XXX .
M. DELU C'S Rules , for the measurement of Heights by the Barometer , compared with Theory , and reduced to English Measures of Length , and adapted to Fahrenheit 's Scale of the Thermometer : with Tables and Precepts , for expediting the practical Application of them .
By Samuel Horsley , LL. D. addressed to Sir John Pringle , Bart. P. R. S. TOSIRJOHNPRINGL E , BAR T. P. R. S. SIR , IN the papers which I have the honor to present to you , nothing more was at first intended , than to reduce the formulae given by Mr. DELU C , in his elaborate work upon the modifications of the atmosphere , for expressing differences of elevation , as indicated by the barometer and thermometer , in Paris toises , to others , which , from the like data , should express such differences in English fathom ; and at the same time be adapted to that scale of the thermometer , which is in general use in this country .
Had I confined myself to this , which was my original Red , Feb. 17 , ginal design , I should certainly have suppressed whatever I might have executed , as I find my learned friend the ASTRONOMERROYAL hath devoted some of his leisure hours to these calculations ; and the reductions , I proposed , are actually performed , in a short and elegant paper of his upon the subject , which is already , I believe , in Dr. MAT Y'S hands .
But I had made but a small progress in my intended work , when it occurred to me , that tables of the equations for the effects of heat and cold , would be very useful .
Such I have taken the pains to construct , and have carried them to as great a length , as can be ever wanted .
They are annexed to the ensuing piece , and will render the application of Mr. DELU C'S rules more easy and expeditious , than any peculiar divisions of the thermometrical scale .
When my tables were finished , I thought it might be still further useful , to give a succinct explanation of Mr. DELU C'S original formulae ; that such as have not leisure to peruse his excellent work , might be furnished with a competent idea of the result of his researches .
This , I found , I could not do , in any way so satisfactory to myself , as by opening , as I went along , the principles of theory , in which the conclusions , he hath arrived at , appear to originate .
And thus I was insensibly led into minute disquisitions , concerning the agreement of Mr. DELU C'S conclusions , from a long train of accurate experiments , with the geometrical theory of the atmosphere , founded on the general laws of gravitation .
In this manner , SIR , the papers before you have taken , as it were spontaneously , the form , in which they now appear ; and a subject , remotely connected with my first design , takes up the far greater part of them .
They can hardly be free from the imperfections naturally incident to productions , in which the plan hath been gradually changed , during the actual progress of the work .
I flatter myself , however , that they are not deficient in two essential points , the precision and the perspicuity of the mathematical reasoning ; and that , however unfinished in some particulars , they are such upon the whole , as the dignity of the subject may , in some degree commend to your protection .
There is perhaps no branch of physical enquiry , intrinsically more sublime and interresting , nor likely to be more important in its uses , than that which immediately regards the constitution of that elastic fluid , which surrounds our globe , and appears to be a principal agent , in many of the most striking operations of nature , and a necessary instrument , at least , in carrying on the wonderful business of vegetation and of animal life .
Upon a subject of so much importance , it must afford satisfaction , to find so exact an agreement , as is evinced , if I am not much deceived , in the ensuing pages , between a multiplicity of experiments , not suggested by any previous speculations of theory , with a theory , whose conclusions in this branch had never before been duly submitted to the test of experiment .
The whole of the following paper is divided into six sections .
The business of the first , is merely preparation for the principal work .
It contains a brief account of the sum of Mr. DELU C'S researches concerning the variation of the heat of boiling water , so far only as they respect the construction and comparison of mercurial thermometers ; and it exhibits an actual comparison of the scale of M. DELU C'S thermometer , with that which is likely to continue in general use here .
In the second , I state the general principles of measuring heights by the barometer ; or in other words , of the modification of the air , with respect to its condensations , at different heights , by the pressure of the superior parts , exclusive of every other cause .
These principles are recapitulated for the sake of perspicuity and order ; but , as they are generally well known , I refer to former writers for the demonstration of them .
This section concludes with the exposition of M. DELU C'S fundamental rule .
The third section treats of the equation depending on the difference of temperature of the quicksilver in the barometers , at the different places of synchronous observation .
It explains , whence the necessity of this correction arises , and considers a case , in which the application of it requires some particular attentions ; namely , the levelling a tract of flat country by the barometer .
The fourth section explains the equation depending on the temperature of the air .
This is likewise traced to its origin in theory ; and upon this occasion , I enquire into the condensations of a fluid , unequally elastic in its different parts , under different degrees of compressive force .
The fifth section contains the reduction of M. DE LU C'S formulae .
I have found , with pleasure , by the perusal of Mr. MASKELYN E'S paper , that we agree in our final conclusions .
To this section , I have subjoined two problems ; the one for estiV O L. LXIV .
Ff mating mating the variations which the density of the air , at any place , undergoes ; and the other for determining the specific gravity of the air of any temperature , at any elevation .
The sixth and last section points out certain consequences , which seem to follow from the discoveries already made in conjunction with the theory established .
These are only propounded to awaken curiosity , and promote enquiry .
I am well aware , how little theory is to be trusted , in its remote conclusions , on account of the necessary deficiencies of the physical data , upon which its reasonings are founded .
The true uses of it are , either to explain the mutual connexions and dependencies of things already known , or to suggest conjectures concerning what is unknown , to be tried by future experiment .
And he who applies it , with due circumspection , to these purposes , will always find it an useful engine .
I flatter myself , that I have assigned the true cause of some very singular phenomenon , remarked by M. DELU C. I have endeavoured , to treat every part of my subject , in the plainest manner , that the nature of it would admit ; and so diffusely , as to be , I hope , intelligible to all , who are moderately well founded in the mathematical sciences ; for I have observed with pleasure , that M. DELU C'S book hath raised a general curiosity upon the points it treats of ; and for that reason , it seemed the more necessary , to depart in this instance , from a practice of late become too general , with mathematicians , to write only for one another .
To which , if I am not mistaken , it is in great measure owing , that these noble and useful studies are less generally attended to , than they were in former times ; when men of eminence , in this , as well as in all other branches , placed their principal glory , no less in communicating knowledge , than in acquiring , or seeming to possess it ; and were content to dedicate large portions of their time , to the removing of those difficulties for others , which they had once surmounted for themselves .
It is true , that in the variety of mathematical discussions , there are some of such a nature , that it would be a desperate attempt to render them intelligible to any but mathematicians ; to those , who to the natural faculty of combining , have added the mechanical and acquired habits of analytic calculation .
Such subjects are best treated with scientific brevity .
But this is no reason for wrapping up others in a similar stile ; and I must observe , that conclusions , at least , are always susceptible of simple perspicuous exposition , however abstruse the mode of investigation , in some cases , may necessarily be , by which they are brought out .
Such , SIR , is the general scope and plan of the following piece , which is presented to you , in testimony of the author 's respect and gratitude .
I have the honour to be , SIR , your much obliged , and most obedient servant .
Temple , Nov. 18 , 1773 .
SAMUELHORSLE Y. SECTIONFIRS T. Of the VARIATION of the point of BOILINGWATE R , and the COMPARISON of THERMOMETER S. THE degree of heat wherewith any fluid boils , is invariably the same , under a given pressure ; but if the pressure be diminished or increased , the boiling heat is diminished or increased .
Water , placed under the exhausted receiver , would be converted into steam , with a degree of heat , far inferior to that , which is necessary to its boiling in the open air ; and under the pressure of its own vapour , confined in PAPI N'S digester , it is said to sustain a degree of heat , without boiling , far exceeding that , which , in the open air , would convert it into steam .
Hence it follows , that in climates , where the pressure of the atmosphere is liable to considerable change , the heat of boiling water , in the open air , will be different , at different times .
Consequently , thermometers , made in different states of the barometer , will disagree ; unless allowance hath been made , for the effect of the variation of the barometer , upon accurate principles .
If care were taken to adjust the boiling point , to the mean height of the barometer , in every country , the instruments of the same country would always be consistent ; but those of different countries would still disagree ; that is , they would express the same temperature differently , though their fundamental intervals should be similarly divided ; for , in every scale , the number of degrees above or below melting ice , by which any given temperature is expressed , will be as the value of each degree inversely ; that is , if each be a given part of the fundamental interval , as the value of the fundamental interval inversely ; but , if the degrees of different scales be different parts of the fundamental intervals , as the value of the fundamental interval inversely , and the number of degrees contained in it directly .
It is necessary here to explain some of my expressions .
By the fundamental interval is to be understood the whole extent of the scale between melting ice and boiling water .
This is not a particular length ; for its length is not the same , even in thermometers made in the same state of the barometer , unless their figures be equal and similar ; but it is a portion of the whole solid content of the thermometer , including so much of the tube , as reaches up to the point of melting ice ; and it is the same portion of the whole solid content in all made in the same state of the barometer ; but in such as have been made in different states , a different portion in each .
The degrees of the scale are subdivisions of this portion of solid content , into lesser aliquot parts .
With respect to the number of these subdivisions , the practice of different countries is different .
By the value of the fundamental interval , or of the degrees of a thermometer , I mean the fractions , which express what parts they are respectively of the whole solid content , terminated as above ; that is , by the point of melting ice : and by the proportion of the fundamental intervals , or of the degrees , of different instruments , compared together , I understand the proportion of these fractions .
tions .
Or every thing may be reduced to a comparison of lengths , if in comparing the instruments of different countries , the comparison be imagined to be made between two of equal dimensions , and similar figure , setting aside that part of each which is above the point of melting ice .
To compare the thermometers , therefore , of different countries , the proportions of their fundamental intervals to each other must be ascertained ; or , we must have some means of finding , upon one scale , the place of the boiling point of another .
For this purpose , a general solution is requisite of the following problem : ` ` The fundamental interval being given for a given height of the barometer , to find the fundamental interval , for any other given height of the barometer .
' ' The solution is furnished by M. DE LUC'S laborious researches .
M. DE LUC fixes the boiling point of his thermometer , when the barometer is at 27 inches French ( a ) , that being its mean height at GENEVA .
He divides the fundamental interval , after the French manner , into 80 equal parts ; and , by a great number of experiments , on the heat of boiling water , at different heights above the level of the sea , the detail of which is to be found in his Essai sir la Variation , &amp; c : he hath found , that the height of his thermometer , plunged in boiling water , may be expressed , in all states of the barometer , by the following formula , viz. 9 9/ 2 00000 log .
ya= T. In which , y denotes the height of the barometer , in sixteenths of a Paris line : T the height of a thermometer , plunged in boiling water , above melt(a ) Recherches , sir les Modificat , de l'Atmosphere , ~ .
451 .
a. ing ice , in hundredths of a degree of M. DE LUC scale ; and a the constant number 10387 ( b ) .
It is proper in this place to inform the reader , that M. DE LUC , by logarithms , always means the tabular or Briggian logarithms , and considers the 7 figures given by the tables , besides the index , as integral figures ; that is , he considers the eighth figure of the logarithm as standing in the place of units .
Throughout this paper , I have conformed myself to this manner of conceiving the tabular logarithms , that , upon the same subject , we may speak the same language .
But it is more usual , with mathematicians , and , in general , it is more convenient , to consider all the figures , after the index , as decimals .
Thus the number , which M. DE LUC expresses by 9 9/ 2 00000 log y , would , in the common mathematical stile , be 99 x 100/ 2 log .
y ; or , 99 x 50 log .
y. It is but seldom that the barometer in this country stands so low as 27 French inches .
30 inches English are little more than its mean height upon the ( b ) Ibid. ~ .
961 .
and ~ .
1143 .
note a. It may seem doubtful whether there is not a small error in the constant number 10387 .
The experiments for ascertaining the variation of the boiling point were made with a thermometer , of a peculiar scale ; and the formula deduced from them was this , 717 5/ 3 0000000 log .
y-5015 = the height of the thermometer , plunged in boiling water , above melting ice , in parts of that peculiar scale .
Recherch .
sir les Modificat .
de l'Atmosphere , ~ .
960 .
Now a degree of this scale was afterwards found to be to a degree of M. DE LUC 's common scale as 80 to 38691/ 4 .
And this proportion between the degrees of the two scales , gives 10368,9 or 10369 very nearly , instead of 10387 , for the value of the constant number a , in the formula for hundredths of a degree of the common scale .
plane country in the neighbourhood of LONDO N. It may , therefore , be proper for the London workmen to fix their boiling point , when the barometer is at 30 inches .
FAHRENHEI T'S division of the scale , which makes 180 degrees between melting ice and boiling water , and places the point of 0 at the 32d degree below melting ice , may be retained ; because long use hath rendered it familiar to us .
A thermometer thus constructed , I shall call BIR D'S FAHRENHEI T : that eminent mechanic , our countryman , Mr. JOHNBIR D , having been , as I believe , the first workman , who took the pains to attend to the state of the barometer , in making thermometers , and having always made it his practice , to fix the boiling point , when his barometer hath stood at 30 inches .
Taking it for granted , that this scale will continue in general use here , I shall give an accurate comparison of it with the scale of M. DE LUC .
T being put for the height of a thermometer plunged in boiling water , above melting ice , in hundredths of a degree of M. DE LUC'S scale , in any given state of the barometer ; let ~ denote the same height , in hundredths of a degree of BIR D'S Fahrenheit .
Put y for the height of the barometer in 16ths of a Paris line .
v for its height in Paris lines .
x in 10ths of a Paris inch .
z in 10ths of an English inch .
