XV .
Observations of the changeable Brightness of the Satellites of Jupiter , and of the Variation in their apparentMagnitudes ; with a Determination of the Time of their rotatory Motions on their Axes .
To which is added , a Measure of the Diameter of the Second Satellite , and an Estimate of the comparative Size of all the Four .
By William Herschel , LL. D. F. R. S. Read June i , 1797 .
IT may be easily supposed when I made observations on the brightness of the 5th satellite of Saturn , by way of determining its rotation upon its axis , and found that these observations proved successful , that I should also turn my thoughts to the rest of the satellites , not only of Saturn , but likewise of Jupiter , and of the Georgian planet .
Accordingly I have from time to time , when other pursuits would permit , attended to every circumstance that could forward the discovery of the rotation of the secondary planets ; especially as there did not seem to lie much difficulty in the way .
For since I have determined , by observation , that the 5th satellite of Saturn is in its rotation subject to the same law that our moon obeys , it seems to be natural to conclude that all the secondary planets , or satellites , may probably stand in the same predicament with the two I have mentioned ; consequently a few observations that coincide with this proposed theory , will go a good way towards a confirmation of it .
I had another point in view when I made the observations which are contained in this paper .
It was an attempt to avail myself of the abundant light and high powers of my various telescopes , to examine the nature and construction of the bodies of the satellites themselves , and of their real magnitudes .
Here phenomenon occurred that will perhaps be thought to be remarkable , and even inconsistent or contradictory .
So far from attempting to lessen the force of such animadversions , I shall be the first to point out difficulties , in order that future observations may be made to resolve them .
Perhaps it would have been better to delay the communiication of these observations , till I had continued them long enough to be able to account for things which at present must be left doubtful .
But as in final conclusions to be drawn from astronomical observations , we ought to take care not to be precipitate ; so on the other hand I am perhaps too scrupulous in satisfying myself , and should probably require the observations of several years before I could venture to be decisive .
It will also be seen by the dates of the first observations , that a further delay in the communication cannot be adviseable ; since much information may possibly be gained by throwing open , to other observers , the road it will be eligible to take for a satisfactory investigation of the subject ; especially as we have reason to congratulate ourselves on the spirit of observation , and increase of large instruments , that seem to have taken place in various parts of Europe .
I shall now transcribe the observations from my journals .
They are as follows .
OBSERVATIONS .
A remarkable Conjunction of two Satellites of 7upiter .
May 14 , 1790 .
1h 30 ' la ; correct sidereal time .
The 2d and 3d satellites of Jupiter are so closely in conjunction , that with a 7-feet reflector , charged with a magnifying power of.350 , I cannot see a division between them .
11h 34 ' lo " .
The shadow of the 1st satellite is still upon , the disc of the planet .
Intenseness of Light and Colour of the Satellites .
July 19 , 1794 .
17h 12 ' 47 " .
7-feet reflector .
The ist satellite of Jupiter is of a very intense bright , white , shining light .
It is brighter than the 2d or 4th .
I speak only of the light , and not of the size .
The colour of the 4th .
satellite is inclining to red .
In brightness it is very nearly , but not quite equal to the 2d .
I make no allowance for its being farther from the bright disc of Jupiter than the 2d .
lo-feet reflector , power 170 .
The 3d satellite is just gone upon the body ; before it went on , it appeared to me to be smaller than usual .
The 2d satellite is of a dull , ash-colour ; not in the extreme , but rather inclining to that tint .
July 21 , 1794 .
16h 56 ' 45 " .
io-feet reflector power 170 .
rfThe 3d satellite of Jupiter is round , large , and well defined .
It is very bright , and ' its light is very white .
The 4th satellite is also round , large , and well defined .
I estimate its magnitude in proportion to that of the 3d satellite to be as 4 to * .
Its light is not white , but inclined to orange .
Brightness and Diameter distinguished .
July 26 , 1794 .
17h 14 ' 41 " .
lo-feet reflector ; power 170 .
The 4th satellite is very dim .
It is of a pale , dusky , reddish colour .
