On the Results of Periodical Observations of the Positions and Distances of Nineteen of the Stars in Sir JOHN HERSCHEL'S Lists of Stars , favourably situated for the investigation of Parallax , contained in Part III .
of the Philosophical Transactionsfor 1826 , and Part I. 1827* .
By Lord WROTTESLEY , F.R.S. 8fc .
Received November 14 , 1850 , -Read January 16 , 1851 .
IN these communications , Sir JOHN HERSCHEL shows , that if the component members of a double star occupy a certain position with reference to the pole of the ecliptic , and one of them be supposed to be situated within a given distance from the earth , a change will be periodically produced in the angle of position , and in the distance of the two stars forming the double star , consequent on the motion of the earth in her orbit ; that the maximum of the change in the angle of position will occur at two periods of the year distant from each other about six months ; and he gives formulae by which the epochs at which that maximum occurs may be computed , and by which the amount of parallax due to a given change in the angle of position may be also found for each star ; also lists of stars favourably situated for the investigation of that element , with the times of the year at which they should be observed , and the amount of parallax , which an observed change of 30 ' in the angle of position would indicate in the case of each star .
On the erection of my observatory in the autumn of 1842 , comprising among its instruments an equatoreal of very considerable power , I determined to employ it in the observations of the double stars contained in these lists , with the view of ascertaining whether they exhibited such decided differences in their positions , when observed at the proper periods , as to give good grounds of hope that some definite conclusion might be arrived at , as to the existence of a parallax in the objects observed , capable of being measured , or at the least , confidently announced as subsisting in fact .
In conducting the observations , however , difficulties arose that ' iad not been anticipated : Sir JOHN HERSCHEL'S lists contain sixty-nine stars ; many of these , at the periods announced in the lists as proper for observation , if observed at all , must be observed at distances from the meridian too great , when the delicate nature of the inquiry is considered ; many of them , having south or small northern declinations , are near the horizon when the proper period for their observation arrives .
Another very serious impediment existed in the obligation imposed of obtaining half of the measures in the early morning hours ; the observer , if employed in the observatory on the preceding night , was sometimes fatigued and unequal to the task ; heavy fogs frequently came on at that time , and an enormous deposit of dew on the interior surface of the object-glass often seriously incommoded the observer ; an evil which , until anopening was made in the tube near it , was irremediable , since the use of dew-tubes failed in preventing it .
To these and other causes it is owing , that after more than six years devoted to this course of observation , I am compelled to apologise for the meagre results which , for reasons about to be mentioned , I have still ventured to lay before the Royal Society in their present shape .
Of sixty-nine stars I have only obtained observations of forty-eight , and of these forty-eight , twenty-nine have only been observed at one period of the year .
It is a most discouraging feature in this class of observations , that , however numerous and trustworthy the measures obtained at one period of the year may be , there may still be a failure to procure , at the expiration of six months from their date , measures worthy to be compared with them ; and the function of the parallax , being the difference between the two results , the value of that element is of course affected by the whole amount of error with which either result is charged .
It has often happened that a star has been observed at one of the assigned periods , and that no corresponding observations have been procured at the expiration of six months , or the next succeeding period ; thus , for example , calling the first epoch in the year at which the star is marked for observation , the early , and the second the late period , a star has been observed two or more times successively at the early period , when no corresponding observations could be procured at the following late period .
The question then naturally occurs , whether the communication of the measures actually obtained may not be deemed premature .
I consider , however , that I have proceeded sufficiently far to demonstrate the impolicy of further perseverance , with the means at my command ; the rather that instruments atre now erected both at Liverpool and Oxford , which are pre-eminently suited to this class of observations , and therefore it would be only a waste of time and force , which may be more profitably employed in other ways , to devote any further attention to the inquiry .
I proceed , therefore , without further apology , to describe the means employed , the mode of employing them , and the results obtained .
The instrument employed was an equatoreally mounted telescope of 10 feet 9 inches focal length , with an object-glass of 7inches clear aperture , of which the flint-glass is by GUINAND , and the crown glass English , the whole having been finished and perfected by DOLLOND .
