259 1871 .
] Altitude Determination of a Ship 's Place .
easy , by a perfectly methodical action involving very little labour , to keep the battery in good and constant action , according to the circumstances of each case .
When , as in laboratory work , or in arrangements for lecture-illustrations , there may be long intervals of time during which the battery is not used , it will be convenient to cease adding sulphate of copper when there is no immediate prospect of action being required , and to cease pouring in water when little or no colour of sulphate of copper is seen in the solution below .
The battery is then in a state in which it may be left untouched for months or years .
All that will be necessary to set it in action again will be to fill it up with water to replace what has evaporated in the interval , and stir the liquid in the upper part of the jar slightly , until the upper specific-gravity bead is floated to near the top by sulphate of zinc , and then to place a measured amount of sulphate of copper in the funnel at the top of the charging-tube .
VI .
a On the Determination of a Ship 's Place from Observations of Altitude .
" By Sir William Thomson .
Received Feb. 6 , 1871 .
The ingenious and excellent idea of calculating the longitude from two different assumed latitudes with one altitude , marking off on a chart the points thus found , drawing a line through them , and concluding that the ship was somewhere on that line at the time of the observation , is due to Captain T. H. Sumner * .
It is now well known to practical navigators .
It is described in good books on navigation , as , for instance , Raper 's ( SS 1009-1014 ) .
Were it not for the additional trouble of calculating a second triangle , this method ought to be universally used , instead of the ordinary practice of calculating a single position , with the most probable latitude taken as if it were the true latitude .
I believe , however , that even when in a channel , or off a coast trending north-east and south-west , or northwest and south-east , where Sumner 's method is obviously of great practical value , some navigators do not take advantage of it ; although no doubt the most skilful use it habitually in all circumstances in which it is advantageous .
I learned it first in 1858 , from Captain Moriarty , R.N. , on board H.M.S. ' Agamemnon/ He used it regularly in the Atlantic Telegraph expeditions of that year and of 1865 and 1866 , not merely at the more critical times , but in connexion with each day 's sights .
Instead of solving two triangles , as directed by Captain Sumner , the same result may be obviously obtained by * * A new and accurate method of finding a Ship 's Position at Sea , ' by Capt. T. H. Sumner .
Boston , 1843 .
" In 1843 , Commander Sullivan , R.N. , not having heard of " this work , found the line of equal altitude on entering the River Plate ; and identifying " the ship 's place on it in 12fathoms by means of the chart , shaped his course up the " river .
The idea may thus have suggested itself to others ; but the credit of having " reduced it to a method and made it public belongs to Capt. Sumner .
' ' ( Raper 's Navigation , edition 1857 .
) Sir W. Thomson on the Determination of a [ Jan. 19 , finding a second angle ( Z ) of the one triangle ( P Z S ) ordinarily solved ( P being the earth 's pole , Z the ship 's zenith , and S the sun or star ) .
The angle ordinarily calculated is P , the hour-angle .
By calculating Z , the sun 's azimuth also , from the same triangle , the locus on which the ship must be is of course found by drawing on the chart , through the point which would be the ship 's place were the assumed latitude exactly correct , a line inclined to the east and west at an angle equal to Z. But , as Captain Mo-riarty pointed out to me , the calculation of the second angle would involve about as much work as solving for P a second triangle with a slightly different latitude ; and Capt. Sumner 's own method has practical advantages in affording a check on the accuracy of the calculation by repetition with varied data .
A little experience at sea suggests that it would be very desirable to dispense with the morning and evening spherical triangles altogether , and to abolish calculation as far as possible in the ordinary day 's work .
When we consider the thousands of triangles daily calculated among all the ships at sea , we might be led for a moment to imagine that every one has been already solved , and that each new calculation is merely a repetition of one already made ; but this would be a prodigious error ; for nothing short of accuracy to the nearest minute in the use of the data would thoroughly suffice for practical purposes .
Now , there are 5400 minutes in 90 ' , and therefore there are 5400s or 157,464,000,000 triangles to be solved each for a single angle .
This , at 1000 fresh triangles per day , would occupy above 400,000 years .
Even with an artifice , such as that to be described below , for utilizing solutions of triangles with their sides integral numbers of degrees , the number to be solved ( being 90s or 729,000 ) would be too great , and the tabulation of the solutions would be too complicated ( on account of the trouble of entering for the three sides ) to be convenient for practice ; and Tables of this kind which have been actually calculated and published ( as , for instance , Lynn 's Horary Tables * ) have not come into general use .
