]\gt ; [ 33 ] .
Magnetic Survey of the West of France , 1868 .
the Rev. STEPHEN J. PERRY , S.J. , F.R.A.S. , F.M.S. , .
Communicated by the President .
Received June 3 , \mdash ; Read June 17 , 1869 .
TIlE Magnetic Survey of the West of France , of which the following is a report , undertaken by the Rev. W. and myself at the request of the ities of Stonyhurst College , who generously undertook to defray all the expenses of the expedition .
A similar survey East of France will be made during the coming autumn , so as to complete the series of observations of the magnetic elements for the whole of France .
The instruments employed were those which have been in constant use at Stonyhurst Observatory for the determination .
the monthly absolute values of the Declination , and Intensity .
They consisted of a dip-circle by BARROW , No. 32 , a unifila and FRODSHAM'S marine chronometer , No. 3148 .
A beautiful transit-theodolite and an aneroid barometer were kindly placed at our disposal by the late .
COOKE .
The dip-circle was provided with three needles of the ordinary construction , the length of each being inches , and the maximum breadth of Nos. \amp ; inch , whilst that of No. 3 was inch .
For the unifilar thele wele five \mdash ; two for declination observations , a third , No. 7 , for vibratin and deflecting , and the remaining two , Nos. 9\amp ; 10 , for suspension in the deflection expeliments .
The same declination was used throughout the whole series of observations , and the only deflected employed was No. 9 .
The reduced obselvations are given at able 1 in the following , in order that the accuracy of the conclusions may be mole reliable , and the results be more easily compared with those of past and future observers .
The moment of inertia of the deflecting magnet , No. 7 , with its stirrup , for different degrees of temperatule , and the coefficients in the corrections required for the effects of temperature and of terrestrial magnetic induction on the magnetic moment of the magnet , were determined at the Kew Observatory by the late .
WELSH .
The moment of inertia of the magnet , with its stirrup , using the grain and foot as the units of mass and of linear measure , is .
Its rate of increase for increase of temperature is for every of FAHR. The weight of the magnet , with its stirrup , is approximately 825 grains , and the len gth of the magnet is nearly inches .
The moment of inertia was determined , independently of the weight and dimensions , by the method of vibration , with without a known , increase of the moment of inertia .
lfDCCCLXX .
The observations will , for clearness ' sake , be arranged under the successive heads of Dip , Intensity , and Declination .
The lilagnetic Dip .
The method of observing differed in nothing from that usually employed .
The magnetic prime vertical was first found by four readings of the needle when vertical , and the circle then set so that the needle might oscillate in the magnetic meridian .
Each end of the needle was read with the face of the circle first East and then West ; next , the axis of the needle was changed on the agate planes , and the same observations again taken ; and lastly , the whole was repeated after reversing the poles of the needle .
Each of readings was l.epeated after lifting the axis of the needle on its Is , and if a difference of 2 ' was found , fresh sets were observed .
The results are contained in the Table ( I. ) .
In the above Table the .
is arranged in the same ol.der as the observations .
The exact site of each separate observation has been carefully noted , but the details SURVEY OF THE WEST OF FRANCE .
are herein omitted for brevity 's sake .
The fullest inquiries have invariably been made as to the exact position of any underground iron pipes that might interfere with the vations .
In the following calculation of the most probable dip at each station , I have omitted the observations taken at Vannes with needle No. 3 , since these were made during heavy rain and under a lofty but covered ball court .
The readings of No. 3 at Amiens have also been excluded , since they were only taken as a velification of the results obtained with No. 1 , and were only single readings .
The latitudes and longitudes in Table II .
are taken from the ' Connaissance des Temps , ' and the differences between the latitude and longitude of Paris , and those of the other stations , have been calculated in geographical miles by aid of the Table in ' Practical Astronomy , ' and each verified by measurement on the geological mnp of MM .
DUFR et ELIE DE TABLE II .
For the latitude and longitude of the Magnetic Observatory at Abbadia I am indebted to the kindness of M. D'ABBADI ] , Membre de l'Institut , who rendered us every assistance during our stay in the South of France .
The pillar on which most of the obselyation at this station were taken stands at about 800 yards N. of the atory .
In the equations of condition , by which the most probable value of the dip at each station can be determined by the method of least squares , I have chosen Paris as the of coordinates for several reasons .
The chief of these arises from the fact that Paris is practically the centre of France , and thus obselvations could easily bc de there at the beginning and end of the Survey of 1868 , and also of 1869 .
Add to this that frequent obseryations have in past times been made there , and that the atafl under direction of M. L VERRIER keep up a continued series of determinations of the netic elements .