This telescope is mounted in the manner usually adopted by the opticians of this country in a fixed observatory , in the immediate vicinity of my residence in Staffordshire ; the polar axis is formed of four mahogany planks 14 feet 3 inches long and 10 inches square in th the mddle , the pivots of which are of hard bell-metal , and rest above and below on is attached to large and solid stoe piers , which are supported by a foundation of brick-work , joined with cement , and formed into a solid mass of great extent by filling up the space enclosed by an outward circuit of brick with stones and mortar pounded orar o together : this nass of brick-work and stone extends beyond the equatoreal room , and forms also the support of the piers of a 5-foot transit .
The Declination and the Hour Circle are each 3 feet in diameter ; the two verniers of the former read off to 10 " in space , and that of the latter to seconds of time .
To the telescope there is attached a parallel-wire Inicrometer with one equatoreal fixed , and two moveable wires ; the screw heads are divided into 100 parts ; the micrometer is provided with the usual position circle , graduated on silver , with its vernier which reads to 6 ' in space .
The value of one part of the micrometer first employed was 0"115628 , but from the 1st of January 1847 the value 0 " 15641 was used , the first having been determined by Mr. BEAUMONT , the former possessor of the instrument , and the latter by myself .
There are six eye-tubes that can be used with the micrometer , with powers varying from 85 to 820 , but the power almost invariably used in the observations about to be described was 450 ; one of 320 , and sometimes a lower power , was occasionally , but very rarely , employed , and when this occurs a notice to that effect will be found in the Table containin the results of observations subjoined .
The telescope is provided with a clock-work motion ; the performance of the objectglass , so far as its powers have been tested , leaves nothing to be desired , but the mounting of the telescope , notwithstanding the precautions above mentioned , does not appear to be equally steady with that of some other recently erected equatoreals .
Whenever the night admitted of it , ten measures both in position and distance were obtained of each object observed ; when in the sequel I speak of a set of observati-ons , the term is to be understood as applying to all the observations of a single night , which almost always comprised that number of measures both of position and distance .
To each individual measure , whether of position or distance , an arbitrary weight was assigned , and registered at the time of making it , the number 10 being supposed to express a result with which the observer was entirely satisfied , and smaller numbers to denote a less degree of confidence in the value obtained ; in point of fact , however , no higher number than 8 seems to have been used , and the weights employed generally range from 4 to 7 .
The sum of the weights of the indi2x2 335 vidual measures , or the weight of the set , has been divided by 10 prior to its entry in the subjoined tables ; while therefore the number 5 should express the weight of a set of average goodness , the number expressive of that weight is reduced in practice to 4 .
The measures when made are registered in the printed skeleton forms , originally proposed by Sir JOHN HERSCHEL , and adopted , I believe , by almost all observers of double stars in this country .
A number of these forms were bound together in a volume , and care was taken to fill up the different columns , so far at least as they refer to measures of position and distance .
Whenever the weather permitted , a set of measures was obtained of each double star , on three separate days , at each period of the year marked out as the proper one for observation .
The common arithmetical mean was used in taking the average of the partial measures , comprising each set , without taking the arbitrary weights into account .
At the commencement of the series , the zero of the position circle was frequently determined ; but it was found to undergo no alteration that might not be attributed to error of observation ; from that time it became the practice to ascertain it about every two months ; but it was never found to vary more than about 2 ' , a quantity very much within the error of observation in determining an angle , as will appear in the sequel ; no zero was required for distance , as it was the practice to take an equal number of positive and negative readings .
The observations were commenced on the 15th February 1843 , and terminated the 9th October 1849 ; and they were made almost exclusively by three gentlemen , who acted successively as my astronomical assistants , during the progress of the work .
