]\gt ; APhotographic of the Elements of the Orbits of Jupiter 's tellites .
By BItyAN COOKSON , M.A. , Mackinnon Student , July , July , 1906 .
( Communicated by H. F. Newall , F.R.S. eceived April April 30 , 1908 .
) the opposition of Jupiter in 1902 , the writer , by Sir David Gill 's kind permission , was engaged in making a series of neasurements of the relative positions of the Galileau tellites the -inch heliometer at the Cape .
Simultaneously with these visual observatious , taken with the raphic telescope .
This papel .
contains a short account of the work done in connectiolll ith the photographs ; a detailed account has appeared as vol. 1 part the 'Anuals of the Cape valory .
' The theory of the motions of Jupiter 's sateLites is one of most in in celestial Dlechanics , and in order to test the istence of some remarkable theoretical inequalities , it is necessary to possess an lccurate of the elements of the orbits .
At the present do not possess this and it was a view of its attainment that vriter uudertook these ations .
The eration with the heliometer was with the object of mass of Jupiter and correcting the best ilable elements of the orbits of the satellitcs , which observation in error .
The object in view in the raphs ) to } ) the visual and raphic methods and to see whether the nnch less ions and convenient methods of raphy would ield 1esulCs of same high order of accuracy as heliometer .
The unber of which were measured and discussed is 35 .
The measureznent of them was carried out at the ) in duplicate by Mr. Lolinger and writer , the micrometer one 0 those used for the of the raphic c all the reductions were made by the writer after his return to from the Cape .
Owing to the special nature of the , the usual method of determining the plate constants could not be .
The usual procedure is to determine the three constants of -centre , scalevalue , and orientation from measurements of three or more stars hose meridian places are known .
In the present case this method was impracticable because the brightness of the satellites made it necessary to cut down 568 Mr. B. Cookson .
A Photographic Deter nination [ Apr. 13 , the aperture of the object-glass ; with the reduced aperture none but bright stars could be photographed in the same field with the planet , and of these there were not enough .
But the same two standard stars as were used in the heliometer investigation were bright enough to leave measurable images on the ) late , and from them two of the constants , nameJy , scalevalue and orientation ( leferred to the earth 's mean equator of 1902 ) , were determined for every plate .
The orientation was also determined independently from measurements of the trail of one of the satellites .
It was important that the orientation should be determined with special care , for on it depend the values found for the inclinations and nodes of the orbits of the satellites .
The remaining constant , .
the plate centre , did not enter , for the quantities which were to be given by measurement were the differences between the co-ordinates of Jupiter 's centre and the centres of the satellites .
Thus the measured co-ordinate of a satellite eferred to Jupiter 's centre as origin and corrected for scale-value , orientation , and refraction , is its co-ordinate referred to axes parallel and perpendicular to a plane parallel to the earth 's mean equator of 1902 and passing Jupitel.'S centre .
The uncertainty in the measurement of the position of Jupiter 's centre , which arose from the ill-defined image of the planet 's disc on the photographic plate , was eliminated by a process of successive approxi- mation , which will be explained presently .
The theory with which the observations were compared was that of Souillart .
Marth 's Tables , which are based upon this theory and were published in the 'Monthly Notices Roy .
Astr .
Soc vol. , were used , with certain corrections , for calculating the values of the longitudes and radius vectors .
The calculated co-ordinates were rectangular co-ordinates referred to parallel and perpendicular to Jupiter 's equator , the origin being at the planet 's centre .
the choice of the equator of Jupiter instead of the earth 's as plane of reference , this part of the work was much simplified .
In order to simplify the numerical work still further , the axes of reference for the measured co-ordinates were rotated into parallelism with axes parallel to Jupiter 's equator .
Col.rections were sought to all the elements of th- orbits excepting the mean motions , which could only be determined by a series of observations extending over a much greater period of time than that covered by the present series .