The 2d satellite is of a bright , white colour .
The 8d satellite is very bright , and white .
The ist satellite is very brilliant , and white .
17h 22 ' 41 " The Magnitudes with 240 .
The 3d satellite is the largest .
The 2d satellite is the smallest .
With 3oo .
The 4th satellite is a very little larger than the 2d , though less bright .
The ist satellite is larger than either the 4th or 2d .
With 400 , the order of the magnitudes is 3X 2 .
With the same power , the order of the light is gi2 4 .
Now and then it appeared to me doubtful whether the 4th satellite was larger than the 2d ; and as their light is of an unequal intensity , it is difficult without much attention , to be decisive about the magnitudes .
Diameter of the second Satellite by entering oil the Disc of thei Planet .
July 28 , 1794 17h 25 ' 4o " .
to-feet reflector ; power 170 .
The 2d satellite is nearly incontact with the following limbof Jupiter .
17h 29 ' 40 " .
It seems to be very near the contact .
With 0oo , very near the contact .
17h 30 ' 40 " It seems to be in contact It is brighter than that part of Jupiter where it enters .
17h 3I 40 " .
It is more than half entered .
17h 33 40 " .
It seems to be nearly qauite , entered .
Its superior brightness makes it seem protuberant .
17h 3 ' 40 " It is certainly quite entered .
17h 35 ' 25 " .
I see a little of the disc of Jupiter on the outside of the satellite , equal to about X of its diameter .
17 h39 ' 40 " .
The 3d satellite is very bright , and of its usual colour .
The 4th satellite is faint , and also of its usual colour .
The ist satellite is very bright , and the , light of it is of its usual intenseness .
The Magnitudes with 300 .
The diameter of the 4th seems to beto that of the 3d , as 2 to 3 ; or perhaps more exactly , as 3 to 5 .
The diameter of the 4th satellite exceeds that of the ist a very little .
With 400 .
With this power the diameter of the 4th satellite certainly exceeds that of the ist .
The diameter of the 4th , is to that of the 3d , as 3 to 5 .
July 30 , 1794 .
1ch 1 ' 37 " .
1o-feet reflector ; power 300 .
The 4th satellite of Jupiter is a little larger than the ist .
It is of its usual colour .
The 2d is less than the ist .
The 3d is larger than the 4th .
July 31o , 1794 .
17h i8 ' 38 " .
lo-feet reflector ; power 170 .
The four satellites of Jupiter are very favourably placed for my purpose .
The ist is less bright than the 2d ; it is a very little larger than the 2d : the difference in the size is but barely visible .
The light of the 2d is very intense and white .
The light of the 3d is very intense and bright .
The light of the 4th is dull ; and seems to be inferior to the usual proportion it bears to the other satellites .
18h 38 ' 38 " .
With 300 .
The 4th satellite is larger than the ist .
The 2d satellite is a little larger than the 1st , or at least equal to it .
The 3d is undoubtedly the largest .
The order of the magnitudes therefore is , 342 1 .
My Brother , ALEXANDER HERSCHEL , looked at the satellites , and estimated the order of their magnitudes exactly the same ; though he was not present when I made the foregoing estinations .
August i , 1794 .
17h 38 ' 37 " .
io-feet reflector ; power 170 .
The light of all the four satellites is very brilliant , the evening being very fine .
With 300 .
The northmost and farthest of the two satellites which are in conjunction , is the smallest : I suppose it to be the 2d .
The southmost and nearest of the two satellites in conjunction , is the next in size : I suppose it to be the ist .
The 4th satellite is a little larger than the largest of the two satellites which are in conjunction ; but the difference is only visible with a great deal of attention .
The 3d satellite is much larger than the 4th .
August 9 , 1794 .
17h 56 ' 32 " .
lo-feet reflector ; power 1i70o The light of the ist satellite is very intense and white .
The light of the 2d satellite is also pretty .
intense and white .
The light of the ld satellite is neither so intense nor so .
white as that of the ist .
The light of the 4th is dull and of a ruddy tinge .
