Pauli Lectures on Physics 1. Bleetrodynamice 2. Optice and the Theory of Electro 4. Thermodynamics an the Kinetic Theory of Gases 4. Stati Mechaniea 5, Wave Mechanice 6. Selected Topi in Field Quantization Pauli Lectures on Physics: ‘Volume 3. Thermodynamies and the Kinetic Theory of Gases Wolfgang Pauli ated by Charles P. Ene ‘Translated by H.R. Lewin and S. Margulies Foreword by Victor F. Weinskopf ‘The MIT Press Cambridge, Mauachoscts, and London, EnglandConse 193 by ‘The Machete Tae of Techalory of his bok may be eprodaced in any form Contents Foreword by Victor F. We opt Preface by the Eiltor 1. Basie Concepts and the Fist Law 1, Teemadynumie variables 5 Qutmy ot bet 4 Tin iw of thermodyneaes 5 Themodyanisthge of sate {Mathematica oration the Iw 4. Aplialoso the t e 2. The Second Law ermal of the sod we 4. Qoattatvepredion the ern ae 16 Remark onthe miting tvs 3. Equilibria 1 Van Holt exes 1 Cvs aration meted sears ss18. Applications fe variation) £0. Commons onthe ond 21, Thermocttie problems 4, Nests Heat Theorem Bibliography Appendix, Comments by the Editor Index Foreword [Rin often said that siomtfic texts quickly become obsolete Why ate the Pauli Ieetres brought tothe public today, whem some of them ere given as long as twenty years ago? The imple: Paulie way of presenting physics is never ‘out of date His famous article on the foundations of quantum mechanics appeared in 1933 in the German encyclopedia Handbuch der Physik. Twenty-five years Inter it reappeared practically unchanged in a new edition, whereas most other ‘contributions to this encyclopedia had to he completely re- written, The reason for this remarkable fact Ties in Paul's style, which is commensurate to the greatness of its subject is clarity and impos, Style in viene writing ba quality that today ison the point of vanishing. The pressure of fat publication isso great that people rush into print with hi idly written papers and books that show little concern for ful formulation of ideas. Mathematical and instrumental techniques have become complicated and dificlt; today most ofthe effort of writing and learning is devoted to the aequ tion of these techniques instead of insight into important concepts. Essential ideas of physics are often lost in the dense forest of mathematical reasoning. This situation need not be 10. Paul's lectures show how physical ideas ean be presented cleaely and in good mathematical form, without heing hidden {in formalin expertise. Pauli was not an accompli 1d lecturer inthe technica senseof the word. t was often dificult to follow his courses. But when the sequence of his thoughts and the structure of his logie become apparent the attentive follower i eft with anew insight into the splendid architecture of reason, which is theo- retical physics. The value ofthe leture notes not diminished by the fact that they were written not by him but by some of his collaborator. They hear the conceptual structure and hhere and there does one miss words and comments of the mas- ter. Neither does one notice the pasing of tan in with the sole exception ofthe lestures on fleld quantization, in which some concepts are formulated in x way thet may appear olfashioned to some today. But even these lectures should be of use to modern students because oftheir compact: ress and their direct approach to the central problems ‘May this volume verve as an example af how the concepts of ‘theoretical physi were conceived and taught by one of the reat men who created them, Victor F. Weisskopf Cambridge, Mauachusets Preface ‘This is conventional course on phenomenological thermo: Alynamies. As Pauli says in the introduction, time docs not appear as variable inthis feamework, except for its direction. ence the subject limited to the equilibrium theory (statis), and ireverible thermodynamics not discussed. But this con. ‘ventional subject is treated by Pauli i the same inductive way 4 the letures on electrodynamice i this series, with the em ‘phasis given tothe historic development andthe logeal truce ture ofthe theory Its for this reason that Paull goes through the labor of Carnot eyes and van't Hoff boxes, The axiomatic formulation is only given a an illustration. 