And for 10387 put a. for 16 put b. for 10 put c. for 12 put d. Also put E and F for numbers expressing the proportion of the English foot to the French foot .
Now M. DE LUC hath found , that whatever be the value of y ; 9 9/ 2 00000 log .
y~= T. But log .
y= log .
v+ log .
b. and log .
v= log .
x+ log .
d log .
c. and log .
x= log .
z+ log .
E log .
F. Therefore log .
y= log .
z+ log .
E+ log .
d+ log .
b. log .
F log .
c. and 9 9/ 2 00000 log .
z+9 9/ 2 00000 log .
E. + log .
d+ log .
b log .
F log .
c-a=T .
But 9 9/ 2 00000 log .
E+ log .
d+ log .
b. log .
F log .
ca= 4171,55 .
the French foot being to the English as 2,1315 to 2 ; vide Phil. Trans. vol. LVIII .
Therefore 9 9/ 2 00000 log .
z 4171,55 = T. And 9 9/ 2 0000000 log .
z 41,7155 = T/ 100 = the height of the thermometer , plunged in boiling water , above melting ice , in degrees of M. DE LUC'S scale , when the height of the barometer , in tenths of an English inch , is z. Now for z write 300 .
Then T/ 100 = 80,902 .
which therefore is the height of the thermometer , in boiling water , above melting ice , in degrees of DE LUC'S scale , when the barometer is at 30 inches English .
And in the same state of the barometer , the height of the thermometer plunged in boiling water , above melting ice , in degrees of BIR D'S Fahrenheit , or , ~/ 100 , is 180 .
Hence the numbers T and ~ are in the constant proportion of 809 and 1800 , whatever be the value of z. For the change produced in the heat of boiling water , by any change of z , being always the same for both thermometers , the temperature expressed by T in parts of one scale , is always the same , as ~ expresses in parts of the other ; and , therefore , putting 1/ L , and 1/ B , for the values of the hundredth part of a degree , of the scales of M. DE LUC and BIRD respectively , the fractions T/ L , ~/ B are always equal , and T , ~ , are in the constant proportion of the invariable numbers L , B : consequently , when the proportion of T and ~ is determined for any particular value of z , it is found generally for all .
Therefore , as was affirmed , T : ~= 809 : 1800 .
And T=80 9/ 1 800~=89 9/ 2 000~ very nearly ( c ) in all values of z. and substituting this value , for T , in the equation exhibiting the relation between z and T , we shall have , for the relation between z and ~ , 99/ 20000000 log .
z 41,7155 = 899/ 2000x100 ~ .
Or , 99/ 10000 x 899 log .
z 92,804 = ~/ 100 = the height of the thermometer in boiling water , above melting ice , in degrees of BIR D'S Fahrenheit , when the height of the barometer , in 10ths of an English inch , is z. And thus M. DE LUC'S formula , for the variation of the boiling point , is adapted to English instruments , and reduced to English measures of length .
( c ) It might be sufficiently accurate for most purposes , to substitute 9/ 20 ~ ( = 81/ 180 ~ ) for 809/ 1800 ~ .
The error of this substitution would be about 1/ 809 ~ ; and consequently would amount to about 2/ 9 of 1~ of BIR D'S Fahrenheit , when z is 300 .
But the error in the substitution I have used is much less , not amounting to 1/ ~ ~ , which makes less than 1/ 44 of a degree of BIR D'S scale , in the same case .
For z write 287,7525 ( the length of 27 French inches in tenths of an English inch ) and ~/ 100 , the height of DE LUC'S boiling point , above melting ice , in degrees of BIRD'S Fahrenheit , comes out 177,989 .
Hence M. DE LUC'S boiling point falls upon 209,989 of BIRD'S scale ; that is , upon 210 very nearly , or insensibly more than two degrees below BIRD'S point of boiling ; and the reduction of either scale to the other , in all inferior temperatures , will be as the table of comparison shows .
By M. DE LUC'S formula , thus reduced , the height of the thermometer , plunged in boiling water , above melting ice , in degrees of BIRD'S Fahrenheit , in any given state of the barometer , may be computed .
But 899 being a troublesome divisor , to render the computation more easy and expeditious , take the following method .
For 11/ 1000000 log .
z. write s. Then s+ 1/ 900 s 92,804 =~ very nearly .
( d ) Upon these principles I have computed a little table , for finding the heights , to which a good BIRD'S ( d ) If , according to note ( b ) , we take 10369 instead of 10387 , for the value of the constant number a in M. DE LUC'S formula , we shall find , ~ log .
z 41,5355 = T/ 100 .
Whence we should obtain s 1/ 900 s 92,198 = ~/ 100 very nearly .
But I abide by the formula given in the text : being persuaded , that M. DE LUC hath purposely adopted the number 10387 , as agreeing better , upon the whole , with his experiments than the other ; though I do not recollect that he hath , in any part of his work , expressly said so .
Fahrenheit will rise , when plunged in boiling water , in all states of the barometer , from 27 to 31 inches English .
Among other uses of this table , it will serve for a direction to instrument-makers , to make a true allowance , for the effect of the variation of the barometer , if they are at any time obliged , to finish a thermometer , when the barometer is above or below 30 inches ; but in general it should be their rule , to watch an opportunity of fixing the boiling point , when the barometer is actually at the height prescribed .
I must , upon this occasion , declare , how heartily I concur with M. DE LUC , in wishing , that some common scale , the same in the number of its divisions , its point of 0 , and its boiling point , might be received , with unanimous consent , by philosophers of all parts of the world ; that , for the future , we might have one general language , for so very general an object of enquiry and discourse , as the different degrees of heat and cold .
To mathematicians the comparison of different scales , is a task of little labour .
But among those who have a taste for physical researches , and are capable of pursuing them , in certain branches at least , with some degree of advantage to science , as well as to themselves , there are many , to whom every little calculation is a toil .
Not having acquired the habits of it , in the early part of their lives , they never can acquire them , in such a degree , as to be able to put a confidence in themselves , and rely upon the accuracy of their own conclusions .
The ease of such persons should , in my opinion , be consulted .
But these are not the only persons interested .
The inconveniences to be apprehended from a diversity of scales are general .
It is not one of the least , that instruments essentially different , as made at different elevations in the atmosphere , will continue to pass under the same name .
The error and confusion , which this may create , is remarkably instanced , in what hath actually happened to the thermometer of the celebrated REAUMUR .
The instrument , which at this day passes , all over Europe , under the name of REAUMUR'S thermometer , is essentially different from his ; yet it was always supposed to be the same , even among his own countrymen , till M. DE LUC detected the mistake .
At what time , upon what occasion , or by whom , the change was introduced , is still unknown .
SECTIONSECON D. Of the GENERAL PRINCIPLES of measuring HEIGHTS by the BAROMETER .
AS it is my design to compare the practical rules , which M. DE LUC hath deduced from experiment , with theory , it is necessary , for the clearer arrangement of the argument , previously to state the general principles , upon which the measurement of heights by the barometer depends .
In doing this , I shall rather aim at perspicuity than at brevity ; referring , however , for the demonstration of whatever hath been advanced before , to writers of approved authority .
In any column of the atmosphere , resting perpendicularly upon any small portion of the earth 's surface , the densities of the air diminish , as we ascend to greater heights ; and if the accelerative force of gravity were the same at all heights , the densities would decrease geometrically as the height increased arithmetically .
This is an obvious consequence from a known property of air , that it expands itself through a greater or a less space , in proportion as the force , by which it is compressed , is less or greater ; that is , that the density of air , is always as the compressing force ( e ) .
And from hence it would follow , that the difference of the elevation of any two places , would be as the logarithm of the ratio of the densities of the air , at each : and the density being every where as the compressing force , and the compressing force as the length of the column of quicksilver sustained by it in the barometer , the difference of elevation would be , as the logarithm of the ratio of the altitudes of the quicksilver in the barometer , at the same time , at the different stations ; that is , as the difference of the tabular logarithms of the numbers , by which those altitudes would be expressed in any given measure ( f ) .
But the accelerative force of gravity diminishes , in the same proportion , as the square of the distance , from the earth 's center , is increased .
It is not the same ( e ) Cotes 's Hydrostat .
Lectures , lect .
ix .
( f ) See Philosoph. Transact. n. 181 .
Phil. Not .
Princip .
Math. lib. ii .
prop. 22 .
Scholium .
Cotes 's Hydrostat .
Lectures , lect .
ix .
Harmon .
Mens .
p. 17 .
Of all these demonstrations , that given by Mr. COTES , in the hydrostatical lectures , will be the most perspicuous to the generality of readers .
It is very diffuse , and he hath been at great pains to reduce it to the most simple principles .
Mathematicians will find the substance of his argument well summed up , by M. DE LA LAND , in the Connoissance for the year 1765 , p. 211 , 212 .
who , I am persuaded , would not have produced it as a new demonstration , had he known it had been given before .
therefore at all heights above the surface ; and the densities of the air must decrease , by a different law , from that which would obtain , if the force of gravity were uniform .
This other law , however , is such , that , to a much greater height than is accessible to man , the gradual variation of the compressing force and densities of the air , will be so little different , from what it would be upon the former hypothesis , that the error of that hypothesis , in the measurement of heights , will be absolutely insensible .
Let the point C [ tab .
IX .
fig. 1 .
] represent the center of the earth .
CA the earth 's semi-diameter .
AB any height above the surface .
At A , place a right line , AD , of any finite length , at right angles with AC .
In the right line AC , towards C , take A~ , such that CA may bear to A~ the proportion of CB to BA .
In a right line drawn through ~ , at right angles with AC , take ~E , of such length , as to bear to AD the proportion of the density of the air at B to the density at A , or at the earth 's surface .
The curve , which the point E always touches , is a logarithmic , of which AC is the asymptote ( g ) .
As I shall have frequent occasion to consider the curve , which thus exhibits the relation between density and elevation , I shall call it the ATMOSPHERICAL LOGARITHMIC .
Imagine this curve described , and take another height Ab , and take A~ = CA X Ab/ Cb , and draw ~e ( g ) Cotes 's Hydrostat .
Lectures , p. 161-167 .
Harm~n .
Mens .
p. 18 .
Phil. Not .
Princip .
Math. lib. ~i .
prop. 22 .
Brooke Taylor , Method .
Increment , Prop. 26 .
parallel to ~E , meeting the curve in e. Then ~ is the logarithm of the ratio of ~E to ~e , or of the density at B to the density at b. But if the greater of the two heights , AB and Ab , bear but a very small proportion to the semi-diameter of the earth , their difference Bb will be very nearly equal to ~ .
For , because CB : BA = CA : A~ ( by construction .
) Therefore , by conversion , CB : CA = CA : C~ .
In like manner , and by inversion , CA : Cb = C~ : CA .
by equi-distance perturbate , CB : Cb=C~ : C~ .
and converting , CB : Bb=C~ : ~ .
by permutation , Bb : ~=CB : C~ .
But when AB is infinitely diminished , CB = CA ultimately .
Also Ab being infinitely diminished , C~ = CA ultimately .
Therefore CB = C~ ultimately , and Bb =~ ultimately .
Q. E. D. Now AB and Ab will always be so small , with respect to CA , if B and b be supposed to represen any accessible places , that CB , C~ , and Bb , ~ , may always , in this case , be considered , as in their ultimate proportion of equality .
It is full therefore to be admitted as a principle , in practice , that the difference of elevation of any two places , is as the difference of the tabular logarithms of the heights of the quicksilver in the barometer at the same time , at both places ; that is , it is the logarithm of the ratio of those heights in some system of logarithms .
And the heights of the quicksilver being given , by observation , the difference of elevation will be known , if that particular system can be determined ; that is , if the modulus of the system , or the length of the subtangent of the curve DEe of that system , can be ascertained , in some known measure , as English fathoms , or Paris toises .
The easiest method of doing this , that theory suggests , is to compare barometers at two stations .
suppose B and b , each of a known elevation AB and Ab , above the level of the sea .
For the logarithms of any given ratio , in different systems , are proportional to the subtangents ; and the difference of elevation , Bb , diminished in the proportion of CB , ( the distance of the higher station from the earth 's center ) to C~ , ( a third proportional to Cb , the distance of the lower station from the earth 's center , and CA , the earth 's semi-diameter ) is the logarithm of the ratio of the density at B , to the density at b ( that is , of the columns of quicksilver sustained in the barometer at B and b ) in the atmospherical system .
Therefore , as the difference of the tabular logarithms , of these columns , to the subtangent of the tabular system , so should Bb , diminished as hath been said , ( that is , so should ~ ) be to the subtangent of the atmospherical logarithmic .
The utmost height , to which we can ascend , above the level of the sea , is so small , that the reduction of Bb may , even in this investigation , always be neglected .
For , if AB were four English miles , which exceeds the greatest accessible heights , even of the Peruvian mountains , and A~ three , ~ would be ~carce one part in 500 less than Bb .
So that , by comparing barometers at different elevations , within a mile above the level of the sea , the subtangent of the atmospherical curve might be determined , as it should seem , without sensible error , by taking simply the difference of elevation , without reducV O L. LXIV .
Hh tion , tion , for the logarithm of the ratio of the observed heights of the quicksilver in the atmospherical system .
Those , however , who have attempted to determine the system , by this method , have hitherto agreed but ill in their conclusions .
