Here we find, as in every analogous case of difficulty, the use of the comparative historical method which I have just employed. We shall enlarge on this hereafter: meantime, I must oifer the observation that the philosophy of the sciences cannot be properly studied apart from their history; and, conversely, that the history apart from the philosophy, would be idle and unintelligible.
In reviewing the different departments of Rule of ar-Physics, I shall follow the rule which deter- rangement in mines the order of the sciences themselves: Physics, that is, I shall take them in the order of the generality of their phenomena, their simplicity, the relative perfection ^ In Electricity, the hypothesis of fluids is rapidly yieldiuj; before the rational idea of Polarity. — J. P. N.
248 POSITIYK PHILOSOPHY.
of our knowledge of them, and their mutual dejiendence. Under this rule, we shall find that the departments that offer themselves first border upon Astronomy, and those that come last upon Chemistry.
First will come Weight, in solids and fluids, regarded statically and dynamically. About this assignment, there are no two opinions; weight being absolutely imiversal. Its phenomena are simple, independent of others, and so exactly understood that science is hei'e almost as jDOsitive as in astronomy, to which it is very nearly allied. Electric phenomena being the most opposite of all to those of weight, in all these particulars, will come last; and they are closely allied to chemistry. Between them will come thermology, acoustics, and optics. Fourier put away, in his study of heat, all fantastic notions about imaginary fluids, and brought his subject up to such a point of positivity as to place it next to the study of gravity. Acoustics might, perhaps, contest its place with thermology, but for the generality of the phenomena of the latter. In regard to positivity, there is little to choose between them: but there are gaps in our knowledge of acoustics which,., also indicate the lower place for it. Our order then is, — Bai*ology, or the science of weight: thermology, or the science of heat: acoustics, optics, and electricity. But we must beware of attaching too much importance to this arrangement, which is really little better than arbitrary, though as good as our present knowledge admits. We shall now proceed to a philosophical review of them, exempt from details; having, in this chapter, analysed the proper object of Physics; the modes of investigation appropriate to it; its position in regard to the other sciences; its influence upon the education of human reason; its degree of scientific perfection: its incomplete positivity at present; the means of remedying this by a sound institution of hypotheses; and finally the rational distribution of its diiferent departments.
BAEOLOGT.
NOTWITHSTANDING the advanced state of our means for the study of Barology, we have no complete theory of weight, but only fragmentary portions of a theory, dispersed through treatises on rational mechanics or physics. It will be of great advantage to bring them together.
The division of the subject is into two r)' ' ■" principal sections, subdivided into three; the Statical and Dynamical consideration of weight, in its application to solids, liquids, and gases. Both philosophically and historically, this division is indicated.
SECTION T.
STATICS.
Taking the statics of gravity first, we must <+ f - point out that we owe our elementary notions Gravitv of positive barology to Archimedes. He first clearly established that the statical effort produced in a body by gravity, — that is, its iveight, — is entirely independent of the form of the surface, and „.. depends only on the volume, as long as the nature and constitution of the body remain unchanged. This may appear to us very simple; but it is not the less the true germ of a leading pi-oposition in natural philosophy, which was perfected only at the end of the last century; namely, that not only is the weight of a body independent of its form, and even of its dimensions, but of the mode of aggregation of its particles, and of all variations which can occur in their compositiou, even bv the diiferent vital operations; in a word, that this quality is absolutely unalterable, except by the circumstance of its distance from the centre of the earth. Archimedes could take none but geometrical circumstances into the account; but, in this elementary relation, his work was complete. He not only discovered that, in homogeneous masses, the weight is always in pr02)ortion to the volume; but he disclosed the best means of measuring, in solid bodies, by his famous hydrostatic principle, the specific co-efficient which enables us to estimate, according to this law, the weight and volume of the body, by means of each other. We owe to him too the idea of the centi-e of gravity, together with the first development of the corresponding geometrical theory. Under this view, all problems respecting the equilibrium of solids are included in the domain of rational mechanics: so that, except the important relation of weight to masses, which could be fully known only to the moderns, Archimedes ought to be regarded as the true founder of statical barology, in relation to solids. There is, however, another leading idea which was not clear in the time of Ai'chimedes, though it became so, soon afterwards; that of the law of the direction of gravity, which men spontaneously considered to be constant, and which the school of Alexandria ascertained to vary from place to place, always being perpendicular to the surface of the terrestrial globe; a discovery which is evidently due to astronomy, by which alone the means ai-e offered of manifesting and measuring, by comj^arison, the divergence of verticals.