I have always found , that observations are better made when entrusted to one competent individual , who after a short time gains experience , and a facility of manipulation , which cannot be acquired by one who only occasionally observes ; and this uniformity in the mode of observing is particularly desirable , where the quantities to be determined depend on comparisons of observations made at different periods , ; it was a subject of great regret to me , therefore , that I was compelled to'make any change in the staff of my observatory during the course of this investigation , and I only observed myself on a few nights at the time of the first appointment of each assistant .
Mr. GODDARD observed from the commencement of the observations to October 2 , 1843 ; Mr. SIMMS from that time to the 11th of June 1844 ; and Mr. PHILPOTT , my present assistant , from thence to the conclusion .
The results of the observations are embodied in five tables , and it now remains to explain the manner in which those tables have been formed .
The First Table .
In the first table the stars are arranged in the order of their R.A. 's ; and the observations of each day , or sets , follow one another in the order of their date .
An asterisk will be found attached to four of the stars in the list ; these are stars , which , on comparing their mean positions as ascertained in the course of this inquiry , with those given by other observers , exhibit changes of such an amount in their angle of position as to afford satisfactory evidence of orbital motion .
This fact being admitted , it is plain that these stars cannot be rendered available in the present inquiry ; it was deemed unnecessary therefore to extend to them the method of reduction employed in the case of the other stars ; the observations , however , are given with the others in their proper place .
The third and seventh columns of the table contain the arithmetical mean of the individual measures , botl in position and distance , obtained on the day specified in the first column .
The quantities inserted in the fourth and fifth columns headed " probable error and " computed weight , " refer to the positions only , and have been computed from the form ulae 0-45494 C2 = n(n--1 X sum of2 , W--p , where P= probable error of a single set of measures , n= the number of measures , g= the individual errors , and W= the weight of the result ; numbers proportional to the computed values of W being inserted in the Table .
The average probable error of all the sets , 211 in number , is 8 ' 98 , and the average weight , 0'0124 .
The quantities inserted in the sixth and eighth columns , headed " assigned weight , " represent the sum of the weights , divided by 10 , assigned by the observer to the individual measures of position and distance forming the set , or the arbitrary weights of the sets .
The magnitudes and colours of the stars were noted on each night of observation , and are inserted in the ninth and tenth columns ; where one colour only is named , it is to be understood as applying to both stars ; and where two are specified , the colour of the brighter component , or , in cases of equal brightness , of the star called A , is placed first in order .
The eleventh column contains the initial of the observer 's name .
The results of each epoch and year are for the sake of distinctness separated by a blank space .
337 ( Continued .
) Year , Month '*~ Position .
3 g. .
Year , Monthn D Position .
i2 g ) Magnitudes .
Colours .
Obserer .
and Day .
Z. , _P XXII 306 * .
1843 .
June 25 .
-479143 2 ... ... ... 6 8-467 6 6-7-9 White .
G. July 3 .
-501 143 56 ... ... ... 6 8353 6 6-7-9 White .
G. 6 .
509 144 17 ... ... ... 5'5 8-461 5-5 6-7-9 White .
G. Dec. 17 .
-958145 46 ... ... .
4-1 8-521 3-5 6-7-8-9 White .
S. 1844 .
Jan. 1 ] .
'027145 34 ... ... 4-5 8-618 4*2 6-7-9 Reddish and white .
S. 13 .
-033144 13 ... ... .
4-9 8-788 4'7 6-7-9 White .
S. Nov.a 27 .
-906146 0 ... ... . .
64 8'670 6-5 7-9 Blue .
P. 1846 .
Dec. 14 .
'950 146 18 ... ... . .
1-1 8-350 1-3 8-10 Pale yellow and pale blue .
P. 22 .
-972146 3 ... ... ... 41 8'335 3-8 7-8-9 Yellow and pale blue .
P. 1847 .
Dec. 28 .
'988146 ]6 ... ... .
1-7 8-024 1-7 7'8-9 Whitish and pale blue .
P. 1848 .
Dec. 21 .
'972 145 37 ... ... . .
4-6 8-251 4'7 7'8--910 Yellow and pale blue .
P. The Second Table.--'his table contains the means of all the results obtained at the same period of each year , when observations have been obtained on two or more days .