The fact is , that the length of this subtangent is very different at different times .
The causes and quantity of its variation will be considered in another place .
It appears , from M. DE LUC'S experiments , that , though otherwise subject to change , it is constant in a given temperature .
And that when the temperature of the air is + 163/ 4 of his scale , the difference of the tabular logarithms of the heights of the quicksilver in the barometer , gives the difference of elevation in 1000ths of a Paris toise ( b ) .
This is the rule , which he hath derived from a great number of experiments made at very different elevations : and the truth of it being admitted , it is a necessary consequence , that the number , which is the modulus of Briggs 's system , expresses the length of the subtangent of the atmospherical curve , such as it is in that temperature , in 1000ths of a Paris toise .
( h ) Recherch .
sir les Modi~ .
de l'Atmosph .
~ .
588 .
SECTIONTHIR D. Of the EQUATION for the DIFFERENCE of TEMPERATURE of the QUICKSILVER in the BAROMETERS , at the DIFFERENT STATIONS .
THE preceeding rule , however , must not be expected to give an accurate result , even in that particular temperature of the atmosphere , to which it is adapted , unless the specific gravity of the quicksilver , in the barometer , hath been the same at both stations , at the time of observation .
If the specific gravity hath been different , at the different stations , the heights of the quicksilver , in the barometer , will not have been proportional to the densities of the air ; that is , to the forces by which they have been sustained : and the calculation , built upon the supposition that they were so , becomes erroneous .
The specific gravity of every material substance varies with its temperature .
If the temperature of the quicksilver , therefore , hath been different , at the different stations , the difference of elevation , found by the foregoing rule , will require correction , though the mean temperature of the air may have been such as it prescribes ( i ) .
No particular temperature of the quicksilver is necessary to the accuracy of the result of the preceeding rule , or to render the ( i ) Observe , that the temperature of the quicksilver , in the portable barometer , will not be necessarily the same with that of the circumambient air , at the place and time of observation ; there will generally be a considerable difference .
correction , I am now considering , 0 ; but only , that its temperature , whatever it be , be the same at both stations .
If the temperature hath been the same , the specific gravity hath been the same ; and if the specific gravity hath been the same , the length of the columns of quicksilver have been as the forces by which they were sustained , whatever the common specific gravity may have been .
Columns of water , sustained in evacuated tubes , of sufficient length , would be proportional to columns of quicksilver in the barometer , at the same times and places , provided the temperature of the two water columns were the same , and that of the two mercurial columns the same ; and , consequently , the difference of the logarithms of the water columns would be precisely the same as of the mercurial columns .
For the water columns , and the mercurial columns , are only expressions of the same absolute magnitudes , the forces by which they are both sustained , in parts of different scales .
But if the temperature of the quicksilver , or of the water , be different at the same time , at different stations ; then , though we compare the water columns with each other , and the mercurial columns with each other , still we compare different things , though we call them by the same name .
We compare fluids of different specific gravities ; that is , we measure the pressure of the air at one place , in parts of one scale , and at the other place , in parts of another .
The error is of the same kind , as if one should attempt to determine the proportion between two parts of a building , by measuring one with a Paris foot-rule , and the other with an English foot-rule , without attending to the difference between Paris feet and London feet ; and the same method must be taken to get rid of the error in both cases : we must ascertain the difference of the scales we have applied , and make allowance for it : we must know the length of the one in parts of the other ; that is , in the particular case in question , we must be able to determine , from the observed height of the barometer , when the quicksilver is of any given temperature , what its height would have been , at the same place , and at the same time , if its temperature had been any other , that may be assigned .
I have thought it necessary to be thus minute , in explaining the principles , upon which the correction in question depends ; because it is a point , which is likely to be misunderstood , though in itself of no great difficulty .
M. DE LUC himself hath fallen into a mistake , not with respect to the quantity of the error , but the manner of allowing for it ; which , however , is of no other bad consequence , than that of lengthening the calculation unnecessarily .
He imagines that a particular temperature of the quicksilver is necessary , that the error should be nothing ; and to this he always reduces the observed heights of the quicksilver , in the barometer , at both stations ( k ) .
But the result of the computation would , in all cases , have been the same , if they had been reduced to any other given temperature ; and therefore , it is always sufficient , to reduce the one to the temperature of the other .
This little oversight I should have touched upon with more ( k ) Recherch .
sir les Modif .
de l'Atm .
~ .
369-374 .
reluctance , were it not of too little importance to derogate , in the least degree , from the general merit of M. DE LUC'S elaborate and invaluable work .
The quantity of the correction in question , is thus determined .
At times , when the barometer , in a temperate air , stood at 27 Paris inches , M. DE LUC tried what change was effected in its height , by changes of temperature , induced by art in the quicksilver , without any alteration in the state of the atmosphere ; and , by repeated experiments of this kind , it was found , that the difference between the length of the column of quicksilver , successively heated to the temperature of boiling water , and cooled to that of melting ice , amounted to half a Paris inch exactly ; that is , to 1/ 54th of the mean height ( l ) .
But the whole extent of the thermometer 's scale , from the temperature of melting ice to M. DE LUC'S boiling point , being 178~ of BIRD'S Fahrenheit , the change of the height of the barometer , due to 1~ of BIRD'S Fahrenheit is always 1/ 54 X 178 = 1/ 9612 .
And if n denote the number of degrees of BIRD'S Fahrenheit , in the difference of the temperatures of the quicksilver , in the barometers , at different stations , the reduction will be n/ 9612 ; that is to say , the height of the warmer column must be shortened by this part of its own length , or that of the cooler augmented by the like part , to reduce either , to the height it would have stood at , at the time and place of observation , with the temperature of the other .
( l ) Recherch .
sir les Modif .
de l'Atm .
~ .
362-364 .
These determinations of the effect of heat upon the column of quicksilver , in the Torricellian tube , show the proportional alterations of the density of that fluid , by given increments , or diminutions , of heat .
For the perpendicular height of a column , of any inelastic fluid , sustained in the Torricellian tube , by a given compressive force , must , by the known laws of hydrostatics , be as the densities of the fluid inversely .
And as this proportion must obtain , whatever be the size , or figure , of the tube , there seems to be no method , by which the change of density , or the proportional expansion of quicksilver by heat , can be measured with more precision .
These conclusions , therefore , may be of use in many physical enquiries ; and there are many cases , in which it may be necessary , to reduce the observed height of the barometer , in one temperature , to another .
Thus when that height is to be made the measure of the variable pressure , or of the density of the air , in some particular place , it will be necessary to choose some standard temperature , to which the observed length of the column may always be reduced .
And it was this consideration , as I have gathered from many conversations with him , which gave occasion to M. DE LUC'S mistake .
He had settled it with himself , at his very first entrance upon these researches , that the point he was to keep constantly in view , as the ultimate object of his whole pursuit , was to find , in the variable length of the mercurial column , a measure of the pressure and density of the air .
This , he saw , was only to be looked for in quicksilver of some one constant temperature ; and thus he became possessed with a general notion , of the necessity of reducing the observations of the barometer , to some standard temperature , upon all occasions .
But , for the particular purpose of computing differences of elevation , instead of attending to the correction of the observed heights of the quicksilver at all , it will be a readier way , to make the corresponding correction immediately , upon the difference of the tabular logarithms .
If any quantity be diminished or increased , in a given proportion , the logarithm of its proportion to any other given quantity ( that is , the difference of the tabular logarithms ) is diminished or increased by a given magnitude ; namely , by the logarithm of the given ratio , in which the variable quantity is altered .
The logarithm of the ratio of x to x+ 1/ 9 612 x , or of 96 12 , to 96 13 , is 452 ( m ) .
Therefore , if n , as before , express the difference of temperature in degrees of BIRD'S Fahrenheit , ~X452 is the correction to be applied to the difference of the logarithms of the observed heights of the quicksilver in the barometer .
This correction is to be substracted , if the temperature , at the upper station , hath been the cooler of the two , and to be added , in the contrary case .
I am indebted to the ASTRONOMERROYAL for the first hint of this elegant method of applying the correction in question .
( m ) The reader will here recollect , that I speak of the tabular logarithms in the language of M. DE LUC .
The logarithm of the ratio of 9612 to 9613 is , in the more common stile , 0.0000452 .
The third of the annexed tables , shows the quantity of this correction corresponding to every value of n from 1~ to 70~ .
The third column , to the right , exhibits the value of it in English fathom , in that particular temperature of the air , in which the difference of the tabular logarithms .
of the heights of the quicksilver in the barometer , gives the difference of elevation in 1000ths of an English fathom .
What that temperature is , will be determined in the sequel .
It is to be observed , that the value of these corrections , in the fathom of any other country , will be the same , in that particular temperature , in which the tabular system measures the difference of elevations in 1000ths of the fathom of that country .
The rule , which I have given for applying this correction , supposes that it is previously known , which of the two stations is the highest ; otherwise it will be doubtful , whether it should be added or subtracted .
This doubtful case may actually happen in levelling a considerable tract of country , that is pretty even .
The table , however , will still give the quantity of the correction .
Add that quantity to the logarithm of the height of the cooler barometer ; and the difference between this logarithm , so augmented , and the logarithm of the height of the other , is the difference of the logarithms of the observed heights , duly corrected : and the station of the cooler barometer , was the lower of the two , if its augmented logarithm exceeds that of the other : in the contrary case , the higher .
It hath been already observed , that the change of the density of quicksilver , by an alteration of its heat , must always be proportional to the increase or contraction of the length of the column in the barometer ; or , in other words , that the alteration , in the perpendicular height of the Torricellian column , is proportional to the change of the solid content of space , occupied by a given quantity of the fluid .
Those who recollect the experiments of the celebrated BOERHAAVE , for measuring this change of volume , cannot but be struck with the great and singular disagreement between his conclusions and those of M. DE LUC .
BOERHAAVE makes the whole expansion , produced by a change of temperature from the 0 of Fahrenheit , ( or the forced congelation of sal ammoniac ) to the heat of boiling water , 1/ 5225/ 53 of the whole volume , which is very little more than M. DE LUC found to be produced by a change from 32~ of Fahrenheit ( or the dissolution of natural frost ) to the heat of boiling water(w ) .
( n ) Vid. Elementa Chemiae , vol. 1 .
p. 174 .
In this place , the author expressly treats of the expansion of quicksilver by heat ; and the proportion , which he here assigns , hath been adopted by the writers of elementary physics .
But , p. 165 , having described an artificial cold , in which the thermometer had been sunk to 40 of Fahrenheit , he says , ` ` Novimus hoc argenti vivi corpus ab illo gradu 40 infra 0 , ad gradum 600 auo incipit ebullire contractum fuisse per parts 640 totius molis 10782 .
' ' And , in the explanation of one of his copper plates , he describes a thermometer in which the quicksilver , in the greatest natural cold , just filled the bulb , which contained , he says , 11520 such parts as the tube contained 96 .
It seems probable , that the number 11124 , which is given , p. 174 , as the whole volume of the quicksilver in the temperature 0 , ( a degree of Fahrenheit being the unit ) hath been deduced from the mean of a variety of experiments upon different instruments , for it differs not greatly from the mean of the two numbers 10782 and 11520 .
Upon this it is to be observed , that BOERHAAVE'S method gave him the expansion of his quicksilver , diminished by the whole expansion nearly of the vessel which contained it .
DE LUC'S gave him the expansion of his quicksilver , diminished by the longitudinal expansion of the scale applied to the barometer , by which the height of the column was measured .
The vessel which contained BOERHAAVE'S quicksilver was glass .
The scale of DE LUC'S barometer was deal .
Glass is considerably altered in its dimensions by heat ; deal but very little , in the direction of its longitudinal fibres : therefore , though M. DE LUC neglected the expansion of his scale , the expansion which he observed of the quicksilver would be nearly the whole ; but BOERHAAVE must have reckoned it much too small , neglecting , as he says he did , the expansion of his glass ( 0 ) ; and the difference is no more than what the expansion of the glass will very nearly reconcile .
BOERHAAVE'S expansion is too small , to be in proportion to M. DE LUC'S , by about 1/ 2 86 of the whole volume , or somewhat more than 1/ 3 0 0 .
The expansion of glass , in length , by 180~ , is , as I am informed , about 1/ 1 00 of an inch upon a foot , or 1/ 1 200 part ; therefore , its expansion in solid content , or volume , will be very nearly 1/ 4 0 0 , by 180~ ; and , by 212~ , about 1/ 3 3 1 , or less than 1/ 3 0 0 .
( 0 ) Seposita engo vitri interea dilatatione , &amp;c. These words may seem ambiguous ; but the meaning seems sufficiently determined by what is said , p. 173 .
l. 2-5 .
Ii2SECTIO N SECTIONFOURT H. Of the EQUATION for the TEMPERATURE of the AIR .
IF the temperature of the atmosphere hath been any other than is expressed by + 163/ 4 of M. DE LUC'S scale , the result of the calculation , formed upon the preceeding rules , requires a further correction .
This correction arises from a variation of the length of the subtangent of the atmospherical logarithmic ; which , as hath been already remarked , is found to be not constantly the same .
That this matter may be the better understood , it will be proper to state , in this place , what this subtangent is , in the nature of things , and upon what physical circumstances its length depends .