Lastly , the nature of the } uarantees a perfect frecdom the influence of igneous ocks , \amp ; c. THE .
STEPlfEN J. PERRY ON THE MAGNETIC Table III .
contains the equations of condition formed from the data in Tables I. , II .
In these equations dip at the central station diminished by ; and whel .
is the angle which the isoclinal lines make with the melidian , and is the increase in angle of the dip every change of a geographic mile in the direction normal to the isoclinal lines .
Solving these equations by the method of least squares , we obtain the following equations : \mdash ; These give Hence the most probable dip at Paris derived from the observations taken at the other stations is By substitution in the above equations we find the values of , and The isoclinal lines are therefore in the direction N. E. to S. W. , the distance between the lines representing the difference of SO ' in the dip will miles .
The substitution of these values of , and in our original equations will us the values of the computed dip at each station .
SURVEY OF THE WEST FRANCE .
' The most probable error for any single observation is then given by the formula , where , and ; Probable error The disturbing influence from geological causes may be of by Table , which is drawn up from the map of MM .
DUFR and ELIE DE BEAUMONT , Jngenieurs des Mines .
TABLE The error at Pau was most probably due to the fine dust that filled the air , the obscrvations having been taken not far from a building in the course of erection .
The amount of error at Angers and at Breast might also partly be attributed to the less favourable situation in which the instruments were used .
Most of the stations were , howevel , well adapted to netic observations , being on sedimentary rocks or later deposits , and equally free from all disturbing influences .
In constructing a map the isoclinal lines from the above values of and , it must be borne in mind that the assumptions of these lines .
straight , of their parallelism , or of a uniform rate of increase of the dip , are only first approximations to the truth so extensive a tract of country as that covered by the Survey .
The curvature of lines towards the north as we approach the west should be very considerable .
The secular variation of the dip has not been taken into account in the above THE REV .
STEPHBN J. PERRY ON Tffl MAGNETIC lations , since the time occupied in the survey was not long enough to cause any considerable change to take place during the interval .
The epoch for all the observations may therefore be taken as September 1st , 1868 .
Table .
contains a comparison between the above results , and those published by Dr. LAMONT in his ' Erdmagnetismus .
' TABLE Supposing the same rate of decrease to hold for the remaining stations , where LAMONT did not observe this magnetic element , we have for the 1st of Jam. 1869 the following results:\mdash ; Laval . . . . .
Breast 66 Vannes Abbadia 62 Pau . . . . . .
Perigueux .
63 Bourges 64 Amiens 66 The dip was observed on March 3rd , 1866 , by M. D'ABBADIE at his magnetic observatory , and found to be , which gives for the annual decrease .
Referring , now , to former determinations of the secular decrease of the magnetic dip Paris , we find , from a memoir by M. G. AIM , " " Sir le magnetisme terrestre\ldquo ; * , that the mean annual decrease between the years 1671 and 1754 was 1 ' 1754 and 1780 , , 1780 and 1806 , , From 1780 to 1830 the yearly diminution in the decrease amounted to , as we learn from a Table in General article in the Report of the British Association for 1838 .
Comparing , now , the results of the Dip Observations taken in 1858 and 1868 with the mean of those obtained by ARAGO , HUMBOLDT , and MATHIEU from 1825 to 1830 , we find the annual decrease of the dip at Paris to be for 1843 , whilst that for 1863-4 is , which shows that there is at present a gradual acceleration in the decrease of this element amounting to a'bout per annum .
Dr. LAMONT gives as the annual diminution for 1858 , which is somewhat smaller than the amount found above .
* Published in the Annals de Chimie et de Physique , 3e serie , .
xvii .
OF THE WEST OF FRANCE .
The Magnetic Intensity .
The method invariably adopted for determining the horizontal component of the earth 's magnetic force was that of vibrations and deflections .
The horizontal , vertical , and total forces are calculated to English measure , one foot , one second of mean solar time , and one grain being assumed as the units of space , of time , and of mass .
The vertical and total forces are obtained from the absolute measure of the horizontal force and the dip .
observed times of vibration are entered in the Table without correction .
The time of one vibration has been obtained from the mean of twelve determinations of the time of 100 or of 200 vibrations .
In deducing from the observed vibrations and deflections the product and ratio of the netic moment of the magnet and the earth 's horizontal magnetic .
intensity , the induction and temperature corrections have always been applied , and the obselved time of vibration has been corrected for the effect of torsion of the suspending thread .
The induction coefficient is The temperature corrections have always been obtained from the formula whele is the observed temperature , and FAHR. the adopted standard temperature .
