On the Astronomical Refractions .
By J. Ivory , A.M. F.R.S. Read June 19 , 1823 .
[ Phil. Trans. 1823 , p. 409 .
] The ancients , Mr. Ivory observes , were acquainted with the existence of atmospherical refraction ; but the first that ascertained its magnitude with tolerable accuracy , and employed it in his calculations , was Tycho Brahe .
Cassini attempted to compute the refraction upon optical principles , and upon the hypothesis of an uniform medium of uniform density , -a supposition which , though very simple , is sufficiently correct to a considerable extent .
The next step was to imagine an atmosphere of a density uniformly decreasing as the height increases .
Kramp was still more accurate in attending to the true effects of pressure and change of temperature ; his methods have been improved and extended with great sagacity by Laplace , and the tables founded on his computations are perhaps the best in existence with respect to the value of the mean refractions .
An uniform atmosphere must be supposed to be five miles in height ; an atmosphere uniformly decreasing in density ten .
Kramp and Laplace consider it as infinite .
The former limits would make the horizontal refraction less than the truth ; the latter supposition much greater .
Mr. Ivory is inclined to suppose some considerably extended , though finite height , which shall give the true refraction at the horizon , and which will probably be also correct for all other cases ; and he thinks it not superfluous to inquire , whether such an atmosphere would afford results sensibly different from those of an atmosphere of infinite extent .
The phenomena of twilight and of meteors indicate , he observes , a height of forty or fifty miles , at which the atmosphere is dense enough to reflect a sensible quantity of light .
At the height of about 25,000 miles the centrifugal force would become equal to the gravitation of the air , and the equilibrium would be no longer possible .
But the great reason that prevents our supposing an atmosphere infinitely extended , is the coldness of the upper regions .
Mr. Ivory considers the elastic force as disappearing from the effect of cold upon the temperature , is 266 centigrade degrees below the freezing point ; and he observes , that if the decrease uniformly amounted to a degree in 95 ( English ) fathoms , as it appeared to do in Gay Lussac 's aerostatic observation , the whole height ought not to exceed 29 miles ; consequently he thinks that the thermometer must fall at a slower rate in the higher than in the lower parts of the atmosphere .
Mr. Ivory proceeds to investigate the motion of light according to the laws of central forces , and to the experiments of Hauksbee and others on the refractive density of the air .
He first shows that the formula employed by the French astronomers , as far as 740 from the zenith , is deducible from any imaginable law respecting the constitution of the atmosphere ; and he gives , for example , the mode of obtaining it from that of Cassini .
He next considers the case of an equable variation of temperature in ascending , which he thinks is rendered probable , as the law of nature , by observations of Gay Lussac and others ; and he inquires into the general methods of integrating the expressions of the refraction in such cases , according to the methods already employed by Kramp and Laplace .
He examines , on various suppositions respecting the height of the atmosphere as connected with various laws expressing the progressive temperatures , beginning always from that which is observable at the earth 's surface the amount of the horizontal refraction , which he finds not so materially affected by these suppositions as to enable us to decide with certainty , from observation , which of them ought to be preferred .
Upon these foundations the author investigates the actual magnitude of the refraction under different circumstances , by means of several infinite series , of which he computes the values ; and he compares his results with those which have been observed by astronomers .
The ingenious hypothesis of Laplace , he remarks , gives us a height of 59I fathoms for a depression of a centigrade degree , or 197 feet for 1 ?
of Fahrenheit , which , he says , is little more than two thirds of the height actually corresponding to this depression ; and the French table , he says , was computed for the foregoing point , and then altered proportionally throughout its extent for a difference of 10 centigrade degrees .
As far as 86 ?
from the zenith , Mr. Ivory 's computations agree very accurately with those of Bessel ; and further , this celebrated astronomer does not recommend the adoption of his table .
The comparison of the new table with the observations of the French Astronomers , and of Dr. Brinkley , appears to be highly favourable to the accuracy of Mr. Ivory 's results .