Imagine then , that instead of the atmosphere , in its natural state , the earth were surrounded with an inelastic fluid , of an uniform density throughout , equal to that of the natural atmosphere , in its lowest parts , which are contiguous to the surface of the earth ; imagine also , that every atom of this homogenous fluid were urged towards the earth 's center , with an accelerative force , equal to that of gravity , at the surface of the earth .
Now the pressure of the atmosphere upon the earth 's surface , or on any given part thereof , being , at all times , a finite , though not a constant force , of which the phenomenon of the Torricellian tube are a sufficient proof , it is evident that it might , at any time , be equalled by the pressure of a finite quantity of this imaginary fluid ; and that , to render the pressure of this fictitious atmosphere , upon the whole , or any part of the earth 's surface , equal to that of the real atmosphere upon the like part , it would be requisite to assign to it some finite thickness or depth .
Now that particular thickness or depth , of the imaginary fluid , which is , at any time , necessary to render an entire column of it a counterpoise for an entire column of the natural air , is , at that time , the length of the subtangent of the atmospherical logarithmic .
In short , the subtangent of the atmospherical logistic , is the length of a column of such a fluid as I have supposed , which would be sustained in the Torricellian tube , by the pressure of the air , at the level of the sea , if we could suppose a tube of a sufficient length .
This is demonstrated by Mr. COTE S , Harmon .
Mens .
p. 18 .
and by no one else , that I know of , with equal simplicity .
It is a manifest consequence from this , that the subtangent must always be as the pressure of the whole cylindrical column ( upon a given part of the earth 's surface ) directly , and the density of the air , at the surface , inversely ; and it may therefore appear to be repugnant to the theory already established , to suppose it subject to variation .
For if the pressure be always as the density , upon which hypothesis the whole theory is founded , that which is always as the pressure directly , and the density inversely , can be no other than a constant quantity .
But M. DELUC 'S experiments prove , beyond a doubt , that the atmospherical subtangent is variable ; therefore the density of the atmosphere , at the surface of the earth , at different times , is not proportional to the whole pressure , at such times , respectively .
And in this there is nothing inconsistent with the foregoing theory , rightly understood .
When the density of the air is said to be as the compressing force , this is to be understood of air in the same state of elasticity ; that is , in which the absolute force of elasticity , under all the different degrees of pressure , is constantly the same .
( The absolute force of elasticity is measured by the force exerted between two particles of given magnitudes and figures , at a given distance .
) But there is no reason , from any experiments , to conclude , that the density of the air will be simply as the compressing force , in different absolute forces of elasticity .
On the contrary , since the proportion of the compressing forces is found to obtain , between the densities , when the absolute elasticity is constant ; this alone is a proof , that the like proportion will not obtain , if the absolute elasticity be changed as well as the compressive force .
Imagine two parcels , similarly shaped , A and B , of different fluids , the same in all their other properties , and similarly , but unequally , elastic ; that is , imagine the integrant particles of A to be equal in quantity of matter and bulk , and similar in figure , to the integrant particles of B , severally ; and if each of the integrant particles of A , be itself an aggregate of lesser integrants , each of the integrant particles of B is to be understood to be a like aggregate of equal and similar lesser integrants , similarly composed ; and whatever forces , except that of elasticity , are exerted between the particles of A , imagine equal forces , of the same kinds , exerted according to the same laws , between the particles of B. Likewise , imagine the equal and similar particles to be similarly situated , in their respective masses .
Such fluids are the same in all respects , elasticity excepted .
Further , imagine an elastic force to be exerted between every two integrant particles of A , and to vary , with the distance , as any power or function of the distance whatever .
Imagine an elastic force , varying according to the same power or function of the distance , exerted between every two particles of B. Thus the fluids are similarly elastic ; but they are to be supposed unequally so ; that is , the force exerted between two particles of the one , at any given distance , is to be greater than the force between two particles of the other , at the same distance .
Now , if the law by which the elastic force varies , at different distances , in one of these fluids , as A , be such , that the densities of A , under different compressions , are to each other , as the compressive forces ; the densities of B , compared with each other , under different compressions , will likewise be as the compressive forces .
But the proportion of the compressive forces will not subsist between the densities of the two different fluids .
The densities will not be the same , when the compressive forces upon A and B are equal ; nor will they be as the compressive forces , when those forces are unequal .
The densities will be as the compressive forces directly , and the absolute elasticities inversely ; that is , the compressive forces will be as the densities and absolute elasticities jointly , which is easily demonstrated .
The compressive force upon any fluid , is to be estimated by the quantity of the pressure , acting perpendicularly upon a given surface , or plane section of it .
Whatever degree of pressure acts upon the mass of an elastic fluid , its particles will approach or recede , till the distances of every two adjacent particles are such , that the whole elastic force , exerted perpendicularly-against every surface or plane section of it , is precisely equal to the pressure acting perpendicularly upon the same surface , in an opposite direction .
And in the particular degree of density , determined by that distance of the adjacent particles , the fluid will remain , while the same degree of pressur is continued .
Imagine therefore the densities of the two masses , A and B , to be the same ; then , from the supposed similarity of the fluids , it follows , that the number of particles , exerting their elasticities upon any equal and similar sections of A and B , must be equal ; and that the distances of the corresponding particles , from each other , in the two masses , must be the same : consequently the whole elasticities , exerted upon equal and similar sections of the two masses , will be as the absolute elasticities of which they are composed .
Therefore the compressive forces are , in this case , as the absolute elasticities .
Call the common density of the fluids D ; the compressive force upon A , P ; upon B , II ; the absolute elasticities a , b , respectively .
Let d denote some other density of the mass A , and p the compressive force corresponding to that density .
Now D and d are different densities of the same fluid A , under different compressive forces P , p. Therefore , p : P= d : D ( by hypoth .
) But P : II = a : b. ( as hath been proved .
) Therefore , p : II =dX a : DX b ; that is , the compressive forces upon the different fluids A and B , when the densities are unequal , are as the densities and absolute elasticities jointly .
Q. E. D. This demonstration is independent of any more particular hypothesis , concerning the law of the elasticity , than barely that it is the same in both masses ; and such , in both , as to make the densities of either always proportional to the forces by which it is compressed .
But to this condition , a particular law of elasticity is requisite ; namely , that the force exerted between every two adjacent particles diminish as the distance between them is increased ; and , adopting this law , the compressive forces might be proved to be in the proportion assigned , that of the densities and absolute elasticities jointly , by the same kind of reasoning as is used to demonstrate the 23d proposition of the second book of the Principia , which is only the most simple case of the more general theorem now proposed .
Hence it follows , that if any causes act upon the atmosphere , to change the degree of its absolute elasticity , provided they act in such manner , as to change it equally at all heights ; so that , though different at different times , it shall always , at any one time , be the same at all different heights ; then , the densities in all different parts of the cylindrical column , resting perpendicularly upon a given small part of the earth 's surface , will , at any one time , be as the compressive forces upon each part respectively .
But the densities , in any one part of this column , at different times , when the absolute elasticities are different , will not be as the compressive forces upon that part , at such different times respectively .
Hence the relation between the decrement of density , and the increment of height , reduced according to the construction of section second , will , at every time , be represented , by the decrement of the ordinates , and the increment of the asymptote of some logarithmic curve .
But of different curves , at different times , or in different states of the absolute elasticity ; that is , the length of the subV O L. LXIV .
Kk tangent tangent of the curve , described in section second , will vary : for the difference of the subtangents is the only thing that constitutes a real difference in two logarithmic curves .
If such curves have equal subtangents , though they have unequal ordinates , they are only different parts of the same curve .
The subtangent of the atmospherical logistic must always be as the pressure of the whole column of the atmosphere directly , and the density , at the earth 's surface , inversely ; therefore it is directly as the absolute elastic force .
For call the subtangent S ; the pressure at earth 's surface , P ; the density , D ; the absolute elasticity , A. Now that S is as P/ D is obvious ( from p. 245 .
) ; but P/ D is as A ( by what hath now been proved ) .
Therefore S is as A. Now heat ( p ) is one cause , which is well known to influence the absolute elastic force .
An increase of heat increases elasticity ; and elasticity is diminished by a diminution of heat .
Accordingly M. DELUC 'S experiments show , that when the temperature of the air is uniform ( or the same at all heights ) the subtangent of the atmospherical curve is increased , or diminished , exactly in proportion to the increment or decrement of the uniform temperature , as indicated by the mercurial thermometer .
His conclusion , from repeated experiments on the mountains near Geneva , is , that , if L denote the difference of the tabular logarithms of the heights of the quicksilver , at two stations , corrected for the difference of the temperature of the quick(p ) BOERHAAVE , Elementa Chemiae , vol. i. p. 456 , &amp;c. silver , and n denote the difference of the temperature of the air , in degrees of his own scale , above or below + 16 3/ 4 , then L~ n/ 215 L , is the difference of elevation in 1000ths of a Paris toise .
The correction n/ 215 L , to be added or subtracted , according as the temperature of the air is above or below 16 3/ 4 ; or according as n is a difference in excess or defect ( q ) .
But in any given value of n , the proportion of n/ 215 L to L is given , being that of n to 215 .
Therefore , L~ n/ 215 L is the logarithm of the ratio of the heights of the barometer , in a system of logarithms , in which the subtangent of the curve is to so many 1000ths of a Paris toise , as are expressed by the subtangent of the Briggian system , as 215 ~n to 215 .
And this being the case , whatever be the value of L , in every given value of n , it is evident , that the variation of the modulus ot the atmospherical system , or of the subtangent of the atmospherical curve , is always as n , the variation of temperature .
But a uniform temperature is a condition of the atmosphere , which rarely obtains , within small distances , at least , above the earth 's surface ; therefore the more usual case of unequal temperatures must be considered .
When different temperatures obtain , at different heights , at the same time , they must render the absolute elasticities , at those heights , unequalThus the hypothesis of M. BOUGUER will take place ; ( q ) Recherch .
sir les Modif .
de l'Atm .
~ .
607-611 .
who , from a great number of experiments , made upon the CORDILEROS , and at various other heights above the level of the sea , concluded , that the absolute elasticity of the air , which he calls , ` ` the intensity of the elastic force , ' ' must be different at different elevations .
His conjectures indeed about the cause from whence the difference might arise , are not the most natural ; and in one point he was evidently mistaken : namely , that he imagined the absolute elasticity to be constant at every given elevation ; and accordingly he hath traced the curve , which , according to his experiments , exhibits the laws of its variations from one height to another .
This , it must be confessed , was trusting his expements too far , which were not made with the most accurate instruments ; and this may have given occasion to his learned countryman M. DELALAND E , to suppose that this curve may rather exhibit the deviations of his experiments from the truth , than any thing which really obtains in nature ( r ) .
His general hypothesis , however , of a difference of absolute elasticity , in different parts of the atmosphere , must obtain , whenever the temperatures of such different parts are unequal ; unless the effect of the inequality of temperature were to be compensated by the synchronous unequal operations of some other cause .
That it does in fact obtain , when the temperature is unequal , is proved , beyond a doubt , instead of being refuted , by M. DELU C'S experiments ; and this is to be considered , as a further instance of their exact agreement with the genuine conclusions of accurate theory .
( r ) Connoissance pour l'annee 1765 , p. 215 .
When the absolute elasticities are different at different heights , the densities will no longer be proportional to the pressures ; and the change of density , throughout the whole column of the atmosphere , will no longer be represented by any one logarithmic .
But to small distances on one side or the other of different heights , it may still be nearly represented by parts of different logarithmics .
Now this is really the case , according to M. DELU C : for , when the temperature , at two different heights , hath been different , he finds , that the difference of those heights will still be expressed in 1000ths of a Paris toise , by the preceeding formula , viz. L~ n/ 215 L , if n be understood to denote the difference between the constant temperature + 16 3/ 4 , and that which is the mean of the different temperatures of the two places of observation ( s ) ; that is to say , though the temperature of that portion of the column of the atmosphere , which is intercepted between the level of the two places of observation , hath not been the same , perhaps , in any two different parts ; yet the variation of density and height will be exhibited , through the whole of this small space , without sensible error , by the curve , which would have represented them strictly , if the entire column had been of the mean temperature of the extremes of this portion of it ; namely , by the logarithmic , whose subtangent is expressed in 1000ths of a Paris toise , by B~ n/ 215 B , B denoting the subtangent of the Briggian system .
Imagine , therefore , that the barometer and thermometer have been observed at ( s ) Recherch .
sir les Modis .
de l'Atm .
~ .
663 .
the same time , at three different heights .
Put the difference of the tabular logarithms of the heights of the quicksilver , at the 1st and 2d elevation L ; at the 2d and 3d , L ' ' .
Imagine the thermometer at the 1st , or lowest height , to have been + 16 3/ 4 ; at the 2d , 16 3/ 4 +2 n ; at the third , 16 3/ 4 2 m. Then the diff. between 1st and 2d elevation , =L+ n/ 215 L between 2d and 3d , = L ' ' + n-m/ 215 L ' ' that is , the variations of height and density between each two of these three places , are nearly exhibited by two different logarithmics , the subtangents of which are B+ n/ 215 B ; B+n m/ 215 B ; therefore , the variations of height and density , throughout the atmosphere , are not represented by any one logistic , when the temperature is unequal ( as they would be , if , notwithstanding an inequality of temperature , the absolute elasticities were the same in all parts ) ; but by parts of different logistics , at different heights ; as they should be when the absolute elasticities are different , in different parts , and can only be considered as uniform to a small distance , above or below any height assigned .