The values of the coefficients and are respectively and The correction for error of graduation of the deflection bar at ] foot is and at foot It has been found necessary to apply the correction for the late of the chronometer at two stations only , .
at Laval and at Bordeaux , where the rate was respectively and ; at the other stations it was always less than In the calculations of the ratio , the third and subsequent terms of the selies have always been omitted .
The value of the constant was found to be The angular measure of one division of the scale in the vibration-appalatus was found to be The value of at is ; this was deduced by Mr. of Kew , observations made with three inertia cylindels .
No correction has been applied for semiarcs of vibration , which vele a ) small .
TIffl REV .
STEPHEN J. PERRY ON THE MAGNETIC TABLE The values of the total force and of its components , as also the changes in the magnetic moment of the deflecting magnet , are at once deduced from the above Table .
SURVEY OF THE WEST OF FRANCE .
Deflections not having been observed at Bourges , it was necessary to compute the value of , in order to eliminate it from the product obtained by the vibrations .
The mean value of during the survey was at epoch Sept. 1st , 1868 ; the mean for the remainder of 1868 was , therefore the variation per lnont , the value on Sept. 14 should have been .
This agrees well with the mean value of at Perigueux on the 12th and Paris on the 16th , Forming , now , our equations of condition from the above data , and solving them by the method of least squares , precisely in the same manner as in the case of the dip , we obtain the equations \amp ; c. , whele the total force diminished by 9 .
From these we obtain ; and this gives for the most probable computed value of the total force at Paris whereas the mean of the direct observations was ; difference By substitution we at once get the quantities ; , and Hence the direction of the isodynamic lines is N. E. to S. W. , and the total force changes for every 115 geographical miles along the normal to these lines .
We can now draw up a Table of the computed errors in the observations .
THE REV .
STEPHEN J. PERltY ON THE MAGNETIC TABLE IX .
The most probable error of any one observation will therefore be the sake of comparison with the results of former observations , and in particular of those of Dr. , we will determine the direction of the lines of equal intensity .
the component of the earth 's magnetism .
By a process identical with that employed for the Total Force , we now find that at central station the Horizontal Force , and , and ; or the direction sought is N. E. to S. W. ; the error in any single observation being now We will next place side by side the values of the Force for1858 and1868 .
TABLE X. Hence the yearly increase of the Horizontal Force in the West of France Dr. LAMONT gives as the value for 1858 , therefore we may conclude that there is a slight acceleration of about per annum in the increase of this element .
SURVEY OF THE WEST OF FRANCE .
Tloe Magnetic Declination .
The method adopted in the determination of this element was the following .
The azimuth of a fixed horizontal mark , situated at a considerable distance , was first l.ead with COOKE'S transit theodolite , and then a transit of both limbs of the sun taken with the same instrument , the time bein noted by chronometer .
The muth circle having been read , the theodolite was removed , and JONES'S unifilar placed on the same tripod-stand .
The leading of the fixed malk and of the collimator magnet completed the observation .
The torsion of the silk thread was removed entirely , as far as that was possible , before each observation .
Bl.est was the only station at which the above method was departed from , and there , on account of the confined space in which the observations wele taken , a distant mark could not be sighted .
It was therefore consideled that more acculate esults would be obtained by observing the sun 's transit by reflection the mirror of the unifilal , according to Dr. LLOYD'S method .
The chief reason for not using this method at other stations was that the line of collimation of the telescope was not perpendicular to the axis of the mirror , and the motion in azimuth of the axis was not sufficient to correct this error .
The correction to be applied to the azimuth leading of the sun on account of this position of the mirror is iven by the formula , where the constant being the altitudes of the sun at the times of obsersm sm vation .
The formula was used for determining the altitudes .
The determination of the meridian line , or of the azimuth of the sum , at each observation depends mainly for its correctness upon the of nd rate of the chronometer employed .
No pains wele therefore spared to pleyent all jolting of the instrument during the voyage , and no opportunity was missed of it with other chronometers .
The following Table is the lesult of these comparisons , made always with the greatest courtesy by those in at the several atories and depots de chronometres .
The chronometer is a lalge-size marine instrument by , No. ; it has been in constant use at Stonyhurst Obselvatory since the beginning of ISC3 , and its daily rat.e is found by the observation of clock stars on every favourable night .
Its mean daily rate previous to the joulney was , and afterwards it increased to The rate is found to vary somewhat during the yeal .
Mr. FRODSHAM kindly examined the instrument before it was taken to France and immediately after its return , and declared it to be in perfect order .