7 his life, he chemical reactions and, as mentioned in the appendix, paper arew out of this occupation with his course. This paper was include in the second German edition, published in 1958, on hich this English translation is based. This edition was an Improved version of notes worked out by student, E.Jucker, and published in 1952. ‘Since Pauli taught this course at BTH in Zvi, equilibeium thermodynamics has again become an active feld of research, Indeed, phase ions are described today by sealing las iularites of the thermodynamic quan tities. This fascinating new field is beautifully described invery recent hook by H. E, Stanley, Introduction to Phase Trane sition and Critical Phenomena (Onford University Press, New York, 1971). Furthermore, new techniques wich a+ Raman scattering are applied to investigate thermodynamic system, and new types of substances such a igud erystals have become Important. Therefore a solid bass of thermodynamics as given in theseIetures is again more important today than at Pals ‘The lst chapter, om kinetic theory of gaten ie logically d= ‘connected from the ret ofthe course, It actually belong to the first chapter of the letures om tatstieal mechanics in thi series The eplit in Paul's Ieetures corresponds tothe iaterrap- tion between term Although the editing of this course did not poue particular problems, the work ofthe translators deserves special acknowl. ccdgment. (Charles P. Ent Geneva, 18 November 1971 Pauli Lectures on Physi ‘Thermodynami ind the Kinetie Theory of GasesChapter 1. Basic Concepts and the First Law Cassical thermodynamics foregoes detailed pitares and, ‘therefore, makes only general statements concerning the cnergetios of heat transfer Te limite itself to states of equ Iibeium and to very slowly oeuring processes, No quan: tity withthe dimension of time appear in thermodyaamies; st the most, time enters via the concepts of "before and Satter” ‘Therefore, in the ease of rapidly occuring pro esses, only initial and final states are disenseed. In theemo- ‘dynamics no considerations concerning the nature of heat fare made. This problom is fist dealt with in the kinetic thoory of gus Definition: Thermodynamic variables are measurable mae- roseopie quantities which characterize a cystam, Beampler; Pressure p, volume V, surface tension, eur face aren A, atres tensor 8, strain tensor fy magnetira- tion Mf, magnetie fel intensity 1, concentration ¢, num ber of moles ‘The temperature ¢ of system must first be defined. Experiment shows that if system is closed, then heat is ‘exchanged within the aystem until a atable thermal state is reached; this etate it known as thermodynamie equlrium.Heat exchange through a heat-conducting connection makes thermodyzamic equilibrium between two systems possible. ‘Th, we may say that two systoms have the same tempo lature I they are in thermadynamie opuilibriam (with one ‘another. “A system in thermodynamic equilibrium possesses one less Agree of freedom [A-I],1 There exists a relation [4-2] eases constant, Fe ti Bede where 2, and y, are the thermodynamic variables that ‘haraeterie the system, Tn the simplest case of « homo- geneous system this relation is 19.) {tone varinble in the above relation is held fixed, then the other voriable an arbitrary measure of temperature. For example, one sueh measure js the volume of a fxed amount ‘of material at a fixed pressure. Since the variation of ‘olume with temperature it diferent for different substan- ‘et, temperature is not defined absolutely by volume at fonstant pressure, he exact thermodynamic definition of the temperature sete in not possible without the second law of thermodynamics, Nevertheless, our tentative definition flways allows the determination of whether one tempera: tare is langer or smaller than another: Normal subetancer expand with increasing temperature. However, there exist fanomalous substances (for example, water between 0° and 1£°0), Dut these ean be recognized as such by means of the following experiment. Two samples of asubstance, A and B, ‘which do not have the same temperature are brought into thermal contact, During the subsequent heat exchange the samples experience volume changes AV, and AV, respec- tively. Hor normal substances we always have AVA <0, ‘but for anomalous subtancos the initial conditions can beso constant chosen that AV,AV,>0. Steady-state heat conduction (refer to Section) oles a further possibility for deciding ‘hich of two temperatures is