The ancients had no accurate ideas about liauids' the equilibrium of liquids; for Archimedes contemplated only the equilibrium of soHds sustained by liquids. His princijile did not result, as with us, fi'om an analysis of the pressure of the liquid against the vessel containing it, thus disclosing the total force exercised by the liquid in sustaining the weight. The theory of gravitating liquids is to be ascribed to the moderns.
The mathematical characters of fluids is that their molecules are absolutely independent: and the geometrical character of liquids is that they are absolutely incompressible. Neither of these is strictly true. The mutual adherence of fluid molecules now forms an interesting section of physics; and, as for the compressibleness of liquids, it CASE OF LIQUIDS. 251 was indicated by several phenomena, — especially the transmission of sound by water, — and it is now proved by unquestionable experiments. The contraction as yet producible is very small; and we do not know what law the phenomenon follows, in its relation to variation of pressure. But this uncertainty does not affect the theory of the equilibrium of natural liquids, owing to the extreme smallness of the condensation. In the same way, imperfect fluidity is no hindrance, provided the mass has a certain extension. We may therefore put aside these exceptions, and proceed to consider the equilibrium of gravitating liquids, in the two cases in which they are studied: in a mass so limited that the verticals are pai'allel, which is the ordinary case; or in a great mass, as that of the sea, in which we have to allow for the variable direction of gravity.
In the first case, there is clearly no difficulty Fir-t ca about the surface: and the whole question is of the pressure against the enclosing vessel. Stevinus, following the principle of Archimedes, showed that the pressure upon a horizontal boundary, or floor, is always equal to the weight of the liquid column of the same base which should issue at the surface of equilibrium: and he afterwards resolved into this the case of an inclined boundary, by decomposing it into horizontal elements, as we now do. From this it appeared that the pressure is always equal to the weight of a vertical column which should have the proposed boundary for its base; and for its height that of the surface of ecjuilibrium above the centre of giuvity of this boundary. According to that, the infinitesimal analysis enables us easily to calculate the pressure against any definite portion of any curved surface. The most interesting physical result is the estimate of the total pressure supported by the whole of the vessel, which is necessarily equivalent to the weight of the liquid it contains. The equilibrium of floating bodies is only a simple application of this rule of measurement of pressure. The immersed part of the solid is a boundary; and it is clear that the pressure of the lici[uid to sustain the body is equivalent to a vertical force equal to the weight of the displaced fluid, and applied to the centre of gravity of the immersed portion. This rule is precisely the principle of Archimedes. The main 252 POSITIVE piiiLOSopny.
problem was geometrically treated by him. The only really difficult research, in this matter relates to the conditions of stability of this equilibrium, and the analysis of the oscillations of the body floating round its stable position; and this is one of the most complex applications of the dynamics of solids. If the question was of the vertical oscillations of the centre of gravity, the study would be easy, because we can estimate at once the way in which the pressure increases as the body is furtlier immersed, and diminishes as it rises, tending always to a return to the primitive state. But it is otherwise with oscillations from rotation, whether of rolling or pitching, the theory of which is a matter of much interest in naval art. Here, the mathematical difficulties of the problem can be met only by abstracting the resistance and agitation of the liquid; and the labours of geometers become merely mathematical exercises, of no practical use.
o 1. The question of the equilibrium of the vast liquid masses which compose the greater part of the earth's surface is clearly connected with the general theory of the form of the planets: but difficulties, unconnected with the figure of the planets, intervene, and cannot be entirely surmounted. Rational hydrostatics shows us that equilibrium is possible when there is the same density at all points equally distant from the earth's centre; a condition which is impossible under our variety of temperatures in different positions. There is no rational result from any practicable study of currents, of varying temperatures, of the compressibleness of liquids, all of which, tliough following unknown laws, are necessary to the solution of the problem. We have no better resource, at present, than in empirical studies: and these, which belong more to the natural history of the globe than to physics, ai-e very imperfect.