The second column contains the mean epoch of the results from which the means are formed .
The positions in the third column are computed in the ordinary mode by multiplying each individual set by its computed weight as contained in the fifth column of Table I. , adding the products , and dividing by the sum of all the weights .
TheJifth column contains the sum of the computed weights of the individual sets , from which the probable error of the mean in the preceding column is calculated from the equation 1 P= The computed weights inserted in the Tables have been multiplied by 1000 .
The mean distances in the seventh column are computed in the same manner , using , however , the arbitrary or assigned weights in column 8 of Table I. , instead of weights computed as above explained ; for it was not deemed necessary , for reasons which will appear in the sequel , to apply the calculus of probabilities to the reduction of any of the observations of the distances .
The quantities in the sixth and eighth columns , headed " Assigned weight , " are the sums of the weights assigned to the individual sets , from which each mean is formed .
In the case of the four binary stars , to which an asterisk is attached in the Tables , the means , both of the positions and the distances , given in this table are obtained by employing the assigned or arbitrary weights , as before described .
In deference to the objections which may be urged to this mode of reduction , I had originally intended to use the common arithmetical mean in all these cases ; but out of fifty-four means of positions computed strictly by the usual formulae , and compared with means obtained as last described , and also with the arithmetical means of the same quantities , it was found that the mean obtained by using the arbitrary weights differed only in thirteen cases more than 9 ' from the strict mean , and that in eight out of the thirteen cases , the result was rendered more erroneous by using the arithmetical mean , instead of the muean derived from the use of arbitrary weights : there were seven cases out of the fifty-four in which the computed and arbitrary weights gave precisely the same result .
The Third Table.-This Table contains the general means of all the observations of each star , both in position and distance , reduced to the mean epoch of all the observations , which will be found in column three .
r'Ihe positions in column four , and the probable errors and weights in columns Jfive and six , are obtained by combining the means and individual sets ( where one set only has been obtained at any one period ) in accordance with the theory of probabilities , as above mentioned , except in the case of the four binary stars , in which the arbitrary weights are substituted for computed ones , as before explained .
Where two stars are supposed to be of equal brightness , it has been usual to register always the smaller of the two angles , by which the position may be designated , according as the one or the other component be considered as the larger star .
In these cases there is some confusion as to the identity of the two stars , which might be obviated by astronomers agreeing always to designate the northern , or in cases of equal N.P.D. , the east star as A. The seventh and ninth columns contain the sums of the assigned or arbitrary weights .
The eighth column contains the mean of all the observed distances , obtained in all cases by using the arbitrary weights .
The tenth column contains the mean of all the registered magnitudes of each component star , the first No. referring to the brightest of the two components , or in cases of equal brightness , to the star called A. The eleventh and twelfth columns contain the R.A. and N.P.D. of each star , taken from the B. A. C. ; and when the star is not found in that Catalogue , from STRUVE'S of 1837 .
The thirteenth column contains the total number of individual measures in position ; the number of measures in distance being equal in every case , except one , that is specified , one column suffices for both position and distance .
In the notes will be found notices of any remarkable differences between the measures contained in this Table , and those given by STRUVE in his Catalogue of 1837:they are confined to differences exceeding 1 ?
in position , and 0"'200 in distance .
_. .
; , l ' .
' _ ... ... .''* ' * .
1-ns M No. Name .
Mean epoch .
Position , .
gi R.A. N.P.D. .
: : .
: .
e^ *0^ " ft ' , hm0I yArietisa ... ... ... . .
1845 .
*483179 18293 117 68 8887 68 53-53 1 45 71 27 118 2 32 Eridani ... ... ... ... ... 1846 .
*220347 372'90 119 49 6-803 49 6'58-2 3 47 93 24 86 3 wAurige ... ... ... ... ... 1846..434351 53-14 102 89 6'401 88 739'7 4 49 52 21 154 4 118 Taurib ... ... ... .
1844 .
-962195 272'35 182 79 4'955 78 7-69'1 5 20 64 59 150 5 41 Aurigre ... ... ... ... 1845 .