With 300 , and 400 , the second is the least , and the 3d is thelargest .
I am in doubt whether the 4th or the 1fst is largest ; with 6oo , I suppose the 1st to be larger than the 4th .
September 30 , 1795 .
20h 1 17 " .
7-feet reflector ; power 210 .
Order of the magnitudes of the satellites of Jupiter 3 2 .
1 , Power iio .
2,1.4 .
With 460 , 3 2 , I , 4.$ October 2 , 1795 .
20h i 22 " .
7-feet reflector ; power 287 .
Jupiter 's satellites 32 1 , 4 The 2d and ld satellites are not yet in conjunction .
20h 4EX ' 22 " .
The conjunction between the 3d and 2d satellites is past .
The distance between them is now one diameter of the 3d .
August 18 , 1796 .
i8h 47 ' 21 " .
7-feet reflector ; power 287 .
The 4th satellite is less bright than the ist ; notwithstanding * Here , in order to denote the different magnitudes of the satellites , I used the notation which has been explained in my First Catalogue of the comparative Brightness of the Stars .
See Phil. Trans. for the year 1796 , Part I. page iS5 .
the latter is so near the planet as to have its light overpowered by Jupiter , while the 4th is at a great distance .
I mean light or brightness , not magnitude .
The ist is very bright .
September 15 , 1796 .
19h 25 ' 21 " .
o1-feet reflector ; power 3oo .
The 2d satellite of Jupiter is a little less than the 1st .
The 3d is much larger than any of the rest .
Power 6oo .
The difference in the magnitude of the 1st and 2d satellites , with this power , is pretty considerable .
September 21,1796 .
19h 24'5 " ia-feet reflector ; power 6oo .
The shadow of the 1st satellite is upon one of the dark belts of Jupiter .
In order to use very high powers with this telescope , I tried it upon the double star CAquarii with 1200 .
The air is very tremulous , but I see now and then the two stars of this double star very well defined .
With the same power , the satellites of Jupiter are very large , but not so well defined as the above star .
The Brightness of the Satellites compared to the Belts and Disc of the Planet .
The 1st satellite , which is lately come off the southern belt , is nearly of the same brightness with that belt ; power 6oo .
With 4oo , it is nearly as bright as the brighter part of the planet , or rather a mean between the belt and the planet .
The 2d satellite is considerably bright ; its colour is whiter than that of the 1st ; it is however not so white as the colouit of the bright part of Jupiter .
The colour of the 4th satellite is as dingy as : that of the belt ; very much less bright and less white than that of the 2d .
The brightness of the 3d satellite is not intense ; its colour , however , is white , though not so white as the bright part of the planet .
Septemnber 24 , 1796 .
20h 5 ' 24 " .
lo-feet reflector ; power 6o~o .
The ist satellite of Jupiter is very bright , and of a white colour ; it is also very large .
The 2d satellite is faint and bluish ; its light is not much brighter than that of the belt .
The gd satellite is pretty bright ; its light is whitish .
It seems to be comparatively less than it ought to be ; or rather , -its apparent smallness is owing to the uncommon largeness of the 1st .
The ist satellite , with 200 , colmpared to the 3d , is proportionally larger than I have seen it before .
September 30 , 1796 .
2oh 8 ' 4"t .
o-feet reflector ; power 6oo .
The satellites of Jupiter are well defined , and the night is beautiful .
The 3d satellite , in proportion to the ist , is much larger than it was September 24 .
1 ascribe the change to an apparent diminUtiOf of the ist .
20h 30 ' 4 " .
The 1st satellite is evidently less in proportion to the 3d , than it was September 24 .
The 2d satellite is considerably bright ; its light is whitish ; much brighter than the belt , but not so bright as the bright part of the disc .
Its magnitude is less than that of the P4th ; but its light is considerably superior .
The 3d satellite is remarkably well defined .
Its light is considerably brighter than that of the belts .
The magnituide of the 1st satellite exceeds that of the 2d .
It is -nearly equal to that of the 4th .
52h 8 ' 4 " .
