XX .
M. De Luc 's Rule for measuring Heights by the Barometer , reduced to the English Measure of Length , and adapted to Fahrenheit 's Thermometer , and other Scales of Heat , and reduced to a more convenient Expression .
By the Astronomer Royal .
M. DE LUC , F. R. S. in a large and valuable Treatise upon the Barometer and Thermometer , lately published at Geneva , the result of many years labour and study , has given a rule for the measurement of heights by the barometer , deduced from his experiments , and far more accurate , than any published before ; since it appears that he could determine heights by it generally to 10 or 15 feet , and that the error seldom , if ever , amounted to double that quantity .
This valuable degree of exactness he has obtained principally by detecting the faults of the common barometer , and , in consequence , improving the construction of it ; and by introducing the use of the mercurial thermometer , to accompany that of the barometer .
The principal faults , which he found in the common barometers , arose from 2 the Red , Jan. the repulsion of the quicksilver by the glass tube , from air and moisture admitted into the tube , and from the variations of the density of quicksilver by heat and cold ; another very considerable error arose , in calculating heights from the barometer , by not allowing for the changes of the density of the air , whose gravity affords us this measure of heights , owing to heat and cold .
The first cause of error , that of the repulsion of the tubes , he remedied , by substituting a syphon-barometer instead of the simple upright tube , the repulsion of the two legs of the syphon , counteracting itself ; the error arising from air and moisture in the tube he cured by boiling the quicksilver , after it was put into the tube , and other precautions ; the errors , in the estimation of the heights , arising from the changes of the density of the quicksilver , and density of the air by heat and cold , he shows how to correct by allowances depending on two thermometers , one attached to the frame of the barometer itself , and the other made to be exposed to the open air , to show its degree of heat ; which thermometers are to be noted both at the top and bottom of the hill .
Lastly , by a great number of experiments made with accurate barometers and thermometers of hs own construction , he has deduced a rule for calculating heights of places ; the exactness of which he has sufficiently proved by a large table of experiments .
But this rule is expressed in French measure , and is adapted either to a thermometer , whose freezing point is 0 , and that of boiling water 80 , or to thermometers of particular scales .
It may be therefore useful , to reduce M. DE LUC'S rule to English measure , and to adapt it to the thermometer of Fahrenheit 's scale , which is generally used in this country .
M. DE LUC , in the winter season , heated the air of his room to as great degree as he could , and noted the rise of the barometer , owing to the diminution of its density , or specific gravity , by heat ; he also noted the height of the thermometer , both before and after the room was heated .
Hence he deduced a rule , that when the barometer is at 27 French inches , which was the case in this experiment , an increase of heat , from freezing to that of boiling water , will raise the barometer 6 lines , or 1/ 5 4 th part of the whole .
It is easy to see , that when the barometer is higher than 27 French inches , this variation will increase in the same proportion ; or will be always 1/ 5 4 th of the heighth of the barometer ; therefore , if the height of the barometer be called B , the rise of the barometer , for an increase of heat from freezing to boiling water , will be B/ 54 ; and , as it will be less for a less difference of heat , therefore , if the number of degrees , marked on the thermometer , between freezing and boiling water , be called K , and the rise of the thermometer from any given point be called H , the correspondent rise of the barometer will be B/ 54 X H/ K , by the increase of heat from the given point by the number of degrees H. If the heat , instead of increasing , was to decrease , then H would signify so many degrees decrease of heat , and the barometer would sink by B/ 54 X H/ K. The fixed temperature of heat , to which .
M. DE LUC thought best to reduce his observations of the barometer , is 1/ 8th of the interval from freezing to boiling water above the former point : and if the thermometer was higher than this degree , he substracted B/ 54 X H/ K ; if it was lower , he added it to the observed height of the barometer ; and thus he obtained the exact height of the barometer , such as it would have been , ~ density of its quicksilver had been the same as answers to the fixed degree of temperature .
He thus corrected the height of both his barometers ( that at the bottom , and that at the top of the hill ) for the particular degree of heat , indicated by a thermometer attached to the barometer , at each station ; for it might and would commonly happen , that the degree of heat would be different at the two stations .