The mention of M. BOUGUER occurred so naturally upon this occasion , that I must have reproached myself with an impiety towards the ashes of a man , whose memory will ever be dear to science , had I not attempted to vindicate his conjectures upon the point under immediate consideration , from an animadversion thrown out , in very general terms , with little tenderness , and only just in part .
At the the same time , I have pointed out , in what his principal mistake consists ; which was , as M. DELUC well observes , that , instead of seeking a rule which should be general for all heights , and vary at every height , in proportion to the temperature , he set himself to find one which should be general for all temperatures , at a particular elevation .
The consequence likewise , which he would have deduced from the unequal elasticity of the particles of the air , that the least elastic would be driven to the bottom , was undoubtedly erroneous .
The excess of temperature may fall sometimes in the lower parts of the atmosphere , and sometimes at greater heights ; and where the greater temperature is , there , caeteris paribus , the elasticity will be greater .
The argument , which M. DELUC directs against the existence of sensible inequalities of elastic force in aggregate masses of the atmosphere , derived from the supposed effect of the winds , throughout all the regions exposed to their agitations ( t ) , militates only against the probability of permanent inequalities in given places , arising from supposed specific differences in the original constitution of the particles of the air ; not against such temporary inequalities , as we ascribe to the occasional energy of extraneous causes .
An inequality of temperature undoubtedly exists , in aggregate masses , more frequently than the opposite .
And from an inequality of temperature , whether in the aggregate or the discrete , necessarily follows , for the time , an analogous inequality of absolute elastic force .
( t ) Recherch .
sir les Modif .
de l ' Atm .
~ .
328 .
Having now sufficiently explained , what the correction is , for a variation of the temperature of the air , and whence it arises , I proceed to reduce M. DE LUC 's formula to BIRD 's Fahrenheit , and a scale of English fathom .
SECTIONFIFT H. M. DE LUC'S RULES reduced to ENGLISH SCALES .
THE whole of this reduction I divide into three problems .
PROBLEMFIRS T. To find the length of the subtangent of the atmospherical curve , in thousandths of a Paris toise , the mean temperature of the air being given in degrees of BIRD'S Fahrenheit .
BY the comparison of M. DE LUC'S scale with BIRD'S Fahrenheit , it appears , that + 16 3/ 4 of the former corresponds to + 69,25 of the latter .
Hence 69,25 is the temperature in BIRD'S Fahrenheit , in which , the subtangent , of the atmospherical curve , is equal to so many 1000ths of a Paris toise , as are expressed by B , the subtangent of the Briggian system .
But the atmospherical subtangent is increased or diminished by 1/ 215 of this quantity for every degree of M. DE LUC'S scale , above or below this given temperature ; and a degree of M. DE LUC'S scale is to a degree of BIRD'S Fahrenheit as 178 to 80 .
Therefore , the subtangent varies by 8/ 3827 of the same quantity , for every degree of BIRD'S Fahrenheit , above or below the given temperature .
Hence , if n denote the difference of the temperature of the air , or , in the case of unequal temperatures , the difference of the mean of the temperatures of the two stations , above or below 69,25 in degrees of BIRD'S Fahrenheit ; then , B~nx 8/ 3827 B is the length of the subtangent in 1000ths of a Paris toise : that is , the subtangent of the atmospherical curve , in the temperature 69,25 ~ n , is to so many 1000ths of a Paris toise , as are expressed by the modulus of the Briggian system , as 3827 ~nx8 to 3827 .
PROBLEMSECON D. To determine the temperature , in which the length of the subtangent of the atmospherical curve is expressed in thousandths of an English fathom , by the subtangent of the Briggian system .
FROM the number 69,25 subtract the 8th part of the number , to which 3827 bears the proportion of 10000 to 617 ; that is , from 69,25 subtract 29,51 ; the remainder 39,74 expresses the required temperature , in degrees of BIRD'S Fahrenheit .
For , let S , ~ represent the subtangents of the atmospherical curves in the temperatures 69,25 and 69,25-29,51 , respectively .
Then S : ~= 3827 : 3827 29,51 x8 ( by prob .
1 .
) But 3827 : 3827 29,51 x8= 10000 : 10000 617 = 10000 : 9383 .
Therefore , S : ~= 10000 : 9383 ; that is , as one Paris toise to one English fathom .
Therefore , whatever multiple S is of the Paris toise , or any part thereof , the same multiple is ~ of the English fathom , or its like part .
And the length of S is expressed in 1000ths of a Paris S~ toise , by the number which is the modulus of the Briggian system ( by prob .
1 .
) ; therefore , the length of ~ is expressed by the same number , in 1000ths of an English fathom .
PROBLEMTHIR D. To find the equation for every degree of BIRD'S Fahrenheit in the mean temperature of the air , above or below 39,74 .
CALL the variation of the length of the subtangent , corresponding to an increment or decrement of one degree of BIRD'S Fahrenheit , V ; and let S , ~ , as before , represent the subtangents corresponding to the temperatures 69,25 and 39,74 .
Now V : S= 8 : 3827 ( by prob .
1 .
) And S : ~= ( 3827 : 3827 29,51 x8 = ) 3827 : 3591 very nearly ( by prob .
1 .
) Therefore , V : ~= 8 : 3591 .
that is , V= 8/ 3591 ~ .
( = 1/ 449 ~ nearly ) .
Hence the length of the subtangent of the atmospherical curve , in any temperature , 39,74 ~n is to its length in the temperature 39,74 as 3591 ~8xn to 3591 ; that is , putting B for the subtangent of the Briggian curve , B~nx 8/ 3591 B is the length of the subtangent of the atmospherical curve , in 1000ths of an English fathom .
And putting L for the difference of the tabular logarithms of the observed heights of the barometer , at two stations , corrected by the equation for the temperature of the quicksilver , and n for the mean difference of the temperatures of the air , in degrees of BIRD'S Fahrenheit , above or below 39,74 , the difference of the elevation of the two stations , is expressed in 1000ths of an English fathom by L~nx 8/ 3591 L. Upon these principles I have made a table , by which the equation nx 8/ 3591 L may be computed for any temperature not above + 80 , nor below 0 of BIRD'S Fahrenheit , and for any height less than 10000 fathom .
It hath already been remarked , that the temperature of the quicksilver , in the portable barometer , may happen to be very different from that of the air , at the place and time of observation .
For this reason , M. DE LUC advises , that every portable barometer should be furnished with two thermometers ; one fixed to the frame of the barometer , to indicate the temperature of the quicksilver ; another , to be exposed , at the time of observation , to the open air .
The use of the annexed tables might be answered by a particular division of the scales of these thermometers .
The thermometers which accompany the two barometers , made for the society , by Mr. EDWARD NAIRNE , under my directions , are BIRD'S Fahrenheits ; but , to the fixed thermometer , I have had a particular scale applied , in which the fundamental interval , between melting ice and BIRD'S boiling point , is divided into 81 equal parts .
The equation for the temperature of the quicksilver is one fathom for every degree , of this scale , in the difference of temperatures .
The thermometers , for the temperature of the air , have each a scale , in which the fundamental interval between melting ice and BIRD'S boiling point , is divided into 120 equal parts .
The point of o is placed at the 5th of these divisions above melting ice .
If n be the mean height of the thermometer , in degrees of this scale , at the two stations , the equation for the temperature of the air is n/ 300 L. + or - , according as n is positive or negative .
The place of the point of o upon the former scale is indifferent .
It was put very low , that the temperatures of the quicksilver , at both stations , might always be above it .
The computation , however , is rendered easier by the tables now given , than it can be by any such peculiar divisions of the scale .
It is to be particularly observed , that M. DE LUC always exposed the thermometer , by which he measured the temperature of the air , to the sun , if it happened to shine ; but then the ball of his thermometer was always quite detached from the frame , which is a necessary precaution in this manner of using it .
If the temperature of the air were measured by a thermometer , exposed in the shade , I cannot but think the quantity of the equation would be different .
His reason for not shading his thermometer from the sun , was this : when one side of the thermometer is exposed to the sun 's rays , the other is in its own shade ; therefore , in this situation , he thought it the just measure of the mean temperature of the air , which is neither the temperature of that part on which the sun 's rays fall , nor of that from which they are intercepted , but less than the one , and greater than the other , or a mean between the two(u ) .
I confess , I should have expected an irregularity , from heat excited by the rays of light , in their passage from the air into the glass , and from the glass into the quicksilver ; and should therefore have exposed my thermometer in the shade ; but the success of M. DE LUC'S experiments seems sufficiently to justify the method he hath taken .
I shall close this section with a brief solution of two physical problems , for which this seems the proper place .
PROBLEMFOURT H. To compare the densities of the air , at any given elevation above the surface of the earth , in different temperatures .
IF the barometer hath been observed , at any given elevation above the earth 's surface , in different temperatures , reduce the observed heights of ( u ) Recherch .
sir les Modif .
de l'Atm .
~ .
533-536 .
the quicksilver to a common temperature , if the temperatures of the quicksilver , as well as the air , have been different , at the different times of observation .
Find the subtangent for each temperature of the air ( by problem 3d ) ; divide the corrected heights of the quicksilver by the subtangents corresponding to the observed temperatures of the air .
The quotients are as the densities in these temperatures respectively ; that is , calling the heights of the mercury reduced , P , II ; the temperatures , T , ~ ; the subtangents , S , ~ ; and the densities , D , ~ , D : ~= P/ S : II/ ~ .
For the densities are as the compressive forces directly , and the absolute elasticities inversely ( by section 4th ) ; and the absolute elasticities are as the subtangents ( by section 4th ) ; whence the truth of the rule is manifest .
The most convenient method , however , for practice will be , to make the density of the air , in some given state of the barometer and thermometer , a standard , with which to compare the densities in all other states of these instruments .
Suppose , for instance , at the level of the sea , 30 inches be taken for the standard height of the barometer , and + 40 for that of the thermometer .
Let D be the density of the air , at the level of the sea , when the barometer is 30 inches , and the thermometer + 40 .
Put P= 30 inches , and S= the subtangent of the logistic corresponding to the temperature + 40 .
Now in another temperature 40 + n , let the subtangent be S+ b/ m S. And let the height of the barometer P be changed into P+ a/ r P ; and let ~ be the density at the level of the sea , in the temperature 40 + n , and height of the barometer , P+ a/ r P. Then the difference between ~ and D , or ~ D , will be Dxmab r/ r xm+ b. Where observe , that b is positive or negative , according as n is positive or negative .
By this problem , the densities of the air at a given height , in different temperatures , are to be compared with each other .
Even good writers have hitherto been generally run away with by a notion , that the air 's density , at a given elevation , at different times , would be always truly measured by the length of the mercurial column ; than which nothing can be more erroneous , as hath been at large explained in the 4th section .
PROBLEMFIFT H. The height of the quicksilver in the Torricellian tube , and the temperature of the air being given , at a given elevation , above the level of the sea , to compare the density of the air with that of the quicksilver , at the time and place of observation .
LETTA represent the semi-diameter of the earth ; T being the center , and A at the surface .
Let B be a place at a given height , A B , above the surface .
Let CD be the observed length of the quicksilver in the Torricellian tube , at the place B , in a given temperature of the air .
It is required to BDF find the proportion between the GN density of the air , and that of the quicksilver , the observed temC perature being supposed to obtain AL uniformly throughout the whole atmosphere ; or at least , through HM out a great portion of it , on all K sides of B. Suppose it done ; and that the density of the air , at the time P and place of observation , is to Q the density of the quicksilver R in the barometer , as the right S line CD to E F. Now the temperature of the air being giT ven , the subtangent of the logistic corresponding to the given temperature is given ( by prob .
3d ) .
Let GH be equal to that subtangent ; and let KL be the perpendicular height of the mercurial column supported in the barometer at A , the earth 's surface , at the time of observation at B. Now since OF is to CD as the density of quicksilver , in the barometer , to the density of the air at B , and CD is the perpendicular height of the column of quicksilver sustained in the Torricellian tube at the place B , at the time of observation , it is evident , that OF is the perpendicular height of a column of an inelastic fluid , of the same density with the air at B , which , if it were urged with an accelerative force through its whole length every where equal to that of gravity at B , would , by its compressive force , sustain the mercurial column C D : for the accelerative forces supposed to act upon the fluids in the columns E F , C D , being equal , and the heights of the columns being , reciprocally , as the densities of the fluids , the compressive forces of the columns will be equal .
But GH is the perpendicular height of a column of an inelastic fluid , of the same density with the air at A , the level of the sea , which , if the whole were urged with an accelerative force equal to that of gravity at A , would sustain the mercurial column KL ( by sect. 4 ) .
Suppose MN equal to the perpendicular height of a column of an inelastic fluid , of the same density with the air at B , which , if the whole were urged with an accelerative force , every where equal to that of gravity at A , would sustain the mercurial column C D. It is evident that OF is to MN as the accelerative force of gravity at A to the accelerative force of gravity at B ; for OF and MN are columns of fluids , of the same density , exerting equal compressive forces .
Therefore , the heights of the columns must be reciprocally as the accelerative forces by which the compressive force is produced .
But gravity at A is to gravity at B as the square of TB to the square of T A. Therefore , E F : MN=T B2 : T A2 .