Corrections were also to the coefficients of the ' Great Inequalities\ldquo ; in motions of the three inner satellites , inequalities which arise from the near commensurability of their mean motions .
Since the equations for the four satellites had no unknown quantity in common , the four sets of 1908 .
] of the Elements of the Orbits of Jupiter 's equations could be solved separately .
The values of the unknowns were found by the method of least squares , and were substituted in the original equations of condition .
If , now , an error had been made in the position of the centre of Jupiter 's disc on any plate , this error wou1 be common to the measurements of all four satellites on that plate , and appear as a residual common to the four equations derived from that plate .
Accordingly , the mean residual for every plate was found and was as a correction to the measured co-ordinates of the four satellites .
] ) corrected the equations were solved .
In this wsy a second approxinla- tion to the values of the unknowns was arrived at .
The mean residuals from this second approximation were now all so small as to have no tical nificance , and the second approximation to the values of the unknowns was taken as final .
The reement between these and those from the heliometer ation was satisfactory ; many of the did not differ by more than probable errors , and the are in general the same .
In point of accuracy , as measured by reement of the the raphic method is more accurate than the .
The } able error of a visual observation of weight unity was about ] ) that of a photographic observation , it be pointed out thati the measurements of the photographs ] a ( micrometelrned to attain the hest accuracy ; a measurement might have ) iven even better results .
On ) hand , for this iCular class of visual obseryations the heliometer is most refined instrument known , but , neyertheless , the ]able error of one observation of a large series has not yet been brought below Thus , in point of accuracy , the } raphic method is superior to heliometel .
But perhaps the heliometer is superior in its mole freedom from systematic error ; in measurements of absolute the heliometer is to be preferred chiefly on account of the of determinit ) the optical distortion of a photographic lens .
A knowledge of the mass of a planet attended by satellites ived at by observation of their times of reyolutio.n and of their meall from the planet .
It is necessary that observation should provide of the mean ances expressed in absolute ular measure , and this purpose the heliometer is the most suitable instrument , provide the ] are enough .
The mass of the system of Jupiter , in terms of the sun 's mass , was determined with great care from the heliometer observations .
value ] ] deduced is 1 : VOL. LXXX.\mdash ; A. 2 570 Mr. B. Cookson .
A Determination [ Apr. 13 , As to the methods of reduction , the photographic is much the laborious of the two , an adyantage which arises chiefly from the use of ular co-ordinates referred to conveniently chosen axes .
At some future date it will be necessary , in order to find the values of the secular variations of the elements , to reobserye the positions of the satellites , and can be little doubt that the best and most convenient method making the observations will be the phtraphic method .
One of the most remarkable of Laplace 's theoretical discoveries was that the well-known relation between the longitudes of the first three satellites oscillated about a mean position , that is , when expressed analytically , , where and are two constants of integration and can be regarded as a kllown function of the masses of the three satellites of the form This periodic ternl was called by Laplace the libration .
Once the masses are known , the period becomes known , but and must by observation .
Eclipse observations failed to show , and the reason ivell was that was too small to be detected by observation .
But the more probable reason is that , since the necessary eclipse observations extend over a considerable period of time , they must be combined by assuming the period to be known ; if the period is not known , or if a yalue is assigned to it , the coefficient camlot be found .
Since the period is a fmlction of the masses , the masses must be known before eclipse observations can be expected to reveal the value of the coefficient .
But values of the masses are still very doubtful : masses of satellites II and are fairly well known to within 4 or 5 per cent. , but tlJe masses of both I and are doubtful to the extent of 50 per cent. , and the period of the atio varies according to the values adopted from to years .
Hence , eclipse observations cannot be expected to disclose the value of 's libration .
The values of can , however , be determined with great by heliometric or raphic observations over only three or four months .
Hence the heliometer or raphic telescope may succeed where eclipse observations have failed .
The values of found by the writer are as follows:\mdash ; With heliometer at epoch 1901.61 heliometer at epoch 1902.60 With at epoch 1902.60 it is not yet safe to conclude that the whole of this is due to Laplace 's 1908 .
