractor , as well as a collection of minutedouble stars encountered in my own 2o-feet sweeps , and which will shortly be published , doubtless many more ; so that it is rather intended to give the following list as a specimen of a more complete one , than as including all , or nearly all , the stars it is desirable to examine with this view .
Meanwhile , to enable others not conversant with algebraic symbols to extend the list for themselves , if so inclined , I shall here set down the whole work of calculation for one star .
Calculation for s55 Piscium .
R. A. -h Is 30 ' , Decl. 70 49'N .
Long. =-I4o 24 ' .
Lat.-A=.+ 60 32 ' ; Angle of position =X= 2980 4 ' ; Distance a1 2'e5 .
sin. obliqY .
+ 90600oi ( z ) tan .
( ?
v-a ) -1P09285 ( 1 ) COS. lat. + 9*99717 cos. R.A. + 999985 sin , long .
+ 888490 ( 2 ) COS. ( -ah)+ 8-90574 ( add ) -19X97775 ( 3 ) ( add ) cos. M + 8.90291 ( i ) are .
comp..0 os l ) at .
CO~nP +o oozS3 ( )zX log .
sin. lat. +8 ' 1X2 4 log .
a+ 1:04798 COs .
a- .
tan .
30'+ 7 94086 sin. + 9'60280 .
( subtr .
) tan .
( G0- ) _1-I i*8656i C 23 ? ?
27 Q0 -i 2700 47 ' ( add ) +8 98884 -98 414 24 ( 3 ) sin. M+ 9*9986x sr.-o ' 274 37 53 275 1 !
( subtr .
) log .
2a+ 8-90023 _+ 6S 95 I2a o0og8 Dates by Nautical 5Dec .
27 Almanac .
June 277 If we wish to avoid the computation of the latitudes and longitudes which this supposes known , they may be taken with sufficient precision from any good map , or from a large celestial globe , as exactness is not required for this purpose in these elements .
Specimen of a List of Stars favourably situated for the Investigation of Parallax by the Method proposed in this Paper .
General NumR .
A. Amount of annual ber in Messrs. Stars ' Name , &amp; c , for Declination Times of the year most proper Parallax indicated by a H. and S 's .
1820 .
for 1820 , for Observation .
periodical variation o Observations .
W30 ' in the angle of pos .
h. M. 0I 35 Piscium ... ... ... . .
o67 49 N June 27 , December 27. .
0.098 20 v Arietis .
1 44 I8 25 N January 31 , August 3 ... .085 25 a Piscium .
1 53 1 53 N January I8 , July I9 ... ..047 38 32 Eridani. .
3 45 3 30 S February I2 , August.5 .
07I 39 Persei ... ... .
3 46 39 29 N February I9 , August 23.08I 46 55 Eridani ... .
35 99S March 6 , September 8 ... 103 47 w Auriga ... ... 4 47 37 36 N March 5 , September 7.0o6 53 Rigel ... ... .'.5 68 25 S February 25 , August 29.084 55 II8 Tauri ... . .
5 I8 25 0N March io , September 13.052 56 32 Orionis .
5 2I 5 48 N March 7 , September g.013 59 33 Orionis ... . .
5 22 39N March 6 , September'8 ... 020 67 Orionis .
5 32 23S March I8 , September 20.025 366 4i Aurig. .
5 58 48 44 N March 21 , September 23.077 69 8 Monocerotis .
6 I4 4 4I N March 21 , September 24.I36 76 38 Geminorum .
6 44 I3 24 N April I , October 4.049 8o z Geminorum. .
79 22 I8 N April 5 , October 9.064 88 II Cancri .
7 58 28 ON April i6 , October 19 ... .041 93 pt Cancri .
8 i6 27 31 N April i , October 23 ... .5 1 94 I8 BODE , Hydrx 8 26 7 I5 N April 26 , October 29 ... 096 96 144Of '45 .
8 39 7I 27 N April 20 , October 23 ... .076 98 57 Cancri .
8 43 31I x6 N April 27 , October 30.o020 99 17 Hydre .
8 47 7 17 S May I 2 , November 14..053 I02 Cancri1I94 ... .
8 57 23 42 N April 30 , November 2 ... o67 114 Lconis 145 ... .
10 II 7 22 N May 23 , November 24 ... o6o 128 9o Leonis .
11 25 17 48 N June 4 , December 6 ... ..039 133 65 Ursae. .
46 47 29 N May 2I , November23.035 134 2 ComOC ... ... .
II 255 2 28 N May 17 , November I8 ... 038 147 118Of145 12 25 75 46 N June 4 , December 6 ... ..053 152 STR .
422 12 40 4 48 N June 30 , December 30 ... 089 I55 STR .
424 ... .
12 44 I6 ON May 29 , November 30..072 I6I 54 Virginis ... .
13 4 17 5S January 13 , July 15 ... .06o 167 8I Virginis ... 13 28 6 57 S January I4 , July i5.038 73 98 of 145 ... ... .14 56 14 N January '5 , July 17 ... ..054 Specimen of a List of Stars-continued .
General NumR .