Again , GH and MN are columns of fluids of different densities , acted upon by equal accelerative forces ; therefore , the compressive forces which they exert , that is , which they sustain , will be as their heights and the densities of the fluids jointly ; that is , if P , Q be as the compressive forces of those fluids , and R , S as their densities , then P : Q=GHx R : MNx S. But the compressive forces sustained by these columns are the pressures of the air at A and B ; also the densities of the fluids are the densities of the air at A and B. Therefore the compressive forces are as the densities ; that is , P : Q= R : S. Therefore GHx R : MNxS= R : S. Therefore GH and MN are equal .
Therefore E F : GH=E F : M N ; but it hath been shown , that E F : MN=T B2 : T A2 .
Therefore E F : GH=T B2 : T A2 .
But AB being given , TB is given , and the proportion of the square of TB to the square of TA is given .
Therefore the proportion of OF to GH is given .
And G H , is given as hath been shown .
Therefore OF is given , and the proportion of the given line OF to the given line C D , or of the density of the quicksilver to the density of the air , at B , is given .
Q E. I. COMPOSITIO N. Find the subtangent GH competent to the given temperature ; find a line , E F , to which that subtangent shall bear the duplicate proportion of TA to T B ; that is , of the earth 's semi-diameter to the earth 's semi-diameter increased by the given elevation of the place of observation , above the level of the sea .
As that line OF to the observed height of the quicksilver , so is the density of the quicksilver , in the barometer actually employed , to that of the air , at the time and place of observation .
Thus the specific gravity of air may be found , comparing it with quicksilver , and by means of quicksilver with other fluids .
Though it is only the most simple case of this problem that can ever come into practice , I chose to discuss it in its most general extent ; tent ; as this is the only one method which gives entire satisfaction to the mind .
SECTIONSIXT H. IHAVE now gone through the exposition and reduction of M. DE LUC'S rules , for measuring heights by the barometer .
The consonance of them with theory appears , upon a strict examination , to be such , as strongly confirms the principles , upon which the theory is founded .
I shall conclude these disquisitions , with pointing out some further objects of enquiry , concerning the modifications of the atmosphere , naturally arising out of M. DE LUC'S discoveries , in conjunction with the theory already established .
1 .
It is probable , that the absolute elasticity of the air may be affected by various causes besides heat .
The degree of humidity must occur to every one , as a circumstance , which may reasonably be suspected to have some influence upon it ; and , perhaps , the state and quantity of the electricity of the air may have more .
2 .
If M. DE LUC'S formulae are to be admitted as universally true , in all imaginable temperatures , there is a given temperature in which the elasticity of the air would be destroyed , and , in any lower temperature , it would be negative ; that is , the repulsion would be changed into attraction .
This given temperature is , 409,13 of BIRD'S Fahrenheit ; for if B+nx 8/ 3591 B , be the length of the subtangent of the atmospherical curve , in 1000ths of an English fathom , whatever be the value of n , then , when n= 448,875 , B8x n/ 3591B = 0 ; that is , the subtangent vanishes ; and the absolute elastic force , which is always as the subtangent ( by ~ .
4 .
) must vanish with it ; but when n= 448,875 the temperature is 409,13 .
Perhaps it may be thought more probable , that the variation of the subtangent , or of the elastic force , is not precisely as the variation of temperature .
If the subtangent changes in a geometrical proportion while the temperature , as shown by the thermometer , changes arithmetically , the subtangent , or the absolute elastic force , will not vanish with any assignable decrement of temperature ; and in that temperature , in which it should vanish , accords to M. DE LUC'S formulae , it will still remain more than 3/ 10ths of what it is in the temperature + 39,74 ; and yet the equation , for an increment or decrement of temperature , amounting to 40~ , will not differ from M. DE LUC'S by more than four fathom in the height of 1000 .
I must repeat , that I am now only pointing out the conclusions of theory , as hints of further enquiry .
I do not mean to substitute this hypothesis as more accurate than M. DE LUC'S in practice ; I do not affirm , that it is more true in theory .
I mean only to suggest , that if M. DE LUC'S formulae are admitted as mathematically true , a consequence will follow , which may seem to some unlikely to obtain in nature .
That however this consequence , if otherwise improbable ( which is not the opinion to which , for my own part , I incline ) is not to be too hastily adopted , upon the bare evidence of its arising out of M. DE LUC'S formulae ; because the general extent of these formulae is more than experiment hath hitherto proved .
And I show , by a very simple instance , that it is easy to imagine other laws , between the temperature and the elasticity , which , to all practical purposes , would be the same as M. DE LUC'S , within the utmost limits of his experiments , and even beyond them , and yet differ from it in remote consequences .
If M. DE LUC'S formulae be universally and strictly true , the consequence I have pointed out must be admitted ; and it will follow , that the force of elasticity owes its first production and origin to temperature : or , at least , that the relation between heat and elasticity , if not the most intimate one of cause and effect , is that which stands the next in the scale of natural affinities , namely , that they are inseparable effects of some common cause .
And these conclusions will hold , if the form of the general expression be true , though the quantity of the equation nx 8/ 3591 B should require a correction : for a change of the fraction 8/ 3591 will only alter the particular temperature , in which the subtangent and elasticity vanish .
3 .
The diminution of the air 's density , as we ascend from the surface , is subject to a limit .
This limit is different , in different heights of the barometer , at the level of the sea , and even in a given height of the barometer , in various temperatures .
But the density even at an infinite height is never 0 , or infinitely small .
In fig. 1 .
through C draw C F , parallel to A D , meeting the atmospherical curve in F. CF is the density of the atmosphere , at an infinite height ( by p. 231 ) .
The proportion of CF to the density at the level of the sea is given in any uniform given temperature of the air .
For the uniform temperature being given , the subtangent of the curve DF is given ; and consequently , the proportion is given , of which the given line CA is the logarithm .
The diminution of compressive force is subject to a corresponding limit .
This curious circumstance hath been remarked by no one , that I know of , but Dr. BROOKE TAYLOR , whose writings are too little attended to , even among mathematicians , from an opinion which prevails of their obscurity .
That consummate geometer seems indeed to have thought it improbable , that a finite density of the atmosphere , at infinite distances , though the necessary consequence of the theory , should actually exist ; and , for this reason , he would imagine such a law of the elastic force , as should render the density of the atmosphere , beyond a certain height , much greater than in proportion to the compressive force ; and circumscribe the whole within narrow limits(w ) .
And if the theory inferred a great density at infinite or even at great heights , such a density as would sensibly disturb the motions of the planets , it would be necessary to have recourse to some such hypothesis ; but , as this is not the case , I see no such necessity .
I know not , for what reason , mathematicians have been afraid to admit the infinitude of the atmosphere of the earth ; whether they thought it would bear hard upon the Newtonian doctrine ( w ) Method .
Increment .
prop. 26 .
Scholium .
of a void ; or , that it implied the infinitude of matter .
But neither the one nor the other of these consequences is to be apprehended : for neither the phenomenon of nature , nor the principles of the Newtonian philosophy , require , that there should be any where a great chasm in the universe ; or that the whole material world should be actually circumscribed within any finite space .
A large proportion of pore , or interspersed vacuity , is sufficient for all purposes .
Nor doth an absolute infinity of matter by any means follow from the hypothesis of an infinite number of finite masses ; and an infinite number of finite masses is all that is implied in the notion of a rare elastic fluid , diffused throughout infinite space .
I agree , indeed , with Mr. COTES , that there are no data from which any great altitudes of the atmosphere can indubitably be concluded , in the way of experiment : but I do contend , that there are no data , from which the supposition of its infinite height can , in the same way , be disproved .
And this may justly be held more probable than the contrary , as being the consequence of a theory which hath never yet , in any instance , proved fallacious ; and this I venture to assert , with the less hesitation , as , besides the evident reason of the thing , the great authority of NEWTON is on my side .
The infinite extent of the earth 's atmosphere is manifestly supposed , in that subtle disquisition concerning the tails of comets , which occurs in the posthumous work De Systemate Mundi .
Instead of framing hypotheses , therefore , to remove imaginary difficulties , I shall pursue the theory as far as it will lead ; and my next step shall be , to exhibit the successive rarefactions of the atmosphere , as we ascend from the earth , in a table .
Heights Volume .
This table shows the in miles rarefactions of the atmo0,000 00,0 1 sphere at different heights , 0,01 40,0 3069 above the surface of the 0,02 80,8 30692 3 earth , upon a suppositi0,03 122,6 3069 4 on that the temperature , 0,04 165,0 3069 5 throughout the whole , is 0,05 208,5 3069 6 that which is indicated by 0,06 252,1 3069 0,07 298,2 30697 + 40~ of BIR D'S Fahren0,08 344,5 30698 heit .
The middle column 0,09 391,9 30699 shows the heights above the 0,10 440,3 306910 surface , corresponding ( ac0,11 489,7 306911 cording to the construc0,12 540,6 306912 tion of the atmospherical 0,20 990,6 306920 0,30 1698,1 306930 logarithmic , ~ .
2 ) to the 0,40 2641,7 306940 decimal parts of the earth 's Semidiameter .
semi-diameter expressed by 0,50 1,00 306950 the numbers in the first 0,60 1,50 306960 0,70 21/ 3 306970 column ; and in the third 0,80 4,00 306980 the corresponding rarefac0,90 9,00 306990 tions , or the proportions of 0,98 49,00 306998 the spaces which a given 0,9836 + 62,00 306998,36+ quantity of air would occu1,00 Infinite .
3069100 pie at the different heights , are expressed by the powers of the number 3069 .
The principle and method of the construction of this table is obvious .
I compute the length of the mean semi-diameter of the earth in thousandths of an English sathoms .
I divide that number by 100 .
The quotient is the tabular logarithm of the 100th root of a number , which is to unity , as the density at the surface to the density at an infinite height , in the temperature supposed ; or , which is the same thing , as the space occupied by a given quantity of air , at an infinite height , to the space occupied by the same quantity , at the surface .
This root I call the radical number .
Then , imagining the earth 's semi-diameter , divided into 100 equal parts , and numbering those parts 1 , 2 , 3 , &amp; c , downwards , towards the center , I compute the heights , above the surface , corresponding to those parts successively , according to the construction of the atmospherical logarithmic , ~ .
2 ; and , writing the resulting numbers in the second column , in the third , I write the powers of the radical number , increasing by unit , in regular succession downwards .
The proportion in which the atmosphere will be rarefied , at given heights above the surface , will be very different , in different temperatures .
It may always be exhibited by a table of this form , but every different temperature will have its own radical number .
The radical number , in the temperature + 40 , is 3069 ; and , for any other given temperature , may be thus found .
Call the tabular logarithm of 3069 , y ; and lot n be the number of degrees of BIR D'S Fahrenheit , in the difference between the given temperature , for which the radical number is to be found , and + 40 .
Then y~ n/ 449 ~ny will be the tabular logarithm of the radical number , for the temperature assigned : observing , that n is to be positive or negative in the denominator of the coefficient n/ 449 ~ n , according as the given temperature is greater or less than + 40 ; and that the term n/ 449 ~ny is negative when n in the denominator of its coefficient is positive , and vice versa .
Thus the radical number , for the temperature 0 , is 6731,2 + .
This table is not intended for any practical purposes ; but merely for speculative amusement .
They , who take delight in the contemplation of final causes , will remark , with admiration , how large a part of the whole rarefaction of the atmosphere is performed on this side of the moon .
Indeed there is comparatively but a very small part performed beyond it : so that the moon revolves at a distance where the resistance , from the earth 's atmosphere , is reduced almost to its minimum .
For if we imagine a series of quantities , consisting of 101 terms , decreasing from the first continually in geometrical proportion ; and such , that the first and greatest shall be to the last and least , as the density of the atmosphere , at the earth 's surface , to the density at an infinite distance ; the density , at the mean distance of the moon , will be less than the 99th of these , or the least but two .
4 .
An increase of temperature raresies the lower regions of the atmosphere , more in proportion than the upper , and brings the constitution of the whole nearer to that of an uniform density .
Though the effect of heat alone , upon every orb of the atmosphere separately , is to rarefy each , perhaps , proportionally , yet its operation upon the different orbs is so modified , by the different degres of pressure they sustain , and by the communication between them , that the greater the temperature of the atmosphere , though uniform throughout , the less will the proportion be of the density at any given height to the density at any greater given height .
For the greater the temperature , the greater the subtangent ; and the greater the subtangent , the less the proportion of AD to C F , of which the given line AC is the logarithm .
( Vid. fig. 1 .
) 5 .
If at any height above the surface of the earth a given alteration of temperature diminish the air 's density in the same proportion , as it increases the absolute elasticity , or vice versa , the pressure of the superincumbent atmosphere , at that height , will remain unchanged .
At all lower heights , the pressure will be less , than in a cooler condition of the atmosphere , and greater at all greater heights .
On the contrary , the pressure at all lower heights will be greater than in a warmer condition , and at all greater heights less .
For let CA ( fig. 2 .
) represent the semi-diameter of the earth , the curve DEF the atmospherical logarithmic for a certain temperature , and GHK the logarithmic for another greater temperature .
Let the ordinates of the two curves AD , AG be as the densities , at the earth 's surface , in the different temperatures , to which the curves belong , respectively .
Then it is evident , the ordinates ~E , ~H , drawn through any other point ~ in the asymptote , will be as the densities , at the height to which A~ corresponds , in the different temperatures , respectively .