It has occurred to me , however , that by dividing the problem into the solution of two right-angled triangles , it may be practically worked out so as to give the ship 's place as accurately as it can be deduced from the observations , without any calculation at all , by aid of a table of the solution of the 8100 right-angled spherical triangles of which the legs are integral numbers of degrees .
Let O be the point in which the arc of a great circle gt less than 90 ' through S , perpendicular to P Z , meets P Z or P Z produced .
t _ If the data were S P , P Z , and the hour-angle P , the solution of the right-angled triangle S P O would give P O and S O. Subtracting P ^ * Horary Tables for finding the time by inspection \amp ; c. , by Thomas Lynn , late Com-mander in the sea-service of the East-India Company .
London , 1827 , 4to .
a=51 ' .
co-hyp .
0 _ 14 31 49 _7 II \lt ; 3 co-hyp .
A. o / 38 0 o " J 89 13 37 58 88 26 37 56 87 40 37 53 86 53 37 50 86 6 37 45 85 20 37 40 84 34 37 34 83 48 37 27 83 2 37 19 82 17 37 11 81 31 37 2 80 46 36 52 80 2 36 41 79 18 36 29 78 34 36 17 77 51 36 4 77 8 35 50 76 26 35 36 75 44 35 21 75 2 35 5 74 21 | 34 48 73 41 4 18 52 12 [ 3 41 52 9 3 5 52 6 2 28 52 4 1 51 52 2 1 14 52 1 0 37 52 0 0 0 1 52 0 1 a=53 ' .
a=54 ' .
co-hyp .
A. O / 3o-hyp .
A. o 36 0 o / 89 16 35 59 88 33 35 57 87 49 35 54 87 6 35 50 86 23 35 46 85 39 35 41 84 56 35 36 84 14 35 29 83 31 35 22 82 49 35 14 82 6 35 6 81 25 34 56 80 43 34 46 80 2 34 36 79 21 34 24 78 41 I 34 12 78 0 33 59 77 21 33 46 76 42 133 32 76 3 133 17 75 24 133 1 if 4\#163 ; 1*4 6 54 12 \#166 ; 3 31 54 9 |2 56 54 6 \#166 ; 2 21 54 4 \#166 ; l 46 54 2 \#166 ; l 11 54 1 a=55 ' .
ir* II a=57 ' .
c=58 .
b. co-hyp .
A. co-hyp .
A. co-hyp .
A. co-hyp .
A. 0 / 35 0 .
; i 89 18 o / 34 0 89 20 o / 33 0 O _ / 89 21 o / 32 0 89 23 0 1 34 59 88 36 33 59 88 39 32 59 88 42 31 59 88 45 2 34 57 87 54 33 57 87 59 32 57 88 3 31 57 88 8 3 34 54 87 12 33 54 87 18 32 55 87 24 31 55 87 30 4 34 51 86 30 33 51 86 38 32 52 86 46 31 52 86 53 5 34 47 85 49 33 47 85 58 32 48 86 7 31 48 86 16 6 34 42 85 7 1 33 42 85 18 32 43 85 29 31 44 85 39 7 34 37 84 26 33 37 84 38 32 38 84 50 31 39 85 2 8 34 30 83 45 33 32 83 59 32 33 84 12 31 34 84 25 9 34 24 83 4 33 25 83 19 32 26 83 34 31 27 83 48 10 34 16 82 23 33 18 82 40 32 19 82 56 31 21 83 12 11 34 8 81 43 33 10 82 1 32 11 82 19 31 13 82 36 12 33 59 81 3 33 1 81 22 32 3 81 41 31 5 82 0 13 33 49 80 23 32 52 80 44 31 54 81 4 30 57 81 24 14 33 39 79 44 32 42 80 6 31 44 80 28 30 47 80 49 15 33 28 79 5 32 31 79 28 31 34 79 51 30 37 80 14 16 33 16 78 26 32 20 78 51 31 23 79 15 30 27 79 39 17 33 4 77 47 32 8 78 14 31 12 78 39 30 16 79 4 18 32 51 77 9 31 55 77 37 31 0 78 4 30 4 78 30 19 32 37 76 32 31 42 77 1 30 47 77 29 29 52 77 56 20 32 23 75 55 31 28 76 25 30 34 76 54 29 39 77 23 21 32 8 75 18 31 14 75 49 \lt ; 7g 3 4 30 20 L *\gt ; /v ... r. 76 20 FJ* \#166 ; -A \#163 ; 5 29 26 1 , flA JUfc- .