-607353 53'15 216 95 7'947 95 7-68'7 60 41 16 1620 6 &amp; Geminorumc ... ... ... 1844 .
-912197 143'58 78 44 7'422 44 3'2-10'5 7 11 67 45 82 7 Anon .
Cancrid ... ... ... 1845 .
-384353 162'86 122 64 3-655 65 8-193 7 56 62 3 113 8e Cancrie ... ... ... ... .
1844 .
-992 32 362'70 137 74 4-833 71 68-6-8 8 18 62 35 137 92 Come Berenices ... 1845 .
'746239 42'60 148 82 3'916 78 6-68-8 11 57 67 42 166 10 39 Bootis ... ... ... . .
1843 .
'407 43 35 ... ... 31 3-896 31 6-8 7-4 14 45 40 40 46 11 6 Serpentis *h ... ... ... 1846 .
-119194 21 ... 117 3-042 115 3'5-5-3 15 28 78 57 215 12 178 ( Bode ) Librse ... 1845 .
'433 7 343'24 95 30 12-038 30 7'8- .
78 15 29 98 10 56 13 p Draconisk ... ... ... 1845 .
'673 14 15 ... ... 103 3-217 101 7'07-1 17 2 35 20 18414 100l Herculis ... ... ... .
1846 .
'247 2 141'02 959 158 14-005 158 6-4-6-4 18 2 63 55 260 15 579 Struve ... ... ... ... ... 1845 .
'456159 293'39 87 70 5-131 70 8-58-5 18 31 48 49 124 16 287 hands1 ... ... ... . .
1845 .
'274291 73'12 103 27 8-192 26 75-8-9 18 58 83 7 56 17 E Draconism ... ... ... 1844 .
'562356 187'02 20 30 3'010 32 4-89-8 19 49 20 7 64 18 of H 95 ... ... ... ... .
1845 .
'118340 532'78 129 106 3-157 103 6-58-8 20 14 35 9 211 19 P XXII 306 The Fourth Table.-This Table has been formed for the purpose of exhibiting the results from which , in conformity with the principles explained by Sir JOHN HERSCHEL , the existence or non-existence of parallax in the stars observed must be inferred .
These results are the differences between two angles of position , the one observed at one period of the year , say the early one , and the other at another period ( the late ) distant from the former about six months .
For reasons which will appear hereafter , it has not been deemed necessary to include the distances in this table .
In so delicate an inquiry as that of parallax , it may not perhaps be considered a legitimate course to pursue , to compare any results except those distant from one another by an interval of about six months , or in other words , results which follow one another immediately at opposite seasons .
It is on differences so obtained that many therefore may be disposed to rely exclusively in discussing the questions raised by this investigation ; and such differences , when furnished by the observations , are accordingly placed first in the table ; but it is in truth only a choice of difficulties ; it has been found impossible in the case of some of the stars to obtain such differences at all ; and in others , the observations are so few , and their weight consequently so small , at one at least of the periods compared , that little reliance can be placed on the results ; it has been thought better therefore to combine in the case of each star , and in the manner already explained , all the observations made in all the years at the early period into one mean result , and all those made at the late period into another , and to insert their difference , leaving it to astronomers to attach whatever importance to the comparison they may think due to it .
This difference follows in every case the more legitimate comparisons at intervals of six months .
The four binary stars are excluded from this Table .
The second column of this fourth Table contains the mean epoch : in the case of thejinal means , that is , the means of all the early and all the late periods , the decimal , following an enumeration of several years , is the mean of all the decimals attached to the years of the individual means or sets , from which the final means are formed .
The third column contains the positions , obtained by using the computed weights of the results employed , as before explained .
The fourth andfifth columns contain the probable errorsand weights of these positions .
The sixth column contains the differences between the positions , which follow one another in succession , and which have been obtained at opposite periods of the year ; or , in other words , the quantity , on which parallax depends , and to which the subjoined remarks on the results of the observations refer .