Now , suppose that the density of the air , at the height B , in the greater temperature , is less than the density at the same height in the cooler temperature , in the same proportion , as the absolute elasticity in the greater temperature exceeds the absolute elasticity in the less .
Then I say first , that the pressure of the superincumbent atmosphere at B , is the same in both temperatures .
For take C~ =C A2/ C B , and draw the ordinate ~E , cutting the curves in E and H , and through E and H draw tangents to the curves , EL , HM , meeting the common asymptote , A C , in L and M. Now the subtangents ~L , ~M , are as the absolute elasticities in the different temperatures ( by p. 250 .
) .
And ~E , ~H are as the densities at B ( by construction ) .
Therefore ~ H : ~E=~ L : ~ M. Therefore ~HX~M =~EX~ L. But the rectangle ~EX~L is equal to the area intercepted by the ordinate ~ E , the curve E F , and the asymtote ~ C , infinitely produced .
And the rectangle ~MX~H is equal to the area intercepted by the ordinate ~ H , the curve H K , and the asymptote ~C infinitely produced .
Therefore these areas are equal .
And these areas are as the pressures of all the atmosphere above B , in the temperatures to which the curves belong , respectively .
Therefore the pressures in these different temperatures are equal .
Q. E. D. I say further , that the pressure of the superincumbent atmosphere , at any height below B , is less in the greater temperature than in the cooler .
Let AP be any height less than B , and take Cp =C A2/ C P , and draw the ordinate p N , cutting the curves in the points N and O. Now p N , ~E being as the densities at P and B , in the cooler condition of the air , and p O , ~H as the densities at the same heights P and B , in the increased temperature , pO is less than to bear to ~ H , the proportion of pN to ~E ( by 4th of this ) : and , by permutation , pO is less than to bear to pN the proportion of ~H to ~ E , or of ~L to ~ M. Therefore pOX~M is less than pNX~ L ; that is , the area intercepted between the ordinate p O , the curve OH K , and the asymptote pC infinitely produced , is less than the area intercepted by the ordinate p N , the curve NO F , and the asymptote pC infinitely produced .
But again , these areas are as the pressures of all the atmosphere above P , in the temperatures to which the curves respectively belong ; therefore , the pressures in the greater temperature , to which the curve OHK belongs , is less than the pressure in the cooler temperature , to which NEF belongs .
Q. E. D. In like manner it may be shown , that the pressures in the warmer temperature , at all heights above B , are greater than in the cooler .
And thus theory might have brought us to expect a phenomenon , which M. DELUC hath actually observed , and was not a little surprised at .
For if the temperature of the atmosphere be at any time gradually augmented , a barometer , placed below the height of stationary pressure , will sink , while another , placed higher up in the atmosphere , will rise .
This is what M. DELUC hath observed in two barometers at the foot and summit of a hill , in settled weather , while the natural heat of the day hath been upon the increase ; from whence he , with reason , concludes , that between these two stations , where the pressure was changing , at the same time , contrary ways , there must have been an intermediate one , which I call the height of stationary pressure , where no change , in either sense , could take place .
I shall hereafter show , at what height the place of unaltered pressure should fall , by theory , for every change of temperature .
It seems a problem worthy of a naturalist , to enquire how far theory doth , in this circumstance , agree with the real operations of nature .
6 .
It may seem , perhaps , still more surprising , but it is no less true , that there will generally be a particular height in the atmosphere where the density will remain unchanged , by a given change of temperature .
To determine in what changes this will happen , and at what height the place of unaltered density , for given changes of temperature , should fall , requires only the solution of the following problem .
PROBLEMFIRS T. To find the intersection of two logarithmics , which have a right line given in position for their common asymptote , and their subtangents given in magnitude ; an ordinate in each curve , drawn at right angles with the common asymptote , through a given point in it , being also given in magnitude .
IMAGINE two logarithmics , CD E , FDG ( fig. 3 , 4 , 5 , 6 .
) , having the right line A B , given in position , their common asymptote ; and suppose the subtangent of each curve given in magnitude .
Through a given point A , in the common asymptote , imagine the right line AC drawn , an ordinate at right angles with the asymptote , meeting the curves in C and F ; and let A C , AF be severally given in magnitude .
It is required to find the point where these curves intersect .
Suppose it done , and let D be the intersection .
Draw DL perpendicular to A B. Through F , the point where AC meets one of the curves FD G , draw FM parallel to A B , meeting the other curve CDE in M. Draw MN perpendicular to A B , and take A H , AK equal to the given subtangents of the curves , CD E , FD G , respectively .
Now AL is the logarithm of the ratio of AF to L D , in the system of the curve FD G ; and ( because NM=A F ) NL is the logarithm of the same ratio , in the system of the curve CD E. Therefore , A L : LN=A K : A H. Therefore the proportion of AL to LN is given ; and consequently , that of AN to AL is given .
But AN is given in magnitude .
For AC and AF are given in magnitude ( by hypothesis ) .
Therefore , the proportion of AC to AF or NM is given ; and AN is the logarithm of that given proportion in the system of the given curve CD E. But AN being given in magnitude , and the proportion of AN to AL being given , AL is given in magnitude .
And it is given in position , and the point A is given ( by hypothesis ) .
Therefore the point L is given ( by 27 .
dat .
) .
Therefore , L D , L D , being perpendicular to A L , is given in position ( by 30 .
dat .
) .
But AL being given in magnitude , the proportion of AF to LD is given ( by logarithms ) .
And AF is given in magnitude .
Therefore LD is given in magnitude .
Therefore the point D is given .
Q. E. I. The construction is obvious .
It is evident , that the points L and N are on the same side of A , if F be at the greater curve , as in fig. 3 and 4 ; but on different sides of A , if the curve to which F belongs be the less , as in fig. 5 and 6 ( x ) .
The calculation of the lengths A L , L D , by means of the logarithmic canon , is very simple .
Putting B for the subtangent of the Briggian system , L for the tabular logarithm of A C , and D for the difference of the tabular logarithms of A C , A F , we shall have , First , AL=AHxAKx D/ B xK H. And again , L~AKx D/ K H= tab .
log .
of L D. In this second expression , the second term is negative , if the greater of the given ordinates belong to the less curve , as in fig. 3 and 5 ; but positive if the greater ordinate belong to the greater curve , as in fig. 4 .
and 6 .
Both these theorems are so easily derived from the preceeding analysis of the problem , that it is needless to add the synthetic demonstration ; but they may be reduced to more commodious forms for practice by the following artifice .
( x ) By the greater and the less curve I mean that which hath the greater or the less subtangent .
to AF being given ( by hypothesis ) , the length of AL will be given , from the same analysis as before .
But D being given , and the right line AB given in position , LD is given in position and magnitude ; and , AL being given , the proportion of LD to AC , and also to AF , is given ( by logarithms ) .
Therefore AC and AF are each given .
The expression for the length of AL is the same as before .
And tab .
log .
LD ~ AK ~ D/ KH = L. In this expression the second term is positive , if the ordinate from A to the lesser curve is to be the greater of the two ; in the contrary case , negative .
Now imagine CD E , FDG to be logarithmics of the atmosphere , in different temperatures ; AC being the density at the earth 's surface in one temperature , and AF in the other : let the AB be the semi-diameter of the earth : let the two logarithmics meet in D , and draw DL perpendicular to A B. Now if the point L be any where in the line A B , between A , which is at the surface of the earth , and B , which is the center , the ordinate LD will represent the density of the air , in the system of both curves , at the distance BA ~ AL/ BA AL above the earth 's surface ; and therefore , at this height , the density is the same in the one temperature as the other .
If L coincide with B , DL represents the density at an infinite height ; but if L falls beyonds B , DL is not among the ordinates , of either curve , which represent densities any where existing .
The expresO o2 sion sion BA ~ AL/ BA AL , which is infinite when AL = AB , now becomes more than infinite , the denominator being negative ; and there is no height in the atmosphere , at which the density is the same in both temperatures .
Again , if L fall above the earth 's surface ( as in fig. 4 .
and 6 .
) , LD is not among the ordinates of either curve which represent densities ; and AL , being negative , the expression BA ~ AL/ BA AL becomes BA ~ AL/ BA + AL ; which expresses a distance below the earth 's surface ; but whether L fall above or below the surface depends upon the state of the density at A , and the temperature jointly .
If the density were to be greater , in the greater temperature , then the greater ordinate at A belongs to the greater curve ( as in fig. 4 .
and 6 .
) , and L is above the surface : but if the density be less , when the temperature is greater , then , of the two ordinates at A , the less belongs to the greater curve , and L is below the surface ( as in fig. 3 .
and 5 .
) ; and in this case , the place of the point L depends upon the magnitude of AN , and the proportion of HK to AK .
If HK : AK = AN : AB , then L and B ( in fig. 3 .
and 5 .
) coincide , and the densities are the same at an infinite height .
If HK be less than to bear to AK , the proportion of AN to AB , AL will be greater than AB , and L will fall beyond B , and the densities are no where the same .
But if HK be greater than to bear to AK , the proportion of AN to AB , AL will be less than AB ; and in this case three will be a height above the earth 's surface , namely , BA ~ AL/ BA AL , at which the densities , in both temperatures , will be the same .
And it is evident , that this height is given , if AL be given .
But AL is given , by the preceeding problem , if the subtangents AH , AK be given in magnitude , and the proportion of AC to AF be given .
But if each temperature be given , each subtangent is given ( by sect. 5 .
prob .
3 .
) and the proportion of AC to AF will be given by the barometer .
( sect. 5 .
prob .
4 .
) EXAMPL E. Brs obTr Tr Brs reduced Densities , by 1744 , served .
in .
out .
to common ~ .
5 .
prob .
4 .
~h temperature March 26 11/ 2 30,17 54 61 30,176 137904 April 26 29,37 56 48 29,37 138039 Hence D= 0.0004080 = 4,08 fathom .
But p= 8 .
s= 21 .
q= 449 .
q+s= 470 .
sp= 13 .
and q+ s/ s p= 36,1 .
Therefore by formula 2d , AL = 150 fathom ; and at a height , insensibly greater , the density was the same in both constitutions of the atmosphere .
7 .
It is manifest , that the changes of density above and below the height where it remains unaltered , are contrary .
If the lower densities are diminished , the higher ones are increased , and vice versa .
Notwithstanding the mathematical evidence of these conclusions , I am persuaded , it will appear to many a physical paradox , that heat should any where condense .
I have already hinted at the solution .
Heat doth not condense any sinite mass of matter , which is at liberty , on all sides , to expand itself ; but acting on a finite mass , which hath not unlimited liberty of expansion , it may condense one part by the rarefaction of another .
Acting on an infinite mass , it must do this ; because it can only transpose .
It neither generates new matter , nor annihilates what already is .
What is taken therefore from one part , must be added to another , and vice versa ; otherwise the quantity of matter , in the whole , must be changed .
Imagine ABCD to be a small portion of any orb of the atmosphere ( fig. 8 .
) AEFB , CMND , CHGA , DLBK , contiguous portions .
Heat drives many of the particles , which occupy the space ABCD , out of it ; but it likewise drives out many of the particles which occupy the contiguous spaces .
And of those which are driven out of the contiguous spaces , many will enter the space ABCD .
If the particles which are driven into ABCD , be more in number than those which are driven out of it , the air , in this space , is condensed , by that very cause , which would rarefy it , if the contiguous portions were annihilated .
Thus condensation in one part may , in an infinite fluid , must ensue from rarefaction in another , if the quantity of matter remains unaltered .
Where then is the wonder , that the like effect should follow from the cause of rarefaction combined with other causes ?
8 .
The cases in which L falls below B , or above the surface , seem to be physically impossible , without an alteration of the quantity of matter in the atmosphere ; for without this , it cannot be every where rarefied , or every where condensed at one and the same time .
9 .
Having found the heights where the density is the same in any two different temperatures , the height where the pressures in different temperatures are equal , will easily be determined .
For this purpose , we have only to seek the solution of the following problem .
PROBLEM II .
Two logarithmics ( DQF , GQK ) intersecting in a given point ( Q ) , ( Vid. fig. 2 .
) , having a right line ( AC ) given in position for their common asymptote , and subtangents severally given in magnitude ; to find the point ( ~ ) in the asymptote , where an ordinate ( ~HE ) being drawn at right angles , to meet both curves , the areas intercepted between the two curves and the common asymptote infinitely extended , beyond the ordinate ~E , are equal .
LET ~L , ~M be equal to the subtangents of the curves DQF , GQK , respectively .
Because the curvilinear areas are equal ; therefore , the rectangles ~L ~ ~E , ~H ~ ~M , are equal .
Therefore , ~E : ~H = ~M : ~L .
But ~M , ~L , being given , the proportion of ~M to ~L is given .
Therefore , the proportion of ~E to ~H is given .
Therefore , the point ~ is given ( by corollary of preceeding problem ) .
CONSTRUCTIO N. Find the point ~ ( by corollary of preceeding preproblem ) such that the ordinate ~HE being drawn , ~E , ~H , may be reciprocally as the subtangents ; or that ~E may be to ~H as the subtangent of the curve GHK to that of DEF : and the thing proposed is done .
Now imagine the curves GQK , DQF , to represent the logarithmics corresponding to different given temperatures of the air , A being at the surface of the earth , and C at the center .