76 50 nr* i1 22 4 0 55 12 3*55*1 56 12 | 3 48 57 12 3 42 58 12 83 3 26 55 9 3 21 56 9 3 16 57 9 3 11 58 9 84 2 52 55 6 2 48 56 6 2 43 57 6 2 39 58 6 85 2 18 55 4 2 14 56 4 2 11 57 4 2 7 58 4 86 1 43 55 2 1 41 56 2 1 38 57 2 1 35 58 2 87 1 9 55 1 ^^7 ^^5 .57 1 1 4 58 1 88 157 0 !
0 32 58 0 89 157 0 0 ,0 58 0 90 1871 .
] Ship 's Place from Observations of Altitude .
from P O , we have Z O ; and this , with S O in the triangle S Z O , gives the zenith distance , S Z , and the azimuth , S Z O , of the body observed .
Suppose , now , that the solution of the right-angled spherical triangle S P O for P O and S O to the nearest integral numbers of degrees could suffice .
Further , suppose P Z to be the integral number of degrees closest to the estimated co-latitude , then Z O will be also an integral number of degrees .
Thus the two right-angled spherical triangles S P O and S Z O have each arcs of integral numbers of degrees for legs .
Now I find that the two steps which I have just indicated can be so managed as to give , with all attainable accuracy , the whole information deducible from them regarding the ship 's place .
Thus the necessity for calculating the solutions of spherical triangles in the ordinary day 's work at sea is altogether done away with , provided a convenient Table of the solutions of the 8100 triangles is available .
I have accordingly , with the cooperation of Mr. E. Roberts , of the * Nautical Almanac ' Office , put the calculation in hand ; and I hope soon to be able to publish a Table of solutions of right-angled spherical triangles , showing co-hypotenuse* and one angle , to the nearest minute , for every pair of values of the legs from 0 ' to 90 ' .
The rule to be presently given for using the Tables will be readily understood when it is considered that the data for the two triangles are their co-hypotenuses , the difference between a leg of one and a leg of the other , and the condition that the other leg is common to the two triangles .
The Table is arranged with all the 90 values for one leg ( b ) in a vertical column , at the head of which is written the value of the other leg ( a ) .
Although this value is really not wanted for the particular nautical problem in question , there are other applications of the Table for which it may be useful .
On the same level with the value of b , in the column corresponding to a , the Table shows the value of the co-hypotenuse and of the angle A opposite to the leg a. I take first the case in which latitude and declination are of the same name , the latitude is greater than the declination , and the azimuth ( reckoned from south or north , according as the sun crosses the meridian to the south or north of the zenith of the ship 's place ) is less than 90 ' .
The hypotenuses , legs , and angles P and Z of the two right-angled triangles of the preceding diagram are each of them positive and less than 90 ' , and the two co-hypotenuses are the sun 's declination and altitude respectively .
We have then the following rule:\#151 ; ( 1 ) Estimate the latitude to the nearest integral number of degrees by dead reckoning .
( 2 ) Look from one vertical column to another , until one is found in which co-hypotenuses approximately agreeing with the declination and altitude are found opposite to values of b which differ by the complement of the assumed latitude .
( 3 ) The exact values of the co-hypotenuse and the angle A corresponding * It ia more convenient that the complements of the hypotenuses should be shown than the hypotenuses , as the trouble of taking the complements of the declination and the observed altitude is so saved .
Sir W. Thomson on the Determination of a [ Jan. 19 , to these values of b are to be taken as approximate declination , hour-angle , altitude , and azimuth .
( 4 ) Either in the same or in a contiguous vertical column find similarly another set of four approximate values , the two sets being such that one of the declinations is a little less and the other a little greater than the true declination .
( 5 ) On the assumed parallel of latitude mark off the points for which the actual hour-angles at the time of observation were exactly equal to the approximate hour-angles thus taken from the Table .
With these points as centres , and with radii equal ( miles for minutes ) to the differences of the approximate altitude from the observed altitude , describe circles .
By aid of a parallel ruler and protractor * , draw tangents to these circles , inclined to the parallel of latitude , at angles equal to the approximate azimuths taken from the Table .
These angles , if taken on the side of the parallel away from the sun , must be measured from the easterly direction , or the westerly direction , according as the observation was made before or after noon .
The tangent must be taken on the side of the circle towards the sun , or from the sun , according as the observed altitude was greater or less than the approximate altitude taken from the Tables in each case .
The two tangents thus drawn will be found very nearly parallel .