The plus sign is attached to the difference when the change in the angle takes place in that direction , which is in conformity with the hypothesis of a parallax in the stars observed ; and the minus sign when the change is in the contrary direction .
2z 2 Results of the observations .
Distances .
In discussing the results of the observations , little need be said of the measures of distance : it was impracticable , consistently with the due prosecution of the more important task of measuring the angles , to take these measures of distance at shorter intervals , and at the epochs proper for eliminating the changes , if any , due to parallax : the small differences between the measures taken at opposite seasons must be ascribed to errors of observation .
The following Table exhibits these differences ; it contains the means of the measures of distance obtained at the early and late periods , with the sums of the assigned weights used in the reduction , and which means correspond , in respect to epoch , with thefinal means of the angles contained in the fourth Table ; the mean epochs themselves therefore have been already given in that Table .
The four binary stars are excluded from the following Table .
TABLE V. Name .
Mean Assigned Name .
Mean Assigned distance .
weight .
distance .
weight .
Arietis ... ... ... ... ... 8943 32 52 Com Berenices ... 3'917 697 8'838 36 3'914 8-4 32 Eridani ... ... ... ... 6881 21 136 178 ( Bode ) Libre ... 12-016 26-6 .
6'745 27'9 12-232 3-0 w Auriga ... ... ... ... ... 6-474 49-2 5 100 Herculis ... ... ... . .
14-064 42-1 6-309 39-1 13-983 115.5 118 Tauri ... ... ... ... ... 4915 347 .
579 ; ... ... ... ... ... ... 5-126 63-8 '07a '062 4'987 42-9 5188 6-0 41 Auriga ... ... ... .
7967 745 094 287 h and s ... ... ... ... 8148 226 7'873 20'5 8-486 3-4 Geminorum ... ... ... 7-461 32-7 Draconis ... ... ... ... .3101 124 152 ?
150 7'309 11-4 2-951 19'3 Anon .
Cancri ... ... ... 3'669 53-1.0 I of H 95 3*145 49'7 02 '-078 4023 3'591 11'8 3168 53-3 2 Cancri ... ... ... ... ... 4-851 59'6 7 4'674 113 Angles of position .
In the discussion of the measures of position , it must be remembered that four of the stars observed , viz. 39 Bootis , B Serpentis , c Draconis , and P XXII 306 , exhibit differences in their mean positions , as compared with those given by other observers , of an amount sufficient to induce the belief that they are binary systems : the observations of these stars may therefore be at once dismissed with the remark , that they are useless for the purposes of this inquiry .
Again , in the case of five other stars , viz. y Arietis , &lt; 2 Cancri , 178 ( Bode ) Librae , 100 Herculis and 579 Struve , the components are of equal magnitude , and parallax is in this case not only a priori highly improbable , but there is nothing in the actual observations themselves of these stars , which can lead us to any definite conclusion as to its existence in fact .
It is proper , however , to direct attention to the large differences in the case of p2 Cancri , accordant in respect of sign .
In order to arrive at any decision on the question , whether the changes in the angles of position of the remaining stars , exhibited in the sixth column of the fourth Table headed Difference , " should induce us to : suspect the existence of a sensible parallax in any of the objects observed , two points are to be considered , -first , the amount of those changes , as compared with the probable error of the results themselves ; and secondly , their direction , considered with reference to the motion of the earth in her orbit .
As to the first point , it would appear , that except the change observed bear a very large proportion to the probable error of the determination , it may be justly treated with little respect , so far as the evidence to be derived from amount only be concerned .
This conclusion is supported , first , by the great discordances between the partial differences , in the instances of the three stars , 118 Tauri , 100 Herculis , and I of H95 , the only objects that can be said to have been satisfactorily observed , if the delicate nature of the investigation be considered .
2ndly .
These discordances are the more remarkable , when we view them in connection with the large ratios which some of the differences bear to their probable errors .
Indeed a bare inspection of the measured angles of position given in Table I. in connexion with the computed weights and errors , will convince us that the differences of the results obtained on different evenings are greater in many instances than those of the separate measures obtained on the same evening .