The heights of the barometers , thus corrected , were what he made use of in his subsequent calculations .
Calling these two altitudes of the barometer B and b , putting log .
B and log .
b , for the logarithms of B , and b , taking only the four first places of figures , after the characteristic , or considering the remaining figures as decimals , and putting C for the mean height of a thermometer , exposed to the air at top and bottom of the hill , the freezing point being 0 , and the point of boiling water at 80 , he finds , by his experiments , that the height of the hill will be given in French toises , when C is 163/ 4 , by simply taking the difference of the logarithms of the heights of the barometer , or will be equal to log .
B log .
b : and in any other degree of heat , will be greater or less , in proportion as the rarity of the air is greater or less , than in the fixed temperature ; or greater or less by 1/ 2 15 th part of the whole , for every degree of the thermometer reckoned from the fixed temperature 163/ 4 ; and consequently the height of the place will be expressed generally in French toises , by this formula log .
B log .
b+ log .
B log .
bXC 163/ 4/ 215 =log .
B log .
bX1+ C-163/ 4/ 215 .
To reduce this formula to English measure , and to the scale of Fahrenheit 's thermometer , we should first premise some particulars .
The French foot is to the English foot , as 1,06575 to 1 , as was found by a very accurate experiment : see Phil. Trans. Vol. LVIII .
for 1768 , p. 326 ; and it is well known , that the point of freezing , on Fahrenheit 's thermometer , is at 32 , and that of boiling water at 212 , or the interval between them 180 degrees .
But M. DE LUC'S point of boiling water 80 , was marked when the barometer was at 27 French inches ; and it is the custom of our principal English workmen to mark the point of boiling water , 212 , on Fahrenheit 's thermometer , when the barometer stands at 30 inches , which is equal to 28 inches , 1,8 lines French measure ; or 13,8 lines higher than M. DE LUC'S barometer , when he set off the point of boiling water on his thermometers ; and it is well known , that the heat of boiling water varies with the weight of the atmosphere : M. DE LUC finds , by his experiments , ments , this rule , that an increase of 1 line in the height of the barometer , raises the quicksilver of the thermometer , placed in boiling water , by 1/ 1134 th part of the interval between the freezing point and that of boiling water : he afterwards indeed found , that this rule would not answer for such large variations of the barometer , as take place in ascending to very great heights above the earth 's surface ; vid .
Essai sir les variations du chal : de l'eau bouill : but it is accurate enough for any small variation of the barometer , on one side or other of its mean height in these lowest regions of the atmosphere .
The change therefore of the boiling point on Fahrenheit 's scale , for a change of 1 line in the barometer , will be 180/ 1134 = 0,16 ; therefore 13 , 8 lines will cause 0,16 X 13,8 = 2,2 degrees of Fahrenheit 's scale ; and a thermometer , whose point of boiling water was marked 212 , when the barometer stood at 30 English inches=28 inches .
1,8 lines French measure , will , when the barometer descends to 27 French inches , sink 2,2 degrees in boiling water , or to 209,8 , or in round numbers to 210 degrees , which is distant only 178 from 32 , the point of freezing .
Hence an extent of 80~d of M. DE LUC'S thermometer answers to an extent of 178 of our Fahrenheit 's thermometer ; and putting F for the degrees of this thermometer , corresponding to C of M. DE LUC'S , we shall have C : F 32 : : 80 : 178 , and C=F-32 X 80/ 178 which substituted in M. DE LUC'S formula , gives log .
B log .
bX1+C 163/ 4/ 215 = =log .
B log .
bX1+F 32 X 80/ 178 163/ 4/ 215 = =log .
B log .
bX 1+80/ 178x215 F-32-178/ 80 16,75 = log .
B log .
bX1+F 32 37,27/ 478,38 = = log .
B log .
bX1+F 69,27/ 478,38 .
Where the answer will still come out in French toises , though adapted to Fahrenheit 's thermometer .
To bring it out in English fathom ( or measure of 6 feet ) multiply the above expression by 1,06575 , and we shall have log .
B log .
bX1+F 69.27/ 478,38 X 1,06575 = log .