The theory of tides will hereafter, when sciences and their arrangements are more perfected, take its place in the department of barology. The periodical disturbances of the equilibrium of the ocean are a proper subject for study in connection with terrestrial gravity; and it can make no difference that the cause of those perturbations is found in the planetary system.
CASE OF GASES. 253 CASE OF GASES. 253 lu studying the third question of equili- ^, ^ brium, that of gases, we meet with a difficulty ' o ■ • • which does not pertain to that of solids and liquids; — we have to discover the gravity of the general medium in which we live. In the case of liquids, we have only to weigh an empty vessel, and then the same vessel filled; whereas, in the case of the atmosphere, a vacuum cannot be created but by artificial means, which must themselves be founded on a knowledge of the weight of the medium to be weighed. The fact can be ascertained only by indirect means; and those are derived from the theories of j^ressure which we have been noticing. Stevinus was not thinking of the atmosphere in elaborating his theory; but, as it answei'ed for heterogeneous liquids, it must answer for the atmosphere. From that date, the means for ^j., ascertaining the equilibrium of the atmosphere, in a positive manner, were provided. Galileo projected the work, in his last years; and it was well executed by his illustrious disciple, Torricelli. He proved the existence and measurement of atmospheric pressure by showing that this pressure sustained different liquids at heights inversely proportioned to their densities. Next, Pascal established the necessary diminution of this pressure at increasing heights in the atmosphere; and Guericke's invention of the air-pump, an inevitable result of Torricelli's discovery, gave us direct demonstration in the power of making a vacuum, and consequently of estimating the sj^ecific gravity of the air which surrounds us, which had hitherto been only vaguely computed. The creation and improvement of instruments of observation is an invariable consequence of scientific discovery; and, in this case, the fruits are the barometer and the air-pvxmp.
One condition remained, before we could apply the laws of hydrostatics to atmospheric equilibrium. We had to learn the relation between the density of an elastic fluid and the pressure which it supports. In liquids (supposed incompressible) the two phenomena are mutually independent; whereas, in gases they are inevitably connected: and herein lies the essential difference between the mechanical theories of the two fluids. The discovery of this elementary relation was made about the same time bv Mariotte in France and Boyle in England. These illustrious philosophers proved by their experiments that the diifereut volumes successively occupied by the same gaseous mass are in an inverse ratio to the different pressures it receives. This law has since been verified by increasing the joressure to nearly that of thirty atmospheres; and it has been adopted as the basis of the whole Mechanics of gas and vapours. But we must beware of accepting it as the mathematical expression of the reality; for it is evidently the same thing as regarding elastic fluids as always equally compressible, however compi-essed they may already be; or, conversely, as always equally dilatable, however dilated they may already be: — suppositions which cannot be indefinitely extended. But thus it is, more or less, with all the laws we have ascertained in our study of nature. They approximate, within narrow limits, we must suppose, to the mathematical reality: but they are not that reality itself, even in the grand instance of gravitation. These laws are sufficient for our use and guidance; and that is all the result that positive science pretends to.
,.^.. Under the law of Mariotte and Boyle, the the problem theory or atmospheric equilibrium tails into the department of Rational Mechanics. We see at once that the air can no more be in a state of real equilibrium than the ocean; and so much the further from it as heat expands air more than water. Yet we must conceive of the partial equilibrium of a very narrow atmospheric column, to form a just general idea of the mode of diminution in regard to the density and pressure of the different strata. Putting aside considerations of heat, and the small effects of gravitation in such a case, we see that density and pressure must diminish in a geometrical progression for altitudes increasing in arithmetical progression: but this abstract variation is retarded by the diminishing heat of the loftier atmospheric strata, which makes each stratum more dense than it would be from its position. Here, therefore, we are stopped by the intervention of a new element which we do not understand, — the law of the vertical variation of atmospheric temperatures— our ignorance of which can be sup])lied only by inexact and uncertain expedients. Great caution is neces- DYNAMICS OF BAROLOGY. 255 sary in using Bouguer's method of measuring altitudes by the barometer; a method very ingenious, but dejiending on such complex and uncertain conditions, and requiring sometimes so much delay, that it is even preferable, when circumstances permit, to enter upon a geometrical measurement, which has so greatly the advantage in certainty. Yet, considered by itself, the method of measurement by the barometer is valuable for its coutributions to our knowledge of the surface reliefs of our ^lobe.