larger. By such means we ob- {ain a monotonie temperature sale; that i i f,> 4 and we Introduce a nev seale through a monotonic transformation Pafl), then {>t also for the new temperatures [4-3] 2. QUANTITY OF HEAT Experimentally we find that there exist equilibria between ‘the phases of a substance: Tiguid = gas, tiguid sexo, sold segue In order to change the equilibrium between two phases, leat must be added oF taken away; this heat it called the feat of transformation (e.g, heat of vaporization oF heat of fonion). ‘Temperature and pressure remain constant during ‘the change. We can use the heat of transformation in Aehning quantity of hnt (at constant temperature. Defwition: 1p order to change the phase of m grams of a substance, n times more heat is required than is needed to ‘change the phaso of Igram, (Heat of transformation is proportional to the amount of material.) ‘This definition is independent of the determination of the ‘temperature scale, sneo the heat of transformation is added or subtracted at constant temperature, For the comparison of quantities of heat at different tem- peratures, we uae the process of steady-tate eat conduo- tion, in which two heat rejervoir with temperatures 4, and t, (o>) ate connected by a heat conductor. Defnition: Tho quantity of heat Q, given up by revervoir 1 ‘equals the quantity of heat Q, absorbed by reservoir 2‘ sot snes se rm rane | Cap. ‘The quantities Q, and Q, an alto be heats of transforma: ‘on, which then allows a direct comparison with the quan- tities of heat defined at constant temperatare, A monotonic transformation of the temperature seale does not alter this Aefinition either. ‘lstorically, quantity of heat was defined by a mixing process, We mix two quantities of a substance, of mass ‘mand my and temperatures f, and 4 (4Qx: Moreheat isaded than taken away, whereby the system does work. During the eycle, heat ie thereby con- ‘verted into work, inthe course of which the amount of teat Gy drops trom t, to 4; that ie, heat goes from a warmer reservoir to a cooler one 2. Q.@jy then we reverse both procostes and obtain the atte ease as above, ‘Therefore, Beeaune of the second law, St must be that oa) 1f the second law were untrue, that is, if heat conduction were reversible, then work could be obtained from heat without compensating changes. Since itis only possible to convert work into heat Without compensating changes, 9 | geurmmanve runes oF ri sees a a" and sinee itis only posible for heat to go from s higher to lower temperature without compensating changes, this is the preferred direction in which actual procersey take place, We ean now alto determine whieh of two temper: atures f, and ts the higher, namely, that one from whiel Tet ean flow to the other without compensating changes TAS]. Since, according to Section 4, heat whieh is added is taken 1s positive, it follows that , Arbitrary quasi-static processes ‘An arbitrary quasi-ataie eylic process can be reduced to ‘changes in a slngle heat reservoir with the help of ausil-2 sux soso ta | ap Jary—in the limiting ease, infinitely many—Carnot cycle, First, we make a decompotition into ony a nite number of individoal processes with temperatures Z,, ab which the ‘quantities of heat 8Q, are added (k— 1, 2. Z). The quan Lilie of heat 3Qu ate removed from’ ingle reservoir at temperature 7, with the aid of Camot cycles, We have Since we have completed a eel, and sinee no changes resalt, beat could neither have been adiled to nor taken ‘way from the heat reservoir at Ty. ‘Thus it must be that Qu=0 [AT], or snot, fa Dt we rte ae ew an co sf Z or quasistatic changes of stato we have ao. |e function of state (Independent of fo | counmas maser or ma mo ee wt an for eyes prosees we have 30 a $3 Reply emaring roomate 1 all remains for ur to investigate how he entopy tango during epily occuring (reverse) process ‘To that end we considera process in a closed bos, in whieh the internal energy romaine constant. Thermodynamics re YG Yi ‘qulres that the intial state be attainable fom the Anal state along @ quasi-tatie path. ‘The change in entropy ‘long this quasi-static path is ; oan[t "Hore Quis the hest removed from «reservoir at temperntare Z,, and Qy<0. Were Q.