Appearances as before .
October15,1796 .
2h3 ' 42 " .
to-feet reflector ; power 6oo .
The 2d satellite is uncommonly bright ; its apparent magnitude is also larger than usual .
The 4th satellite is very faint ; it is not brighter than the belt , but is of a bluish , ruddy colour .
The apparent magnitude of the 2d satellite , after long looking , is very nearly equal to that of the ist ; but at first sight it seems to be larger , owing to its superior brightness .
The apparent diameter of the 2d satellite is certainly larger than that of the 4th , 23h h5 42 " .
The light of the 1st satellite , compared to that of the 2d , is considerably increased since the last observation .
It is now nearly as bright as the A24 October i6 , 1796 .
oh 23 ' 49 " .
io-feet reflector ; power 6oo .
The ist , 2d , and 3d satellites of Jupiter seem all considerably bright .
rfhe 3d is much larger than the ist , and the ist a little larger than the 2d .
The intensity of the light seems to be pretty equal in all the three ; that of the 2d , however , is perhaps a little stronger than that of the ist ; for , notwithstanding its apparent less diameter , it seems to make as strong an impression as the ist , October 25 , 1796 .
21h i 48 " .
lo-feet reflector ; power 6oo .
The ist satellite of Jupiter , compared to the 3d , is small .
The 3d satellite is bright and large , The 2d is brighter than the ist .
Compared to its usual brightness and magnitude , it is very bright and small .
The ist satellite , compared to its usual brightness and magnitude , is faint and small .
The air is so tremulous that the power of 60o is too high , and the necessary uniformity required in these observations will not permit a lower to be used .
Perhaps one of 400 might be more generally employed ; and it may be proper to use it constantly .
Novembers3 , I796 .
23h S ' 47 " .
ia-feet reflector , power 6ao .
The 4th satellite of Jupiter is large and bright .
The 3d satellite is large and bright .
The 1st satellite is pretty small , and not very bright .
The 2d satellite is small , and considerably bright .
The brightness and magnitude of each satellite refer to its own usual brightness and magnitude .
Before we can proceed to draw any conclusions from these observations , we ought to take notice of many causes of deception , and of various difficulties that attend the investigation of the brightness of the satellites .
The difference in the state of the atmosphere between two nights of observation , cannot influence much our estimation of the brightness of a satellite , provided we adopt the method of comparative estimations .
If we endeavour by much practice to fix in our mind a general ideal standard of the brightness of each satellite , we shall find the state of the atmosphere in different nights very much disposed to deceive us ; but if we learn to acquire a readiness of judging of the comparative brightness of each satellite with respect to the other three , we may arrive at much more precision , since the different disposition of the air will nearly affect all the satellites alike .
But here , as we get rid of one cause of deception , we fall under the penalty of another .
The situation of those very satellites to which we are to refer the light of the satellite under estimation , being changeable , permits us no -longer to trust to their standard , without a full scrutiny of the causes that may have producedan alteration in them .
In the foregoing observations it will also be seen , that I attempted to compare the intenseness of the light of the satellites with the different brightness of the disc of Jupiter ; but these endeavours will always fail , on account of the little assurance we can have that the parts of the disc , setting aside its quick rotation , will remain for any time of the same lustre .
A very material difficulty arises from the magnifying power we use in our estimations .
If it be a low one , such as for instance iSo ( for a lower should not even be attempted ) , then we run the risk of being disappointed , in bright nights by the sparkling of the brilliant light of the satellites .
Besides , we cannot then see the bodies of them , and judge of their comparative magnitude , with the same power that we view their light .
If we choose a high magnifier , we shall be often disappointed in the state of the atmosphere , which will of course occasion an interruption in the series of our observation , of which the regular continuance is of the greatest consequence .
If we change our power according to the state of the atmosphere , we introduce a far worse cause of confusion ; for it will be next to impossible to acquire , for each magnifying power , an ideal standard , of comparative brightness to which we can trust with confidence .
If the magnitudes are not attended to , and carefully contradistinguished from the intenseness of light , we shall run into considerable error , by saying that a satellite is large , when we mean to express that it is bright .