A. Amount of Annual ber in Messrs. Star 's Name &amp; for Declination Times of the year most proper Parallax indicated by a H. and S 's .
1820 .
for 1820 .
for Observation .
periodical variation of Observations .
30 ' in the angle of pos .
I_ .
m. o i76 STR .
456 ... ... .
14 13 6 56 S January 23 , July 26 ... .
o.0o6I STRm 457 ... . .
14 49 i6 N January 14 , July I6.064 188 39 Bootis ... ... 1 4 44 49 27 N April z , October 5.040 193 44 Bootis ... ... 1 4 58 48 21 N February IO , August 13.020 ' I94 STR .
474..-. .
14 59 9 55 N February 2 , August 4.042 201 n Corona B ... .
I5 i6 30 57 N January 28 , July3I.014 205 'Serpentis ?
... .
15 z6 1I 9N February 6 , Augustg 9..027 2o6 Librae 178 .
5. .
IS 30 8 II S February II , August 14.104 21I 1T .
85 ... ... . .
I5 47 I 39 S February 7 , August io.08S 212 III .
IO3 ... ... .
I5 48 3 56 N January 3I , August 2..132 2z8 g5 Ophiuchi. .
I6 I5 23 31 S February zS , August 29.036 240 i6 46 19 IS S March 3 , September 5.064 245 39 Ophiuchi ... I7 7 24 5S March 8 , September IO .
II I z6z Ioo Herculis ... I8 I z6 5N March 22 , September 24.125 265 I. 86 .
O I8 I2 25 z8 N March 2I , September 24.040 269 39 Draconis ... ,8 21 58 42 N April z8 , November I.031 271 _ I8 30 4I 7N March zz , September z5.053 274 , I8 36 10 39 S March z8 , October I..049 z80 I8 42 10 47 N April 5 , October 8 ... 042 287 I8 58 6 S3 N April 2 , O ctober.077 295 I1I .
57 ... . .
' .
19 19 20 46 N April 7 , October io..o6z 306 fl Aquilm ... ... .
19 4I II 22 N April 8 , October I I.02 I 3'I s Draconis .
9 49 69 48 N April so , October 23 ... 023 312 4. .
Cygni ... .
19 51 51 38 N January 8 , July 9.038 313 I. 96.1 ... ... . .
19 56 35 32 N May I9 , November 2 ... 022 317 11 .
96 ... ... . .
20 3c 19 N April 30 , November 2 ... 045 320 I. 95..20 14 54 48 N January 13 , July 15 ... ..035 323 g Capricorni II 20 20 I8 24 S April 22 , October z6 ... 036 326 20 32 38 5N May z6 , November27..085 343 STR .
751 .
* -22 I6 65 So N March 2 , September4 ... 033 349 Aquariiz2I3 22 34 9 II S May 28 , November 28..03I 352 _ 22 59 31 SI N February24 , August28.076 3 54 94 Aquarii 23 10 I+ z6 S May so , November zi ... 133 356 107 Aquarii 23 37 19 4I N June 3 , December 5i..,.045 359 o Cassiop ... ... 23 50 54 45 N January I , , July 13 ... ..ozIf I am not mistaken , then , we have here a method of investigating parallax very much more delicate than any which has yet been proposed-open to no insuperable objectionsdemanding no laborious reductions-and depending on a branch of practical astronomy , which is now happily cultivated with the assiduity it so richly deserves .
It may not then be too much to hope , that in a few years we shall have data for a positive decision , respecting at least a certain number of individual stars , and thus be enabled to form a much better judgement as to the probable maximum of parallax in any .
Meanwhile , the number of stars to which the method is applicable must not be judged of from the specimen above given .
The only position micrometer of which I have any knowledge from practice is the wire micrometer ; but the double image micrometer , if used in a manner , for the knowledge of which I am indebted to Captain KATER , and which consists in bringing the ist image of one star almost but not quite in contact with the 2nd image of the other , and in the same straight line , presents the singular advantage of an attainable exactness proportional to the distance of the stars ; and by employing angles of position thus measured , stars of the 4th , 5ath and 6th classes are rendered equally , or more available than those of the ist , 2nd and srd ; so that our range of objects becomes , by the use of this instrument in this manner really unlimited .
It is possible , that by some an apology may be thought necessary for a communication like the present-offered as it is to the Royal Society as a mere proposal-not followed out into actual practice .
A variety of other pursuits , however , and circumstances less under my own control , have hitherto prevented , and will in all probability continue to , prevent , my engaging actively in the very extensive and laborious series of observations required for this purpose .
Should I do so in future , to which however I by no means intend to pledge myself , my progress must of necessity be extremely slow .
But should the method itself appear really to possess the advantages I am inclined to ascribe to it , practical astronomers I am sure will hardly impute its publi- .
cation , even unaccompanied with observations , as a fault ; and if their energy and perseverance should anticipate me in its application , I hope I should be the first to acknowledge that the merit of the discovery of parallax must rest with him , who , whatever be the method he may pursue- , shall first point out the star in which it exists , and establish it to the satisfaction of astronomers by unequivocal observations .
J. F. W. R. London , Dec. 8 An , ,