DYNAMICS.
We have now to consider the laws of notion of gravitating bodies; and of Solids, in the first place.
The last elementary notion about gravity, — that of the necessary proportion between weight and mass, — which was still wanting to statical barology, was established by the admirable observation that all bodies in a vacuum fall thi'ough the same space in the same time. Proceeding from this, we must examine the discovery of the laws of motion produced by gravity. We shall find „.
herein, not only the historical origin of physics, but the most perfect method of philosophizing of which the science admits. Aristotle observed the natural acceleration of bodies in their fall; but he could not discover the law of the case, for want of the elementary principles of rational dynamics. His hypothesis, that the velocity increases in proportion to the sj^ace traversed, was plausible, till Galileo found the true theory of varied motions. When Galileo had discovered the law that the velocity and the space traversed were necessarily proportioned, the one to the time, and the other to its square, he showed how it could be verified in two ways; by the immediate observation of the fall, or by retarding the descent at will by the aid of a sufficiently inclined plane, — allowance being made for the friction. An ingenious instrument, which affords a convenient A^erification, was afterwards offered by Attwood: it retards the descent, while leaving it vertical, by compelling a small mass to move a very large one in that direction.
By this one law of Galileo, all the problems relating to the motions of falling bodies resolve themselves into questions of rational dynamics. They, indeed, compelled its formation, in the seventeenth century; as, in the eighteenth, questions of celestial mechanics thoroughly developed it. In all that relates to the motion of translation of a body in sjiace, this study is due to Galileo, who established the theoi'y of the curvilinear motion of projectiles — allowance being made for the resistance of the air. All attempts, liowever, to ascertain the effect of this resistance have hitherto been in vain; and therefore the study of the real motion of projectiles is still extremely imperfect.
As for the motions that gravity occasions in bodies that are not free in space, — the only important case is that of a body confined to a given curve. It constitutes the problem of the pendulum, which we have already considered, as the immortal achievement of Huygheus. Its pi-actical interest, as the basis of the most perfect chronometry, presented itself first: but it has besides furnished two general consequences, very essential to the progress of barology. First, it enabled Newton to verify the proportion of weight to mass with much more exactness than would have been possible by the experiment of the fall of bodies in a vacuum; and, in the second place, the pendulum has enabled us, as we saw before, to observe the differences of the imiversal gravity at different distances from the earth's centre. It is by the use of this method that we are continually adding to our knowledge of the measure of gravity at various points on the globe, and, therefore, of the figure of the earth.
In these different sections of dynamical barology, solid bodies are regarded as points — considerations of dimensions being discarded: but a new order of difficulties arises when we have to consider the particles of which the body is really formed. With regard to cases of restricted motion, as the pendulum, the thing to be done is to find out under what laws the different points of the body modify the unequal times of their respective oscillations, so that the whole may oscillate as one point, real or ideal.
CASE OF FLUIDS. 257 This law, discovered by Huyghens, and afterwards obtained by James Bernouilli iu a more scientific manner, easily transfoi'ms the compound into the simple pendulum, when the moment of inertia of the body is known. The study of the pendulum is involved in all the questions of the dynamics of solids. To give it the last degree of precision, it is necessary to consider the resistance of the air, though that resistance is small in comparison with the case of projectiles. This is done, with ease and certainty, by comparing theoretical oscillations with real ones exposed to the resistance of the air; when, of course, the difference betweeu the two gives the amount of that resistance.
We have seen enough of the difficulties of y, ■,.
hydrodynamics to understand that the part of dynamical barology which relates to fluids must still be very imperfect. In the case of the gases, and especially of the atmosphere, next to nothing has been attempted, from the sense of the impi-acticable nature of the inquiry. The only analysis which has been proceeded with, p,,.., in regard to liquids, is that of their flow by ' very small orifices in the bottom or sides of a vessel: that is, the purely linear motion, mathematically presented by Daniel Bernouilli, in his celebrated hypothesis of the parallelism of laminae. Its principal result has been to demonstrate the rule, empirically proposed by Torricelli, as to the estimate of the velocity of the liquid at the orifice, as equal to that of a weight falling from the entire height of the liquid in the vessel. Now, this rule has been reconciled with observation, even on the supposition of an invariable level, only by an ingenious fiction, suggested by the singular phenomenon of the contraction of the liquid filament. The case of a variable level is scarcely entered upon yet; or any which involves the form and size of the orifice. As for more complex cases, — their theory is yet entirely in its infancy.^ ' The subjects here spoken of by M. Conite have recently received remarkable elucidation through experiments entered ui)on chietly by the instrumentality of the " liritish Association for the advancement of Science." There cannot be much liope, in the present state of our knowledjje, of ascending otherwise than empirically towards any general law or rules which will either comprehend phenomena, or be of use in practice. — J. P. N.