It is so common to call stars that are less bright than others , small , that we must be careful to avoid such ambiguities , when the condition of the satellites is under investigation .
Nor is it possible to throw the size and light into one general idea , and take the first coup d'oeil in looking at them , to decide about the general impression this compound may make .
When our attention is forcibly drawn by a considerable power to the apparent size of the satellite we are looking at , its brightness can no longer be taken in that general way , but must be abstracted from size .
Let us now see what use may be drawer from the observations I have given .
It appears in the first place very obviously , that considerable changes take place in the brightness of the satellites .
This is no more than might be expected .
A variegated globe , whether terraqueous like the earth , or containing regions of soil of an unequal tint , like that side of the moon which is under our inspection , cannot , in its rotation , present us with always the same quantity of light reflected from its surface .
In the next place , the same observations point out what we could hardly expect to have met with ; namely , a considerable change in the apparent magnitude of the satellites .
Each of them having been at different times the standard to which another was referred , we cannot refuse to admit a change so well established , singular as it may appear .
The first of these inferences proves that the satellites have a rotatory motion upon their axes , of the same duration with their periodical revolutions about the primary planet .
The second either shows that the bodies ' of the satellites , are not spherical , but of such forms as they have assumed by their quick periodical and slow contemporary , rotatory motions , and which forms in future may become a subject for mathematical investigation ; or it may denote , in case geometrical researches should not countenance a sufficient deviation from the spherical form , that some part of the discs of these satellites reflects hardly any light , and therefore in certain situations of the satellite makes it appear of a smaller magnitude than in others .
1ere then we see evidently that a considerable field for speculation , as well as observation , is opened to our view ; and almost every attempt to enter upon the work must seem premature , for want of more extended observations .
However , from those that have been given , such as they are , I will show how far we may be authorized to say , that the satellites revolve on their axes in the same time that they perform a periodical revolution about the planet .
I shall take the usual method of throwing the observations of each satellite into a graduated circle .
The zero of the degrees into which I suppose it divided , is in all observations assumed to be in the place of the geocentric opposition .
In order to bring these observations to the circle , the places of the satellites have been calculated from my own tables of the mean motion in degrees , and according to epochs continually assumed from the geocentric conjunctions pointed out in the configurations of the Nautical Almanac ; and the nearest of these conjunctions have been always used .
This method is fully sufficient for the purpose , as greater precision in the calculation is not required .
The observations extend from July 19 , 1794 , to November 3 , 1796 ; and therefore include a period which takes in 470 rotations of the ist satellite ; 234 of the 2d ; 1 i6 of the 3d ; and so of the 4th : that is , provided we admit that these rotations are performed in the same time the satellites revolve in their orbits .
In the following table are the calculated places of the satellites ; the correct sidereal times , given with the observations , having been turned into mean time .
Table of the Positions of the four Satellites of Jupiter at the Time of the Observations .
I IIIIIV : Time of Observe .
II }[ I III I IV 1794 .
0 1796 .
I0 01 010 July 19d b 2I1 1271346|1791 466 Aug. i8 ' $h 21b MI I1||1 9~~~ 17 + 19 JUly 21 .
8 .
57 ?
1 27819 Sept. I S. 7 .
44.1 36 1 328 11i981 July 26 .
8 .
56 | I24 |333 1i69 205 Sept. 21 .
7 .
19 1072 I 2i4 | 138 1 2i0 July z8 .
8 .
59 i 171 I76 270 |248 SePt .
24 .
8 .
38 7 74 1 K631305 1 275 JulY 30 .
10 .
27 f 231 1 IYI 259| I.-29z Sept. 30 .
7 .
27 | 206 1 46 244 1 36 July 31 .
8 40 5O 7 .
44 6 .
28 1 130._ S Aug. i. 8 .
56 265 221 I"I | 334 Oct. IO .
I1 41 1 Aug. 9 8 .
421 83 1310 IJ2 138 Oct.i i6 .
J0 .