] of the Elements of the Orbits of Jupiter 's tellites .
libration : we must filsG be satisfied that the quantities do represent mean longitudes and that they include no elloneous or unknown inequalities in longitude .
This would entail a revision of the whole theory of the satellites .
) servation , then , is in advance of , and interesting questions cannot be settled until theory is reyised and further elaborated .
Heliometric and raphic nleasrcmenGs are of more value than photometric measurements of eclipses in deter1uining the positions of the satellites ' orbits .
The nodes and inclin ations of the orbits are determined by the duration of eclipses , and an observation of the Cion of an eclipse is liable to considerable error .
But with the heliometer or photographic telescope the position of the orbit is given by measurement of or its equivalent in nlar co-ordinates ; and it is clear measurements of the or rect ( co-ordinates of a satellite at its reatest elongation will provide reliable information concerning the position of the orbit than the duration of ) The quantity which it is of most interest to determine is the motion of the node of the second satellite measured ou Jupiter 's equator .
Hence , it is of importance to find observation the position of the orbit at epochs considerable inlCervls .
It is well known that the effect of an protruberance of the ary is to cause a rade motion of the of a satellite 's orbit the equator of the primary : the inclination ] nains constant , so that the pole of the satellite 's orbit round the pole of the of the .
The } .
table the percentage of the llotion of the llodesvhich is due to the compl'ession of Jupiter and bing e of the other satellites:\mdash ; Satellite .
Yearly motion .
iter .
1 .
II .
\mdash ; - I II 12 1II 2 IV 0.44 I II 12 1II 2 IV 0.44 I II 12 1II 2 IV 0.44 I II 12 1II 2 IV 0.44 I II 12 1II 2 IV 0.44 per cent. per I 97 II 82 2 IV 0 .
per oent . .
cent. per .
per \mdash ; 13 Thus in the case of the second satellite which has an ) at an inclination of to ) 's equator , the node annulu , and of this l1lotion 8 per cent. is due to the ] ) ressio of Jupiter , 4 ) the inence of satellite ) thab of to that of orbits of satellites I III have only small tions to the planet 's equator , and are , therefore , not suibal ) olio of the Hon- .
R. J. Strutt .
Helium [ Apr. 23 , nodes .
But the motion of the nods of II can be measured , and affords the best means of finding the compression of Jupiter .
The fifth satellite discovered by Barnard is so near to the primary that the node of its orbit revolves through about per annum , and second order terms begin to make themselves felt .
A careful measurement of this motion would be of much value .
for a comparison of the compression of Jupiter , deduced from the motion of the node of with that deduced from the motion of the node of II , might provide information concerning the distribution of mass in Jupiter .
Helium and xctivity in Rare Common Minerals .
the Hon. R. J. ( Received April ) , \mdash ; Read May 7 , 1908 .
) CONTENTS .
PAGE S1 .
Introduction 572 S 2 .
Experimental Methods S.3 .
Statement of Experimental Results 578 S4 .
Helium and Radio-activity of Elements 591 S 5 .
Presence of Argon and Neon in Minerals 592 S 6 .
Summary of Conclusions 694 S 1 .
Intro The original discovery of helium in cleveite and other minerals by Ramsay , and the subsequent explanation of its presence as due to production situ by radio-active change , are of fundamental importance , and are too well known to need anything than mention here .
No exception has been known hitherto to the rule that helium is found in radio-active minerals exclusively\mdash ; minerals , that is , in which either uranium , or thorium , or both , are present in appreciable quantity .
It seemed , however , that valuable information might be from a fuller examination of the subject than has yet been made .
In the first place , Butherford ested with great plausibihty that subatomic changes might be going on in some of the ordinary elements emission of particles of somewhat lower velocity than those of the radio-active elements .
In such cases , owing to the abrupt disappearance of and photographic action below the critical velocity , the activity