> 6, then heat would have been con verted into work without compensating changes, which con- tradjts the second law of thermodynamics. ‘Ths, §,< 8,5 ‘hat i the entropy Kar ines. Concerning the second lay, the following is to be noted: ‘The thermodynamic function of state ean also be deter- ‘mined without using ireversible processes, on whieh the” seco ta | an statements of Clausius and Thomson are bated, ‘This done by weakening the prinipls of impossibility somewhat by only stating something about quasistatic processes, Tn 4 quasistatic cycle proces 1, Heat cannot be convert into work unles, atthe same time, a corresponding quantity of heat is taken from a warmer reservoir to a colder one; 2, Heat cannot go from a eolder reservoir to & warmer one ‘unless, at the same time, a corresponding amount of work oes into heats 3, Work cannot be converted into heat unless, at the same time, a corresponding quantity of heat is taken fom a colder reservoir to a warmer one; 44. Heat eannot go from 4 warmerreservair toa colder one unless, at the same time, a corresponding quantity of heat is converted into work, [If we postulate the validity ofthe above four axioms for ‘quasi-static changes of state then an enteopy function exes, It we require that only axioms 1 and 2 are true for rapidly ‘oveurring processes, then an entropy funetion exists nev: ertheless. In a closed system, the entropy can change only ‘in one direction; it can either inerease oF it can decrease ‘However, both possiblities are never simultaneously avail: able [48] ‘A further remark eoneeras the question of whether the temperature defined by the Carnot procestes must necessa ily be positive [6]. (The same question eeurs forthe sgn of the integrating factor in the formulation of thermedy- ramies by Carathéodory (Seetion 11).| From the axiomatic ‘viewpoint, Mrs, . Bhrenfest-Afanasjewa showed that thi does not follow solely from the remaining aarumptions [Z. Physik 38, 985 (1025) and a correction in Z. Physik 34, 695 (1995)]. However, in order that this quetion not ‘be a purely formal one, the sttistioal methods (eanonial ensemble) which form the subject matter of the following 9 | gunn ratsonow or mae mcoes am ” fet of lectures mast bo appliod here. From this broadened viewpoint N. F. Ramsey showed that special syxteme can be characterized by a negative temperature under certain clrvumstances [Phyt. Kee. 108, 20 (1056) 40. MATHEMATICAL FORMULATION OF THE SECOND LAW ‘The mathematical statement of the second law is that 9/T—a8 in an ezact diferent, "This statement is no longer tied to quasl-statie procetes, 1 defines entropy in complete generality. We have ag _asecay P+ Z man TF where [A-2) B= Bley f4youy2eal aD gua =0. TE we in troduee the functions 1ae wd (E12, 090 E+) » and 2 wun 3 (5x) ‘hen we can write as= Seeds, (diferentil form of Ptah. Singe dts an exact diferential, it ls necessary that [A:2] ope” mm con ta | cap However, this notation is very inespedient, ‘Therefore, ve shall introduce other functions with more convenient Independent variables later 11. AXIOMATIC FOUNDATION OF THERMODYNAMICS FOLLOWING chsaTaeovonr® 4 Pfs linear aiferential forme et the functions Xi 4)..4,) be given. Tn order that ‘he expression (Pfat?'s form) 3 ley yy tea be the exact diferential af of a funetion fle, the conditions rast be flied for all ¢ and. 8 » consequenoe of ee Ba, Beda,” Bas, 11 80 i not an exact dient then the case when fan tom t(a, 4459 4) (the reiprocal of whieh 36 called an Sntegrating jator) can be introduced to make Wop Ban ny tan exact diferntial i of especial interest in thermodynamn- Jes. The conditions for the exiatenco of toch an integrating factor fr are given by aun ay Ay 2p oy "The fanetion ¢ ean be liminated by forming the sum (au + 0) + aR) = 0. ‘where, for example, (ay = ex fd 2) ‘This condition must be fale fr all posible triples Uf It cam be shown that these conditions forthe existence of fan integrating factor 1/r are aio suficient. or two vari ables this condition is not applicable. Now, we want to consider those curves 8Q-= ¥ ada, whieh pase through ahead: f fixed point. Two eases must be distin 1. If there exists an integrating factor (v0), then it follows from 3Q-=rdf that, when 89=0, then df= 0 also; his implies that f(x, 245.0) 5)- constant, Accordingly, all curves 890 lie in an (s-~1)-dimensional hyperplane through the fixed point. Thus, in the neighborhood of the fixed point there exist points arbitrarily close whieh do not Len the hyperplane and, therfore, cannot be reached from ‘the fixed point along eurves 390. Since the fixed point may be chosen arbitrarily, the folowing statement is tru’ tan integrating factor exists (+0), then, for every point = in the w-imensional spac, there are points arbitrarily close ‘to 2 which cannot be reached along curves 3Q=0. 