I. s From this cursory review of Barology we may carry away some idea of its spirit, and of the progress of which it admits. Imperfect as our survey has been, we may perceive that this first province of Physics is not only the purest, but the richest. We may observe in it a character of rationality, and a degree of co-ordination which we shall not meet with in other parts of the science. It is because we look for a consistency and precision almost like those that we find in astronomy that we consider barology so imperfect as we do. It has long attained its position of positivity: there is no one of its subdivisions which is ^.,., ^ not at least sketched out: all the general ^f n^r-^f^rnr means or investigation, Observation, Jiixpenment, and Comparison, have been successively applied to it; and thus its future progress depends only on a more complete harmony between these three methods, and on a more unifomi and close combination between the mathematical spirit and the physical.
THERMOLOGT.
THERMOLOGT.
IVTEXT to the phenomena of gravity, those of heat are, " ^ nnquestionahly, the most universal in the province of Physics. Throughout the economy of terrestrial nature, dead or living, the function of heat is as important as that of gravity, of which it is the chief antagonist. ^,.. The consideration of gravity presides over the geometrical and meclianical study of bodies: while that of heat prevails in its turn, Avhen we investigate deeper modifications, relating to either the state of aggregation or the composition of molecules: and finally, vitality is subordinated to it. The intelligent apj^lication of heat constitutes the chief action of man upon nature. Thus, after harology, thermology is, of all the parts of physics, the one which most deserves our study.
The earliest scientific observations in thermology are almost as old as the discoveries of Stevinus and Galileo on gravity, the invention of the thermometer having taken place at the beginning of the seventeenth century: but, owing to the complication of its phenomena, it has always been far distanced by barology. At the end it ■ i • t of the seventeenth century, the indications of the thermometer could not be compared, for want of two fixed points,, the necessity for which was, at that time, nointed out by Newton. The greatest difference between the two studies, however, was in their spirit. Wliile philosophers were already inquiring, with regard to weight, what were its phenomena and their laws, those who were studying heat were looking for the natui-e of fire, and reducing the facts of the case to something merely episodical. It was only at ihe end of the last centurv, that the great discovery of Black imparted anvthing of a scientific character to thermology, while barology was almost as much advanced as it is now. Our philosophers still entertain some of the old chimeras; but now very loosely, and, as they say, to facilitate the study of the phenomena. The labours of Fourier, however, must soon establish a thoroughly scientific method; and this result cannot but be aided by the fact that the two great modern hypotheses about the nature of heat are in direct collision. It is certain that, of all the provinces of physics, thermology is the nearest to a complete emancipation from the anti-scientific spirit.
Of all the branches of Physics which admit Mathematics Mathematical Analysis, this is the one which exhibits the most sj^ecial application of it. Barology enters into the province of rational mechanics; and so, in a less degree, does acoustics. The analytical theory of heat now offers a scientific character as satisfactory as that of gravity and sound; and it may be treated as a dependency of Abstract Mechanics, without any resort to chimerical hypotheses. Our present business is, however, with the purely physical study of Heat.
MUTUAL THERMOLOGICAL INFLUENCE.
rp ^,. Physical thermology consists of two parts, distinct, but nearly connected. The first relates to the mutual infiueuce of two bodies in altering their respective temperatures. The second consists in the study of those alterations: that is, of the modifications or entire change which the jihysical constitution of bodies may undergo in consequence of their variations of temperature, stojjping at the j^oint at which chemistry intervenes, — that is, where the molecular composition becomes affected. The first of these consist in the theory of warming and cooling.