TABLE XI .
At Toulouse the chronometer was only compared as to rate by the Director of the bservatory , the absolute times of comparison not registeled .
Altitudes of the sun were taken at nearly all the stations with COOKE'S small altazinuuth , but these were only used as a check on the above results and for intermediate stations .
tfhese direct determinations of the rate of our instrument were exceedingly useful , from showing us that the two plincipal disturbances probably do not interfere in least with the results obtained , since the first took place between Poitiers and Bordeaux : and the second between Amiens and London .
In the following Table of obselvations the first line of azimuths at each station was read on the theodolite circle , and the second line on the unifilar .
The centre division of the was always made to coincide with the centre line of the telescope , except , where it was .
to the apparent left .
The zero of the scale of the collimatol ' magnet was , each scale-division being At Abbadia the azimuth of the mark .
the Biarritz Lighthouse , was accurately ermined by D'ABBADIE , President de la Societe bIeteolologique de SUBVBY OF THE WEST OF FRANCE .
TABLE XII .
The azimuth of the sun at the time of observation , calculated by the formula , where , gives the results in Table XIII .
TABLE XIII .
for scale-reading .
20 !
9 44 An application of the method of least squares , similar in all details to that made use of above , furnishes the equations \amp ; c. Hence ; per geographical mile , .
the rate of change in the declination along the normal to the isogonic lines very 8 geographical miles , and or the direction of the isogonic lines is N. E. to S. W. We might naturally expect that obselvations of this magnetic element would lead to much less satisfactory results than those of the Dip or Total Force ; for the almost continuous , and often energetic , action of disturbing forces render any limited number of absolute readings of the declination at different stations a very uncertain guide in calculatin the exact direction taken by the isogonic lines ; whereas in the case of the vertical and horizontal components of the total force the daily bances are of a much less exaggerated character , and the maximum perturbations of not frequent occurrence .
The subjoined Table will show the probable amount of error at each station .
TABLE The probable error of any single observation will therefore be .
We are now able to determine the secular variation of the declination .
TABLE SURVEY OF THE WEST OF FRANCE .
At Angers an iron pipe was lying N. and S. at a distance of about 11 yards from the place of observation .
Neglecting , therefore , this station , we find the yearly decrease of the Declination in the West of France to be M. G. , in the ' Memoir \ldquo ; already quoted , gives the mean annual.variation of the Declination at Paris for past years .
Epoch .
Declination .
Yearly variation .
1663 to 1767 to 1767 to 1785 19 16 to 22 1785 to 1805 22 to 22 1805 to 1817 22 5 to 22 1817 to 1825 22 19 to 22 this summary a step further we have 1825 to 1858 , to , yearly diminution to to , yeally diminution whence it appears that the declination is rapidly on the decrease , with a mean yeally acceleration in the decrease of The value of the yearly diminution , as given by Dr. LAMONT , is for 1858 , which shows a steadin in the variation of this magnetic element .
A glance at the maps which accompany this report will show at once the changes that ten years have produced in the position of the lines of equal declination , dip , and intensity .
The distance between the lines remains in all cases almost constant for the same element , and the amount is moreover identical for the isoclinal and isogonic lines , the values of being respectively and .
The angle moved through by the lines is more considerable in the cases of the isoclinal lines than for the others , and the direction of this motion is away from the astronomical meridian in the case of the isogonics and isoclinals , and scarcely perceptibly towards the astronomical meridian for the lines of equal horizontal force .
In the maps of LAMONT'S ' Erdmagnetismus ' the lines are curved , and the epochs the dip , horizontal force , and declination are severally August 1848 , June 1848 , and 1854 ; whereas I have taken January 1st , 1858 , as the common epoch , and drawn the lines straight , for the sake of comparison with the lines for 1868 , which are laid } without any modification from the calculated values of and .
The dotted lines , which belong to the survey of 1858 , can therefore only be looked on as first approximations to the results derived from LAMONT'S observations .
The broken lines are those obtained by the above calculations .
The numbers marked at each station are the means of the observed values , and they serve to show the degree of approximative correctness attained by the adopted method of reduction .
A comparison of these maps with those of the latest surveys of the British Islands manifests a striking resemblance between the isoclinal lines in France and England .
For in the Report by General SABINE for the British Association 1861 , we find that the direction of the isoclinals in England was \mdash ; 65o 5 ' in 1837 , and in 1860 ; therefore , supposing the rate of change to remain constant , we have as the value for 1868 , being the amount found for France in August and September of the same year .