B log .
bX 409,11 +FX 1,06575/ 478,38 = = log .
B log .
bX 409,11 + F/ 448,87 , or in round numbers = log .
B log .
bX 409 + F/ 449 = = log .
B log .
bX1+F 40/ 449 ; which will express the height between the two stations in English fathom .
In the foregoing expressions , B and b , as has been mentioned before , signify heights of the barometer , at the lower and higher stations , both corrected according to Mr. DE LUC'S directions , for the difference of heat between a fixed temperature , ( namely 1/ 8th of the interval between freezing and boiling water ) , and the present heat , indicated by the thermometer attached to the barometer at each station ; but it is not necessary , to correct both barometers for the effect of heat , but only one for the difference of heat of the two ; which will be more convenient also on another account , because the difference of heat , at the two stations , will be generally small , and the correction to reduce one barometer to the heat of the other will consequently be small also ; whereas the difference of the present heat , and the fixed temperature , and consequently the correction of both barometers , may be frequently very considerable : this is evident : because if the heat of the barometers , at both stations , was the same , however different from the fixed temperature chosen by M. DE LUC , no correction would be necessary ; the mercury in the barometer in both stations , being expanded in the same proportion , and consequently the difference of the logarithms of its height , at both stations , being the same , as if the heat of both barometers had agreed with that of the fixed temperature .
I shall now therefore suppose the upper barometer is to be corrected , to reduce it to the temperature of the lower one , and that b signifies the height of this barometer , as observed , and not yet corrected ; the correction , from what has been said above , calling D the difference of height of the thermometer attached to the barometer at the two stations , will be ~D b/ 54 K , according as the thermometer stands highest at the lower or upper station ; and the upper barometer corrected , instead of b , will be b~D b/ 54 K , which substituted in the formula , gives log .
B log .
( b ~D b/ 54 K ) X1+F 40/ 449 .
But the correction , on account of the difference of heat of the barometer at the two stations , may be reduced to a still easier expression , in which the variable quantity b , the height of the upper barometer , shall not appear .
The fluxion of a logarithm is to the fluxion of its natural number , as the modulus of the system to the natural number ; and 4343 is the modulus of the common logarithms , when the four places , next following the characteristic , are taken as whole numbers , instead of decimals , which is meant to be done in the use of the foregoing formula .
Therefore D b/ 54 K being very small with respect to b , we shall have , variation of log .
b : variation of b ( = D b/ 54 K ) : : 4343 : b very nearly , and thence variation of log .
b=~D b/ 54 KX 4343/ b =~ 4343 D/ 54 K. Which ( putting K= 178 ) =~ 0,452 D. Hence log .
( b ~D b/ 54 K ) = log .
b~ 0,452 D ; which being substituted in the formula above , will give the difference of height , of the two stations , in English fathom , in a more convenient expression , namely log .
B log .
b~ 0,452 DX1+F 40/ 449 ; where the upper sign , - , is to be used , when the thermometer of the barometer is highest at the lower station , and the lower sign , + , is to be used , when the said thermometer is lowest at the lower station .
The first case will be most common ; especially where the difference of height of the two stations is considerable .
It should also be observed , that when F , the height of Fahrenheit 's thermometer , is less than 40~ , + F-40/ 449 becoming negative or substractive , must be applied in the calculation accordingly .
It may perhaps be convenient to repeat here the meaning of the algebraic terms , used in the foregoing formula , that any person may make use of it , withoug having occasion to recur to the foregoing investigation .
B signifies the observed altitude of the barometer at the lower station , and b that at the upper station ; Log. B and Log. b , signify their logarithms taken out of the common tables , by assuming the four first figures next following the characteristic as whole numbers , and considering the three remaining figures , to the right hand , as decimals ; D signifies the difference of height of Fahrenheit 's thermometer , attached to the barometer at the top and bottom of the hill ; and F signifies the mean of the two heights of Fahrenheit 's thermometer , exposed freely for a few minutes to the open air in the shade , at the top and bottom of the hill .
The formula , for the measure of heights , may also be changed , and adapted to thermometers of particular scales , for the convenience of calculation , as M. DE LUC has done ; but these scales will be different from his .