^.,,^,, No thermological effect is produced be-First: Mutual, IT,P • 1 XT influence tween bodies or precisely the same temperature. The action begins when the temperatures become unequal. The warmer body then raises the temperature of the cooler; and the cooler depresses that of PROPAGATION OF HEAT. 261 the warmer, till, sooner or later, tliey reach a common temperature. Though the final state may usually be at an unequal distance from the two extremes, action is not, properly speaking, the less truly equivalent to reaction in a contrary direction. This case again divides itself into two. The bodies may act at a distance, greater or smaller; or they may be in contact. The first case constitutes what is called the radiation of heat.
The direct communication of heat between two isolated bodies was long denied l;>y philosophers, who regarded the air, or some other medium, as indispensable to the effect. But there is now no doubt about it; as thermological action takes place in a vacuum: and the small density and conducting power of the air could not account for the effects observed in the majority of oi'dinary cases. This action, like that of gravity, extends, no doubt, to all distances, in conformity to the fundamental approximation between these two great phenomena, pointed out by Fourier: for we can conceive of the jilanets of our system as exerting an appreciable mutual influence in this respect: and it seems as if the temperature of the whole solar system were attributable to the thermometrical equilibrium to which all the parts of the universe ai-e for ever tending.
The first law relating to such an action con- t> v ■ sists in its rectilinear propagation. Though Heat the term radiation has been connected with untenable hypotheses, we may retain it, provided we carefully restrict it to the meaning that it is in a right line that two points act thermologically on each other. It thus implies that in placing liodies to prevent this mutual action between two others, the absorbing body must be placed in a right line. — -This radiating heat can be reflected like light, and in conformity with the same rule: and it undergoes the same refractions as light, with some modifications.— Another question about this action relates to the influence of the direction of radiation considered in regard to the surface of either the warming or the warmed body. The experiments of Leslie, confirmed by the mathematical results of the case, have established that, in either case, the intensity of the action is greater as the rays approach the perpendicular, and that it varies in j^roportion to the sine of the angle that they form with each siarface.— The last consideration, and the most important, is the difference of temperatures between the two bodies. When this dift'ereuce is not very great, the intensity of the action is in precise proportion to it: but this relation appears to cease when the temperatures become very unequal.
When the radiation is not direct, but transmitted through an intermediate body, the conditions just noticed become complicated with new circumstances relating to the action of the interposed body. We owe to Saussure a fine series of experiments — hardly varied enough however, — upon the effect of a set of ti-ansparent coverings in changing remarkably the natural mode of accumulation or dispersion of heat whether I'adiant or obscure. More recently, M. Melloni has pointed out an essential distinction between the transmission of heat and that of light, in proving that the most translucent bodies are not always those through which heat i^asses most easily; as was previously supposed, p _,.■ It is well for physicists to separate the bv Contact radiation of heat from its propagation by contact, for analytical purposes; but it is evident that the separation cannot be found in nature. They are always found in connection, however unequally. Besides that the atmospliere can hardly be absent, and is always establishing an equilibrium of heat between bodies that are apart, it is clear that it is only the state of the surface that can be determined by radiation. The interior parts, which have as much to do with the final condition as the surface, can grow warmer or cooler only by contiguous and gradual propagation. Thus, no real case can be analysed by the study of its radiation alone. And again, the action by contact of two bodies can take place only in those small portions which are in contact, while the bodies are acting upon each other by the radiation of all the rest of their surfaces. Thus, though the two modes of action are really distinct, the analysis of either is rendered extremely difiicult by their perpetual coml>iiiation.
Of the three conditions noticed above, relating to the action exerted at a distance, the only one which applies equally to the ])roj)agatioii of heat by contact is the difference of temperatures. The temj^erature of the parts in COXDUCTIBILITV. 263 question can have so little inequality, that the law which makes the action increase in jjroportion to the difference may be regarded almost as an exact expression of the fact. As for the law relating to direction, it probably subsists here too, though we cannot be perfectly assured of it. But that law which relates to the distance must be totally changed; for, on the one hand, the action of the almost contiguous particles cannot be nearly so great as the variations which we are able to estimate would lead us to suppose; and, on the other hand, when we compai'e the various small intervals between them, the diminution is certainly much more rapid than in the case of distant bodies.
Whatever may be the mode of the warming and cooling of respective bodies, the final state which is established under the laws just noticed is numerically determined by three essential coefficients, proper to every natural body, as its specific gravity is in barology.