Draw a line dividing the space between them into parts proportional to the differences of the true declination , from the two approximate values taken from the Tables .
The ship 's place at the time of the observation was somewhere on the line thus found .
To facilitate the execution of clause ( 2 ) of the rule , a narrow slip of card should be prepared with numbers 0 to 90 printed or written upon it at equal intervals , in a vertical column , equal to the intervals in the vertical column of the Table , 0 being at the top and 90 at the bottom of the column as in the Table .
Place number 90 of the card abreast of a value of co-hypotenuse in the Table approximately equal to the declination , and look for the other co-hypotenuse abreast of the number on the card equal to the assumed latitude .
Shift the card from column to column according to this condition until the co-hypotenuse abreast of the number on the card equal to the assumed latitude is found to agree approximately enough with the observed altitude .
When the declination and latitude are of contrary names and the azimuth less than 90 ' , or when they are of the same names , but the declination greater than the latitude , the sum , instead of the difference , of the legs b of the two triangles will be equal to the complement of the assumed latitude ; and clause ( 2 ) of the rule must be altered accordingly .
The slip of card in this case cannot be used ; but the following scarcely less easy process is to be practised .
Put one point of a pair of compasses on a position in one of the vertical columns of the hypotenuse abreast of * A circle divided to degrees , and having its centre at the centre of the chart , ought to be printed on every chart .
This , rendering in all cases the use of a separate pro tractor unnecessary , would be useful for many purposes .
1871.1 Ship 's Place from Observations of Altitude .
that point of the column of values of b corresponding to half the complement of the assumed latitude ; this point will be on a level with one of the numbers , or midway between that of two consecutive numbers , according as the assumed latitude is even or odd : then use the compasses to indicate pairs of co-hypotenuses equidistant in the vertical column from the fixed point of the compasses , and try from one column to another until co-hypotenuses approximately agreeing with the observed altitude and the correct declination are found .
It is easy to modify the rule so as to suit cases in which the azimuth is an obtuse angle ; but it is not worth while to do so at present , as such cases are rarely used in practice .
The following examples will sufficiently illustrate the method of using the Tables:\#151 ; ( 1 ) On 1870 , May 16 , afternoon , at 5h .
42m .
Greenwich apparent time , the Sun 's altitude was observed to be 32 ' 4 ' : to find the ship 's place , the assumed latitude being 54 ' North .
The Nautical Almanac gives at 1870 , May 16 , 5h .
42m .
Greenwich apparent time , the Sun 's apparent declination N. 19 ' 10* .
On looking at the annexed Table ( which is a portion of the solutions of the 8100 right-angled spherical triangles ) under the heading a = 56 ' , and opposite b = 64 ' , the co-hypotenuse ( representing the Sun 's declination ) is 19 ' 11* , and opposite b \#151 ; 18 ' ( differing from 54 ' by the complement of the assumed latitude ) , the co-hypotenuse ( representing the Sun 's altitude ) is 32 ' 8f , which are sufficiently near the actual values ; we therefore select our sets of values from these columns as follows :\#151 ; Co-hyp .
A. r 1 .
6=54 19 11 61 2^=Sun 's hour-angle .
6=18 32 8 78 14=Sun 's azimuth ( S. towards W. ) .
a = 56 ' \#166 ; * | 2 .
6=55 18 42 61 5=Sun 's hour-angle .
v. 6=19 31 55 77 37=Sun 's azimuth ( S. towards W. ) .
from which we have the following:\#151 ; O J o t Greenwich apparent time ( in arc ) 85 30 ... ... ... ... 85 30 Sun 's hour-angle ( 1 ) 61 23 ... ... ... .(2 ) 61 5 Diff. = Longitude 24 7 W. 24 25 W. Sun 's altitude ( observed ) 32 4 ... ... ... ... 32 4 Sun 's altitudes ( auxiliary ) ( 1 ) 32 8 ... ... ... ... ... ... .(2 ) 31 55 Diff. = - 4 +9 Sun 's declination from N. A. 19 10 ... ... ... . .
19 l6 Sun 's declinations ( auxiliary ) ( 1 ) 19 11 ... ... ... ... ..(2 ) 18 42 Diff + 28 Sir W. Thomson on the Determination of a [ Jan. 19 This example is represented graphically in the first diagram annexed .
The second set of values could have been selected equally well from the contiguous columns ( a=5 7 ' ) , which on trial will be found to give an almost identical result .
Again , ( 2 ) , on 1870 , May 16 , afternoon , at 5h .