39 1 256 1 243 1 334 1 I795 .
Oct.25 .
7 .
zq 26i 72 1 59 l Sept. 30 .
7 .
37 1 294 | 6Z 2 2I9 o100 Nov. 3 .
9 .
" 306 1 270 | 1 ; I| 58 Oct. 2 .
7 32 1341 1 2641 31 91 143 |_ It will be necessary now to explain in what manner , with the assistance of this table , the observations of the brightness and magnitudes of the satellites have been reduced to the expressions they bear in the four circles of the figures contained in Tab. VIII .
and IX .
By way of uniformity I judged it would be best to reduce the estimations of magnitude to those of brightness ; as it may be justly supposed that when a satellite is at any given time larger in proportion to another than it was at another time , it will also be brighter than it was at that other time , due regard being had to the light of the satellite to which its magnitude has been compared .
To manage the space allotted to the figure advantageously , I have used the abbreviations formerly employed in my catalogue of Nebulae , v B , c B , B , p B , p F , F , c F , v F , for all the gradations of light that are necessary to express the brightness of the satellites at the time of obe servation .
It will be easily remembered that B and F mean bright and faint ; and p , c , v , stand for pretty , considerably , and very .
Now , when the observation mentions the brightness of the satellite , I place it in the figure as it is given .
In that of the first , for instance , July 19 , 1794 , we find the satellite called very bright ; I therefore put down in fig. 1 .
( Tab. VIII .
) at 127 degrees , v B. But where the brightnessis not expressed , I have recourse to the comparative magnitude , if that can be had .
By fig. i. ( Tab. IX .
) it appears that the 2d satellite is less subject to a change of brightness than either the ist or 4th : it becomes , for that reason , a pretty good standard for the light of these other satellites .
Therefore , in the observation of October 2 , 1795 , for instance , where the 1 st satellite is described as undoubtedly less than the 2d , I put down very faint , or v F , at 34p degrees of the circle in fig. i. ; for in the observation of July 19 , before mentioned , when the satellite was called very bright , it was at the same time described as undoubtedly larger than the 2d .
In this case , as regard must be had to the relative state of the satellite we refer to , the four figures I have given will assist us in determining the condition of the light of the satellite we wish to admit as a standard .
In reducing the 2d satellite to the circle , I have generally used a reference to the magnitude of the ist , where marks of '.eDCXCV1L Z z brightness were wanting ; and sometimes also to the magnitude of the 4th , and even of the 3d .
The 3d satellite can hardly be ever compared to any but the 2d in magnitudeand this only in its degree of excess .
The magnitude of the 4Wth satellite has been generally compared with that of the 1st : and also sometimes with that of the 2d .
To make an application of the contents of the figures , will now require little more than a bare inspection of them .
The ist satellite appears evidently to have a rotation upon its axis that agrees with its revolution in its orbit .
It cannot be supposed that , in the course of 47o revolutions , all the bright observations could have ranged themselves in one half of the orbit , while the faint ones were withdrawn to the other .
The satellite appears in the middle of the duration of its brightness , when it is nearly half way between its greatest eastern elongation , in the nearest part of its orbit ; or when advancing towards its conjunction .
I have pointed out this circumstance by a division with dotted lines , and the words bright and faint , inserted within the circle , fig. i. This satellite , therefore , revolves on its axis in 1d 18h 26',6 .
The 2d satellite , though much less subject to change , on account as we may suppose , of having only a small region on its body which reflects less light than the rest ; has , nevertheless , its rotation directed by the same law with the ist .
It will hardly be necessary to take notice of a single deviation which occurs at 163 degrees , fig. 2 .
; as from the proximity of the satellite to the conjunction , a mistake in the estimation may easily take place .
I generally made it a rule ; not to make allowance for the influence of the superior light of the planet ; but it seems that we can hardly abstract sufficiently on such occasions .
Two similar cases occur , in fig. 3 .
at 179 ; and fig. 4 .
at 5 degrees .