2. I there does not exist an integrating factor, then all neighboring points can be reached along curves 300. 1 order to prove this statement, we use the fact that the ‘Pf linear diferential forms can be brought into the nor- ral form BQ = ade, t days 2 dayton Heidt Eden,* me weconm tan | ap ‘if the number of variables is odd (2-41), and BO made ndeyt eda a tebe if the number of rariables ia even (2n) (that is here E is eto), We will carry out the proof for (2n-+1)" variables, Let Pct fy ny Hays) be the fixed point and Pa, 2 1 in) the point t0 be reached along curves 89 =0. Let the running coordinates be 25555 y new Fits Wh 24585445 Bess Bo comitonty that iy mya ty Semty ny fer Bays (Chereore BQ 0), and go from P* ‘tw the point Pia, yy Eins Bey Zou Whote coordinates Fas Bus Bay od satisfy the condition sot —a) + Blot a) + +8 oa) + Mthea Now, weleare 24 Hy Constant that Faso Zag™ but ad G0 along the murfce defined by Alea) + Beles) Balter thd + Mees from P to P* (ey yyy Boy Ee thvaly i whieh case again 30=0. Lastly, We agin leave 25 54m Sn onstaRy hat iy a=, Fy 355 F— yy aeee— Dogs ANA KO from P* to Poy, 2) 25) y faa in Which case 30 0 alo, Consequently, P’ canbe reached from P* along carves 3Q-= 0, It everywhere in the above proot zy ie Istt out and & 4s set equal to zero, wo obtain the analogous proot for 2 variables, 2, Applicaton to thermodynamicn We need the following concepts 1, Adiabatic wats: Changes of state are possible only through mechanical means. Adiabatic wale do nat conduct eat 2. Heat conducting wale: All ronadiabatic walls are heat conducting, 8. Quav-tatic changes of state: These changes of state are ‘very slow, infinitely slow in the limiting ease, ao that the {intermediate states form a continuous sequence of equiib- slum states 4, Nonoquiibrium changer of slate: All non-quai-statio changes of state are nonequilibrium (rapidly occurring). ‘The quasistatic changes of sate of a closed system within suiabotio walls, thats, the adiabatic quua-tatie processes, lead tothe existence of adiabatic curves, along which 890, or quasi-static changes of state within heateondueting wall that i, for processes involving heat conduction, there ‘thermal equilibrium between the inside and outside. Tf ‘oro anbatancnt are separated by a heat-condueting wall ‘thon there exist a relation [4-2] F.¥.B Pio Up pvige oo Meat-condueting wall ‘This relation must have the form t(p, ¥)=H(R,P) for the following reason: If (p, Vi) and (py, V) are diferent states fof one aubtance, and if (Pu, Fa) and (Day Fe) are diferent Sates of the other substance, then experiment shows that ‘the three conditions [4-2] Fie FrPy hi =0 Fen Fab F=0 | imply Fy YB P=0. Pep FnPu =OHowever, this i possible only if the relation JO has the form tp, ¥)~i(7, P)= 0. In order to formulate tho fst law, we ean now restrict ourselves to adiabatic but not necesarily quas-static changes of state. For auch a process we deine for. Hore, it isto be noted that there always exitts an adiabatic [4:9] path which either leads from 1 to 2 or from 2 to 1 For adiabatic [4:9] changes of state, we have WaSyden and aB +30 -B, for homogencous substances AE +paV—0, Therefore, the equation for adiabatie changes of a homogeneous eub- anlar. where the temperature function t(p,¥) 1s sill completely arbitrary and need not be monotonic. For two homoge: ‘ous substances which are suparated by a heat conducting wal, that i, which are in thermal equilibrinm, the equae tion for adisbatic changes is (2) or fr} Y BVI | BGP fet | axtane oon rusomne canaratnony nL For nonadiabatic changes of stato, that i changos in volving heat eondustion, we deine pee "This definition determines the unit of heat. If the tem- perature is held xed, tip, ¥)=constant, then we obtain ‘the equation of the bothers. Nonadiabatic changes of sate can always be traced back to heat conduction, ‘For the second law, Carathéodory replaces the formals tions of Clausins and ‘Phomaon with the following require ‘ment which st fire ball be valid only