42m .
Greenwich apparent time , the Sun 's altitude was observed to be 30 ' 30 ' : to find the ship 's place , the assumed latitude being 10 ' North .
The Sun 's declination from N. A. is N. 19 ' 10 ' , and the half complement of the assumed latitude 40 ' .
By a few successive trials , a=56 ' will be found to contain values of co-hypotenuses approximately equal to the Sun 's declination and altitude at the time* and which are equidistant from 40 ' ; we therefore select the following sets of values from this column as follows : \#151 ; Co-hyp .
A. ( 1 .
O 6=54 O 1 19 11 61 23= Sun 's hour-angle .
a = 56 ' * 6=26 30 10 73 32=Sun 's azimuth ( N. towards W. ) .
2 .
6=55 18 42 61 5 = Sun 's hour-angle .
6 = 27 29 53 72 58 = Sun 's azimuth ( N. towards W. ) .
from which we have the following :\#151 ; Greenwich apparent time ( in arc ) Suns hour-angle ( 1 ) 85 30 61 23 ... .(2 ) 85 30 61 5 Diff. = Longitude 24 7 W. 24 25 W. Sun 's altitude ( observed ) Sun 's altitudes ( auxiliary ) ( i ) 30 30 30 10 ... , ( 2 ) 30 30 29 53 Diff. xst + 20 + 37 Sun 's declination from N. A. Sun 's declinations ( auxiliary ) ( 1 ) O / 19 10 19 11 ... .(2 ) o 4 19 10 18 42 Diff. = - 1 + 28 In this case the sun passes the meridian to the north of the ship 's zenith , the azimuth , from the Tables being less than 90 ' , is measured from the north towards the west .
In this case also the second set of values might have been taken from a=57 ' , which will be found on trial to give a position nearly identical with the above .
This example is represented in the second diagram annexed .
Again , ( 3 ) , on 1870 , May 16 , afternoon , at 5h .
42m .
Greenwich apparent time , the Sun 's altitude was observed to be 18 ' 35 ' : to find the ship 's place , the assumed latitude being 20 ' South .
The Sun 's declination from N. A. is N. 19 ' 10 ' , and the half complement of the assumed latitude is 55 ' , to be used because the Sun 's declination and the assumed latitude are of different names .
Proceeding as in the previous 266 Prof. Story Maskelyne on the [ Jan. 26 , example , we find the column a\#151 ; 56 ' again to contain values of co-hypotenuses approximately equal to the given values ; and therefore have :\#151 ; I2 ' which give Co-hyp .
' o o / 5=54 19 11 6 = 56 18 13 6=55 18 42 6=57 17 44 A. O i 61 23 = Sun 's hour-angle .
60 47 = Sun 's azimuth ( N. towards W. ) .
61 5 = Sun 's hour-angle .
60 30 = Sun 's azimuth ( H. towards W. ) .
o t o / Greenwich apparent time ( in arc ) 85 30 ... ... ... .
85 30 Sun 's hour-angle ( 1 ) 6123 ... ... ... ( 2 ) 61 5 Diff. = Longitude 24 7 W. 24 25 W. Sun 's altitude ( observed ) 18 35 ... ... ... ... 18 35 Sun 's altitudes ( auxiliary ) ( 1 ) 18 13 ... ... ... ..(2 ) 17 44 Diff. = +22 + 51 Sun 's declination from N. A. 19 10 ... ... ... ... 19 10 Sun 's declinations ( auxiliary ) ( 1 ) 19 11 ... ... ... ... .(2 ) 18 42 Diff. = - 1 +28 This example is represented in the third diagram annexed .
January 26 , 1871 .
General Sir EDWARD SABINE , K.C.B. , President , in the Chair .
The following communications were read :\#151 ; I. , l On the Mineral Constituents of Meteorites .
" By Nevil Story Maskelyne , M.A. , F.R.S. , Professor of Mineralogy , Oxford , and Keeper of the Mineral Department , British Museum .
Received November 3 , 1870 .
( Abstract .
) In the memoir now offered to the Society the author gives the results of his investigation of the meteorites of Breitenbach and of Shalka .
A preliminary notice of two of the minerals occurring in the former , which is of the Siderolite class , was read before the Society in March , 1869 ( Proc. R. S. vol. xvii .
p. 370 ) .
After entering upon the probable history of the Breitenbach Siderolite , and endeavouring to identify it with certain other Siderolites that have been found , or have been recorded as found , in the region extending from