It is indeed not impossible but that occasional changes , on the bodies of the satellites themselves , may occasion some temporary irregularity of their apparent brightness : it will , , however , not be necessary to make such an hypothesis , till we have better authority for it .
The brightest side of this satellite is turned towards us when it is between the greatest eastern elongation and the conjunction .
It revolves , consequently , on its axis in 3d 18h 17',9 .
The 3d satellite suffers but little diminution of its brightness , and is in full lustre at the time of both its elongations .
It is however not impossible but that , after having recovered its light , on the return from the opposition , it may suffer a second defalcation of it in the nearest quadrant about half way towards the conjunction .
The two independent observations at ii and 152 degrees , fig. 3 .
seem to give some support to this surmise .
It revolves on its axis in 7d 3}1 59',6 .
The 4th satellite presents us with a few bright views when it is going to its opposition , and on its return towards the greatest eastern elongation ; but otherwise -it is generally overcast .
Its colour also is considerably different from that of the other three ; and it revolves on its axis in 16d 18h ' 1 .
It will not be amiss to gather into one view , all the obser-v vations that relate to the colour of the satellites .
The ist is white ; but sometimes more intensely so than at others .
The 2d is white , bluish , and ash-coloured .
The gd is always white ; but the colour is of different intensity , in different situations .
The 4th is dusky , dingy , inclining to orange , reddish and ruddy at different times ; and these tints may induce us to surmise that this satellite has a considerable atmosphere .
I shall conclude this paper with a result of the observation of the diameter of the second satellite , taken by its entrance upon the disc of the planet , July 28 , 1794 , and marked in fig. 2 .
at 176 degrees .
The duration by the observation is fixed at 4 minutes ; in which time it passes over an arch in its orbit of i6 ' 52",9 .
Now as its distance from the planet is to its distance from the earth , so is 16 ' 52",9 to the diameter of the satellites ; or the mean distance of the 2d satellite may be rated , with M. DE LA LAND , at 2 ' 57 " , or 177 " .
Then putting this equal to radius , we shall have the following analogy .
Radius is to 177 " , as the tangent of i6'52",9 is to the angle , in seconds , which the diameter of the second satellite subtends when seen from the earth .
And by calculation , this comes out o",87 ; that is less than nine-tenths of a second .
I have not been scrupulously accurate in this calculation , as the real distance of Jupiter at the time of observation should have been computed , whereas I have contented myself with the mean distance .
Nor am I very confident that the angle of the greatest elongation , admitted to be 2 ' 57 " , is quite accurate ; but I judged it unnecessary to be more particular , because the time of my observation in the beginning of the transit upon the disc , I find was only taken down in whole minutes of the clock .
The end , however , is more accurately determined , by the observation which was made 4-5 " after the immersion ; when a part of the disc , equal to about 4 of the diameter of the satellite , is said to be visible .
It seems that observations of this kind , made with very good telescopes , charged with high powers , are capable of great precision .
For the remark that a margin of Jupiter , equal to about 4 of the diameter of the satellite , became visible in 45 " of time , adds great support to the accuracy of the observation of the foregoing 4 minutes : and , at all events , it is evidently proved , from the whole of the entrance upon the disc , that the diameter of this satellite is less , by one half at least , than what from the result of the measures .
of former observers it has been supposed to be .
A method has also been used , of deducing the diameter of the satellites from the time they employ to immerge into the shadow of the planet ; but this must be very fallacious , and ought not to be used .
I should not pass unnoticed the apparent magnitude of the satellites .
The expressions that have been given of them may be collected into the following narrow compass .
1,4,2 41 3-,4 ; 12 4 , 2,1 3--4 ; 1 ; 2 1,4,2 3-2,1,4 3-- !
2-I,4 12 4.1-2 1 2-4 3-1 , 22 -l1 From which we may conclude , that the 3d satellite is considerably larger than any of the rest ; that the ist is a little larger than the 2d , and nearly of the size of the 4th ; and that the 2d is a little smaller tha-ni the ist and 4th , or the smallest of them all .
WM .
HERSCHEL .
Slough , near Windsor , April 30 , 1797 .