The values of and , show , however , that the lines are closer packed in France than in England , and this is confirmed by the observations taken with needle No. 1 at Kew and at Stonyhurst immediately after our return from France ; for if the angles be computed on the assumption of the same data holding in England as in France , we shall have : Observed .
Computed .
Difference .
Kew Stonyhurst The isogonics of France are nearly in the same direction as those of Scotland , as seen in Dr. STEWARr'S Repolt of Mr. WELSH'S Survey ; but they are nearly twice as close in Scotland as in France , the values of being in 1868 for the latter country , and for the former in 1858 .
The greatest difference is in the isodynamics , their angle for England being and , whilst for France at the same epoch , 1868 , the values and The following is a complete Table of all the elements at the Epoch Jan. 1st , 1869 .
TABLE XVI .
In none of previous reductions have I ventured to introduce a correction for any magnetic disturbance that might have existed at the time of observation , or for diurnal range , since thel.e were no self-recording magnetographs in France by which these corSURVEY OF THE WEST OF rections could be accurately determined , and any correction founded on the supposition of the simultaneous similar action of the disturbing forces in England and in France might appear somewhat arbitrary .
That a correction , however , be applied with advantage is rendered more than probable by the results of the comparison of the Kew and Lisbon magnetograms , and by the great similarity between the daily range in England and in Italy which I remalked in some of the ence declination-curves sent by Signor DONATI .
It would seem , from the comparisons made by Dr. STEWART and Senhor CAPELLO , that the declination and horizontal-force disturbances at Kew and Lisbon are simultaneous , in the same direction , and in the propoltion of to 1 for the declination , to 1 for the horizontal force ; whilst , on the other hand , Dr. STEWARr remarks that there is " " very little likeness between the vertical-force curves We will therefore assume that the perturbations of the declination and horizontal are simultaneous in England and in France , and in the proportion of to 1 and to 1 respectively , and we will take the corrections from the rams obtained at Stonyhurst Observatory during the survey , since the Kew and Stonyhulst curves may be considered as almost identical .
From hourly measurements of the undisturbed portions of our declination-curves from Dec. 16th , 1868 to Jan. 16th , 1869 , we obtain as the mean leading of the ordinate for Jan. 1st , 1869 ; and taking the difference between this and the ordinate at the time of each observation , we obtain a number which , when divided by and multiplied into 28 ' , the coefficient of the declination magnetograms , gives us the correction in arc for reducing the French observations to their mean value on Jan. 1st , 1869 .
No correction for torsion of thread has been found necessary .
The results obtained from the corrected observations compared with the values found above , will best show what is gained by this correction .
From uncorrected obseIvations .
From corrected observations .
Error of Paris . . . .
Probable error of any observation These figures of themselves would scarcely justify us in giving much weight to this correction .
The horizontal-force magnetogram seems to offer a surer means of improving our results .
The standard ordinate has been obtained from hourly measurements of the curves on undisturbed days in Aug. and Sept. 1868 , the gradual loss of magnetism of the suspended magnet .
taken into account .
The coefficient of the culve is In this case the amount of difference between the mean observed and the computed values at each station is generally diminished by the correction ; but , owing to the ] ness of these differences and to the trace of the magnetograph having been lost at the AIDCCCLXX .
time of the observations at Angers and Bordeaux , where the differences were greatest , the total change.of the values is inconsiderable .
The following comparison will give an idea of the results:\mdash ; From uncorrected observations .
From corrected observations .
Error at Paris Mean probable error A similar correction might be applied to the Total Force and the Dip ; but since the simultaneous variations of the vertical force are less understood than those of the horizontal force or the declination , the attempt could scarcely be expected to lead to any satisfactory result in the present state of our knowledge .
APPENDIX .
The observations made with the same instruments at Loyola on Sept. 3 and 4 are not taken into account in the above reductions , since we ished to confine ourselves entirely to the West of ance .
It may , however , be of some use to subjoin a synopsis of the results obtained .
Sept. 3rd , G.M.T. Azimuth of sun .
Azimuth of mark read with the altazimuth .
Azimuth of magnet Azimuth of mark read with the unifilar .
Correction for zero-point of scale of collimator magnet The declination is not deduced from want of sufficiently reliable data as to latitude and longitude .
Sept. 3rd , , temp. Fahr. Observed deflection Distance between centres of magnets 1 , , , , , Fahr. Observed time of one vibration Hence Moment of magnet .
Horizontal force Vertical force . . . . .
Total force . . . .
Sept. 4 .
No. 1 Needle .
Observed dip MDCCC I , XX \amp ; 1 ' \mdash ; - 4 ' \mdash ; .
, blev