This circumstance , which , it is believed , is not unfrequent in observations requiring the same delicacy of estimation as those under consideration , renders the application of the calculus of probabilities embarrassing and its conclusions uncertain .
It is certain , in fact , that in the greater number of instances given in this paper , the computed probable errors of the results on which the parallaxes depend , form no probable criterion of the magnitude or even of the existence of parallax .
The only probable explanation that can be given in such cases , where purely instrumental errors can scarcely have place , must arise from an unsuspected bias of the observer in the selection of the line of direction passingthrough the centres of the stars , which causes an estimation constant on the same evening , but differing on different nights .
The weight of each evening 's result , calculated according to the amount of discordance of the individual measures , may still be an adequate representative of the degree of dependence to be placed on it , as far as the steadiness of the atmosphere and other favourable circumstances are concerned ; but the probable errors of combined and final results , calculated on ordinary principles , will not form a good criterion of the accuracy of such results .
Supposing , however , the bias of the observer on different evenings likely to be as often in excess as in defect , or not to be of the nature of a constant error , a better estimation of the probable error of the final means would be obtained by considering each evening 's set of measures as a single result without regard to their number , and then comparing the separate results in the usual way* .
If the results in the present paper had been of a more positive character , a re-reduction of some of them on this principle might have been desirable , but as the case stands this explanation is sufficient .
It is proper to add , that I have no reason to believe , that a single individual measure , either in position or distance , however discordant , has been ever rejected on that or any other ground ; but all sets containing less than four measures have been invariably rejected .
It would seem , then , that though , in one point of view , comparisons at opposite seasons must be treated as possessing value , yet we are after all driven to rely in a great measure on the final means ; for notwithstanding the well-founded objections that may be urged to combining observations made in different years , the discordances above alluded to show , that a greater number of measures are required to eliminate error , than can possibly be obtained in any one , or even two years .
Now if we examine the final means of the two periods in Table IV .
, we shall find five stars only , viz. 32 Eridani , 41 Aurigae , 8 Geininorum , Anon .
Cancri , and I of H 95 , of which we can affirm that they have any pretensions whatever to be considered as fulfilling the condition above stated:-with respect to 8 Geminorum and Anon .
Cancri , the small weights of the results of the late period render any conclusions drawn from them extremely liable to doubt : this is the more to be regretted , in the case of the last-mentioned star , inasmuch as the smallness of the distance of its components renders it a very eligible object for observation .
When however we examine the results obtained with reference to the direction of the changes observed , they seem to be entitled to rather more consideration .
In the nf and np quadrants an apparent motion to the westward of the larger star , assumed to be the nearest to the earth , must increase the angle of position , and in the sfand qad ts the contrary effect must take place:-therefore at that epoch of observation when the sun is to the west , and therefore the earth to the east , of the star observed , the maximum angle may be expected to occur in the two first named quadrants , and the minimum in the others ; in the case of double stars , whose components are equal , it is plain that two different conclusions may be drawn from the periods at which the maxima occur , according as the one or the other be taken as the nearest star .
Omitting therefore the five stars , whose components are of equal magnitude , omitting also the four binary stars , and 2 Comae Berenices and s Draconis ( which exhibit differences less than their probable errors ) , this interesting result appears , that of the eight that remain , there is only one , that is , 41 Aurigae , the changes in whose angles , however small and little entitled to confidence they may be , do not conform in direction to those , which would take place , were a sensible parallax admitted in the brighter of the stars themselves .
This is most probably only an accidental coincidence , and I am very far from wishing to estimate it at more than its real worth ; but in the case of 32 Eridani and I of H 95 , where large , and to a certain extent trustworthy , differences concur with normal directions of motion , it may not perhaps be too much to assert , that this constitutes them objects of interest to the astronomer , possessed of adequate means to prosecute an inquiry , which , I fear , I must be said rather to have attempted than to have succeeded in .
Wrottesley , October 22 , 1850 .