And having found the point ~ , take AB = A~ ~ CA/ CA A~ ; then B is the place where the barometer will stand at the same height , in both conditions of the atmosphere .
LIMIT S. Draw the ordinates , AGD , CFK , and QR .
If the point Q be any where below the ordinate through A , it is evident , the point ~ will lye above the point R. And if the subtangents ~L , ~M , be reciprocally as the densities AD , AG , the points ~ , A , and consequently B , A , coincide , and the pressure , at the level of the sea , remains unchanged .
If the greater subtangent ~M be greater than to bear to the less ~L , the proportion of the greater density AD to the less AG , ~ will fall above A , and B consequently will be below it ; for A~ becoming negative , the expression CA ~ A~/ CA A~ becomes CA ~ A~/ CA + A~ And there is no place in the atmosphere where the pressure is the same in one condition as the other .
But it is , at all heights , greater in the warmer condition ( by 5th of this ) .
If the point Q fall below the central ordinate CK , ( fig. 7 .
) and the subtangents ~L , ~M , be reciprocally as CF , CK , the points ~ and C will coincide , and B will go off to an infinite height .
If the greater subtangent ~M be less than to be ~L as the greater density CF to the lesser CK , ~ will be found below C , and the ordinate ~E is not one of those , by which the density of the air , at any height , in either condition , is represented .
The expression CA ~ A~/ CA A~ is more than infinite , the denominator being negative ; and there is no height at which the pressure is the same , in both conditions of the atmosphere ; but it is at all heights less in the warmer condition ( by 5th of ths ) .
But whenever Q falls below the superficial ordinate , whether it be above or below the central ordinate , if ~M be less than to be to ~L as AD to AG , and greater than to be to ~L as CF to CK , ~ will fall between C and A , and an equality of pressure , in both conditions of the atmosphere , will take place at a finite height , determined as above .
If Q ever falls above the superficial ordinate through A , ~ will be more above it , and B will be below the earth 's surface , and there will be no height at which the pressure will be the same ; but it will every where be greater in the warmer temperature .
The calculation for determining A~ is obvious from the foregoing solution of the problem .
For putting B for the subtangent of the Briggian system , D for the difference of the tabular logarithms of AG , AD , and ~ for the difference of those of ~L , ~M , we have , AR=~L x ~M x D/ B x LM ( by prob .
1 .
) and R~=~L x ~M x ~/ B xLM ( by cor .
1 .
) Therefore , A~ =D~x ~L x ~M/ B x LM .
Therefore , putting ~L=B ~ p/ q B , and ~M=B ~ s/ t B. and substituting D~ for D in the formulae for AL , deduced from problem 1 , we shall change them into formulae for A~ .
EXAMPLE S. In the example of the preceeding problem D= 4,08 , and ~= 121,817 fathom .
Whence D~ is negative , and the point B falls above the surface , and the pressure was no where the same , but at all heights greater in the warmer condition .
EXAMPLE II .
Br .
obTr .
Tr. Br .
reduced Densities , by served .
in .
out .
to common ~ .
5 .
prob .
4 .
dh temperature March 26 20 30,07 58 59 30,058 13796,6 28 21 30,21 54 50 30,21 14138,2 Hence p=10 .
s=19 .
q=449 .
q+s=468 .
s-p=9 .
q+ s/ s-p=52 .
D=0.0106238 .
~=0.0084332 .
D-~=0.0021906=21,906 fathom .
Therefore , A~ = 1164,42 fathom = 1,32 miles ; and at a height insensibly greater , the pressure was the same in both conditions .
In this manner may the heights be determined , in which either the density or pressure of the air , in a given temperature , is the same as in any greater or lower temperature ; provided the proportion be known between the densities , in the different temperatures , at the level of the sea , or at a given elevation above it ; otherwise both the problems are indeterminate .
And I know no method of ascertaining this proportion , but by actual observation of the barometer .
The change of temperature determines the proportion in which the subtangent of the atmospherical logistic is changed ; but in whatever proportion the subtangent of a logistic is changed , the ordinate , at a given point of the asymptote , may be of any imaginable length , or may bear any imaginable proportion to the subtangent .
There is no given geometrical relation , in the nature of the curve , between the length of the one and the other ; and experiment hath not yet brought to light any physical relation in the particular case in question .
Diligent observation of the barometer and therP p2 mometer mometer may perhaps , in future , give more satisfactory information upon the subject .
Perhaps I have been too minute in detailing consequences from an hypothesis , of what probably never doth obtain ; namely , that the atmosphere may be considered as equally heated in all its different parts ; and that the variation of the absolute elasticity of its component particles is proportional to the change in the degree of uniform temperature , as if the absolute elasticity were influenced by no other cause .
But such is the necessary order of enquiry .
Theory must begin with the most simple cases , considering what would be the effect of some known cause , acting singly , and in the most simple manner ; and comparing the conelufions from what it is supposed , with the effects which are , the difference leads us to the discovery of other causes , and to an estimation of the separate effects of each , and of the compound effects of all the known causes any how combined .
I shall only add , that as the state of an atmosphere , unequally heated in its different parts , may be represented by parts of different logarithmics , it is possible , that instead of one point of unaltered density , and one point of unaltered pressure , we might find several , if synchronous observations could be made at several successive elevavations sufficiently different .
10 .
If the atmosphere of the earth reaches to infinite heights with a finite density ; for the same reasons , that of Jupiter and every other planet , will reach also to infinite heights , above the surface of the planet with a finite density .
The atmosphere therefore of every planet will reach to the surface of every other planet , and to the surface of the Sun ; and the atmosphere of the Sun to the surfaces of them all .
All these atmospheres will mingle , and form a common atmosphere of the whole system .
This common atmosphere of the system will be infinitely diffused , since the particular atmospheres that compose it are so .
It will reach therefore to every fixed star ; and , for the same reason , that of every fixed star will reach the central body of our system , and of every other system .
The atmospheres of all the systems will mix .
The universe will have one common atmosphere , a subtle elastic fluid , which pervades infinite space , and being condensed near the surface of every larger mass of matter , by the gravitation towards that mass , forms its peculiar atmosphere .
To certain distances from every one of these great bodies , the condensations of this infinite fluid will follow the laws of the preceeding theory nearly ; but to certain distances only .
For that theory considers only the effects of the attraction of a single sphere , and assigns the law of the variation of density , such as would obtain , if one spherical body existed in the midst of an infinite fluid ; and such as cannot generally obtain , unless that hypothesis were true : for many great spheres being immersed in a common atmosphere , the attraction of any one , at great distances from it , becomes but an inconsiderable part of the whole cause , by which the density is modified , the joint forces of them all .
And in many other circumstances , besides the condensation , the modifications of every particular atmosphere may depend upon those of others , innumerable and infinitely remote ; as the effluvia and exhalations of each great mass , mingling with its atmosphere , may be distributed , in certain proportions , among all the rest .
So that there is probably no branch of physics , in which human discovery , in its utmost extent , must always bear so small a proportion to what will still remain unknown .
TABLE I. EQUATION of the boiling point .
Barometer .
Equation .
Difference .
31,0 + 1,57 0,78 30,5 + 0,79 07,9 30,0 0,00 0,80 29,5 0,80 0,82 29,0 1,62 0,83 28,5 2,45 0,85 28,0 3,31 0,86 27,5 4,16 0,88 27,0 5,04 The numbers in the first column to the left express heights of the quicksilver in the Barometer , in English inches and decimal parts .
The second column gives the Equation to be applied , according to the sign prefixed , to 212~ of Bird 's Fahrenheit , to find the true boiling point , for every such state of the barometer .
The heights of the barometer decrease by 1/ 2 inches from 31 to 27 inches .
The boiling point , for all intermediate states of the barometer , may be found , with a very sufficient accuracy , by taking proportional parts .
For which purpose , the differences of the equations are given in the third column .
EXPLANATIO N. The principal table consists of three columns .
The numbers in the first column signify degrees , either of DE LUC'S scale , or the scale which hath been hitherto chiefly in use among the French , which is divided as M. DE LUC'S is ; but 28 French inches being generally mentioned , by the French mathematicians , as the mean height of the barometer at PARI S , it is to be supposed , that their boiling point will generally agree to that state of the barometer .
The numbers in the next column to the right , which is marked L reduced , give the degrees of BIRD'S Fahrenheit corresponding to the degrees of M. DE LUC'S scale , expressed by the numbers of the first column ; and the numbers in the column marked P reduced , give the degrees of BIRD'S Fahrenheit corresponding in like manner with the PARIS scale .
The heights in the first column diminish by 5~ successively .
But the reduction may be computed to every degree , by means of the little table on the left ; which gives the value of single degrees , either of DE LUC'S , or of the PARIS scale , in degrees of BIRD'S Fahrenheit .
The titles L F , PF signifying DE LUC'S degrees , or Paris degrees , reduced to Fahrenheit 's , respectively .
If there should be occasion to reduce heights of BIRD'S Fahrenheit , either to M. DE LUC'S , or the PARIS scale , this may be done by the principal table , and the small one on the right hand ; which exhibits the value of single degrees of BIRD'S Fahrenheit in degrees , both of DE LUC'S and the Paris scales : the titles F L , FP signifying degrees of BIRD'S Fahrenheit reduced to M. DE LUC'S degrees , or the PARIS degrees , respectively .
EXAMPL E. EXAMPLE 2 .
To find the point upon a BIRD'S To find the point upon the Paris Fahrenheit , which corresponds to scale , which corresponds to + 55 + 58 of DE LUC'S scale .
of BIRD'S Fahrenheit .
By principal table + 55 = 154,37 By principal table + 54,45=10 By little table on 55-54,45=0,55 3= 6,67 the left hand By little table on the right hand 0,5 = 0,22 Therefore + 58 = 161,04 By ditto 0,05= 0,02 Therefore + 55 =10,24 TABLE III .
EQUATION for the temperature of the QUICKSILVER .
Degrees Correction of Bird 's of Diff. Fathom Fahrenh .
Logarith .
The numbers in the first column 1 452 0,452 0,003 are to be understood to express differ2 904 0,904 0,006 ences of temperature , in the quick3 1355 1,355 0,009 silver of the portable Barometers at 4 1807 1,807 0,012 the two stations , indicated by Ther5 2259 2,259 0,016 mometers fixed in the cases of the 6 2711 2,711 0,019 Barometers , in degrees of Bird 's 7 3163 3,163 0,022 Fahrenheit .
The second column 8 3614 3,614 0,025 gives the corresponding corrections 9 4066 4,066 0,028 of the difference of the tabular lo10 4518 4,518 0,031 garithms of the observed heights of 20 9036 9,036 0,062 the quicksilver in the barometer , to 30 13554 13,554 0,094 be added or subtracted , according as 40 18072 18,072 0,125 the higher barometer hath been the 50 22590 22,590 0,156 warmer or cooler of the two .
The 60 27108 27,108 0,188 numbers in the third column are the 70 31626 31,626 0,219 1/ 1000th parts of those in the second .
The manner of using them is explained in the general precepts .
The fourth column gives the reduction of the observed height of the column of quicksilver , in decimal parts of an inch , for the differences of temperature expressed by the numbers of the first column , when the height of the quicksilver is 30 inches ; and the reduction for any other height of the quicksilver may be thus found .
Suppose the height of the barometer 30 ~a inches , in a certain temperature b , and this is to be reduced to the temperature b~ n. From the fourth column collect the number corresponding to n ( by repeated entry , if need be ) .
Call that number M. Then M~ a/ 30 M is the reduction .
GENERAL PRECEPTS , for the CALCULATION of HEIGHTS by these TABLES .
FROM the common tables of logarithms , namely , those which exhibit the Briggian logarithms to eight places , write out the logarithms of the numbers , which express the observed heights of the quicksilver , in the portable barometers , at the two stations .
2 .
Subtract the lesser logarithm from the greater .
3 .
Divide the remainder by 1000 .
The quotient is a certain number of fathom .
4 .
Take the difference of the temperatures of the quicksilver , as indicated by the thermometers in the cases of your portable barometers , and look for the correction corresponding thereto , in the third column of TABLE III .
and add that correction to the number last found , if the barometer at the higher station hath been the warmer of the two ; otherwise subtract it .
Call the sum in the first case , the remainder in the latter , the approximate height .
5 .
Add together the temperatures expressed at the two stations by the thermometers in the open air , under their proper signs .
Take half the aggregate , and call it the temperature of the air .
6 .
In the principal table of the equation for the temperature of the air ( TABLE IV .
) , look a-top for the decade of degrees next less than the temperature of the air , found by the last rule ; and from the column underneath it ( by repeated entries if need be ) collect the corrections for the units , decades , and centuries of fathom in the approximate height ; and if you have chiliads of fathom in your approximate height , find the correction for them , by taking the decuple of the correction for the corresponding centuries .
Add these several parts of the correction together , and call the sum N. 7 .
In the little table look , either a-top or in the outer vertical row to the left , for the units of degrees in the temperature of the air , over and above decades , and collect the correction for these , for the units , decades , centuries , &amp;c. of fathom in the approximate height , by taking the numbers found in the little table , and proper multiples thereof .
Add these numbers together , and call the sum n. 8 .
If the sign of the correction found in the great table be + , add N+n to the approximate height ; but if the sign of the correction found in the great table be - , subtract Nn from the approximate height .
The sum in the former case , the remainder in the latter , is the correct height , according to M. DE LUC'S rules .