The thermometer , attached to the barometer , had better be divided with the interval between freezing and boiling water , consisting of 81,4 degrees ( = 180 X , 452 ) the freezing point may be marked 0 , and the point of boiling water will be 81,4 ; for then , if the difference of height of this thermometer , at the two stations , be called d , we shall have d= , 0,452 x D. for d : D : : 81,4 : 180 : : 0,452 : 1 , and the number of degrees expressed by d , will show immediately the correction for the difference of heat of the two barometers .
If the thermometer , designed to show the temperature of the air , be divided with the interval between freezing and boiling water = 200 , and the freezing point be marked 9 , and the boiling point + 191 , and the heights of this thermometer , at the two stations , be called G and I , we shall have F 40/ 449 =G+ 1/ 2x500 =G + 1/ 1000 .
For F 40 =F 32-8 , is the height of Fahrenheit 's thermometer , reckones from 8 degrees above freezing , and 449 : 500 : : 180 : 200 : : 8 : 9 , and the fraction F 32 8/ 449 , if both the numerator and denominator be increased in the ratio of 449 to 500 , will become =F 32 8x 500/ 449/ 500 =F 32 500/ 449 9/ 500=G + 1/ 2 x 500=G + 1/ 1000 , because G+1/ 2 +9 =F 32 X 500/ 449 .
Therefore , if the thermometer of the barometer has the freezing pointmarked 0 , and the point of boiling water 81,4 , and the difference of its height , at the two stations , be called d ; and the thermometer for measuring the temperature of the air , be divided with the interval of 200 between the freezing point and that of boiling water , and the first be marked 9 , and the latter + 191 , and the degrees , shown by this , at the two stations , be called G and I ; the formula , that will give the height of the upper station above the lower one , in English fathoms , will be log .
B log .
b~dx1+G+ 1/ 1000 ; which consequently multiplied by 6 , will give the height in English feet .
It is to be observed , as before , that d or +d is to be used , according as the thermometer , attached to the barometer , is highest at the lower or upper station ; and if G and I should happen to fall below 0 of the scale , or to be substractive , they must be applied accordingly in the calculation .
I shall now add nothing more , but to give the rule for finding heights by the barometer , according to the sormulae delivered above , in common language ; first , as adapted to the two thermometers of particular scales .
Take the difference of the tabular logarithms of the observed heights of the barometer , at the two stations , considering the 4 first figures , exclusive of the index , as whole numbers , and the three remaining figures to the right as decimals , and substract or add 452/ 1000th of the difference of the altitude of the Fahrenheit 's thermometer , attached to the barometer at the two stations , according as it was highest at the lower or upper station ; thus you will have the height of the upper station above the lower , in English fathom , nearly ; to be corrected , as follows : make this proportion ; as 449 is to the difference of the mean altitude of Fahrenheit 's thermometer , exposed to the air at the two stations , from 40~ , so is the height of the upper station found nearly , to the correction of the same ; which added or substracted , according as the mean altitude of Fahrenheit 's thermometer was higher or lower than 40~ , will give the true height of the upper station above the lower , in English fathoms ; and multiplied by 6 , will give it in English feet .
The same rule , adapted to the thermometers of particular scales , is this : Take the difference of the tabular logarithms of the observed heights of the barometer , at the two stations , considerint the 4 first figures ( exclusive of the index ) as whole numbers , and the three remaining figures to the right as decimals ; and substract or add the difference of the thermometer , of a particular scale , attached to the barometer , at the two stations , according as it was highest at the lower or upper station , and you will have the height of the upper station above the lower one , in English fathom , nearly ; to be corrected as follows : make this proportion ; as 1000 is to the sum of the altitudes of the thermometer of a particular scale , exposed to the air at both stations , so is the height of the upper station above the lower , found nearly , to the correction of the same ; which added or substracted , according as the sum of the altitudes of the thermometers , exposed to the air , is positive or negative , will give the true height of the upper station above the lower in English fathoms ; and multiplied by 6 , will give it is English feet .
NEVIL MASKELYNE .