for quas-statiechanges fof state: For arbitrary initial states, there exist neighy Doring states which eanzot be reached from the initial state bby means of quast-tatic adiabatic changes (39=0). Wecan ‘loo say that there exitt neighboring states which cannot bbe reached from the initial state along adiabatic curves. ‘Thus, according to Carathéodory, the existence of states ‘which cannot be reached adiabatically implies the exist- nee of aa integrating factor. For a aingle homogeneous tubstanee, the equation for adiabatic changes is (Boas |G)+rjar-o Since a linear diferential form with two variables can al- ‘ways be written with an integrating factor, it follows that (Ge) +(e) a)” where r=1( 7) and o=o(t ¥). ‘For two Homogeneots tubstances which are in thermal ‘equilibrium, the equation for adiabatic changes of the con bined system fs, as given above, (a) Galas ler) (=) a}an-0: om In Si non art he nna mane cua weban sserata n= (Bae (28) saan, sub rodat ro (ars (@B) nan awe, sing to hor, an intotig ter tote compet nie mar ts 18 Wet FF ~[(e)+ Galle) from which it follows that Jove[@) saan, Fg = 20a + 2.00, mj By introducing new variables f, , and oy instead of the variables 1, Vi, and Ve, we Bnd ae en ba Boe ‘From this we see that o is independent oft and, further, that r/r and 1,7 are likewise independent oft. ‘Therefore, n= MEI), = MOE» + =HZe1,00) 113] 1 remains only to show that Zie,,09)= Zio). If we sub atitate the expressions given by Iq, {11.2} into Bq, (11.1) we obtain Zde=2,do,+S,dey, trom which i follows that Wojto)E=E, and (2oltodE=E. If we now differentiate with respect to a, and ay, respec- ‘This quantity is exactly the Jacobian, and the fact that 1 equals zero implien 2oy.0))= Zio) We now define the thermodynamic temperature to within ‘8 conitant maltipliative factor [4.6], as the reciprocal of ‘he integrating tuctor: so=7 In adaition, we make the following further deiaitions: AS=Holds, a= Zloiido,, ASy—FAo)de,, 8 -f@ ao + constant, = [Blood conan, ae -[2tede.+ constant. We then have SOSA and 39— Tas Tas, 7a5,, ‘These two expressions are exactly equivalent to the sec- fond law, We ean thus aay that i two system are i ther- ‘modynaimie equilibrium, then the seoond law requires the cexatence of an integrating factor. If the existence of adiabatically unattainable states is requited also for rapidly occurring adiabatic changes ofoo ta | aap 2 stato, then the following statement is valid: For all pos- sible changes of state (rapid and quasistatic) in a closed system the entropy can only increase or decrease; both pos sibiliuies are never simultaneously aralable [A]. Let the Initial stato be specited by (V3, VS, and the final state by (Fu Ty 8). Wointroduceanintermesiatestate, Pa, Fay S, which is quasistaticaly attainable (aince AS—0). Without ‘hanging the volumes V; and V, we carry out & rapidly cceuring process (stirring or rubbing); that is, Vy, Voy ‘8%. (iy Fy 8). Ifthe entropy could both increase oF de- crease a8 a result of a rapid proces, then all states would ‘oe attainable through rapid processes, However, there must De a restriction. The only poastble one consists ia allowing only one algebraic algn for entropy changes; this means that the entropy can either inerease or decrease, but not both. ‘The conventional normalization is that the entropy fan only increase in a closed aystem (A ‘2 FREE evencr 1. Definition We define the tree energy Fas FaE-TS, aay Tt follows trom this definition that ap=an—ras—sar land, if the second law is introduced, that [4-2] ane. Sat—aW=— SATS nde, you~. (122) ‘Therefore, P= Ploy. relations Bi» = It wo substitute these relations into Eq. [12.1}, then we ny). Further, we have the = . roe, "era()-maG pen ‘The funetion F ie used when the tomperature 7 appears as fn independent Variable, and the entropy 8 is used when the internal energy # appears as an independent variable. 1b, Application to homogeneous eubetances Asoume that m=, 27, y=p) P= FIV, T) Men, ap——pav-sar, (2a and cat Fa tntn of sate, we hve (@2),--Gr) ‘This identity fa very important for homogeneous substances. From Ege. (12.3) and (12.5}, we have [A] (),-(G)-*(ézar) 7+ 7G), 029 or the speci heat at constant volume, ¢, we have [4-5] . (25) ‘and forthe specific heat at constant pressure, o, we have (Gt (G0) +(e). from these we obtain *(irhlen) 7 ee) (en), 29 eneTe must be noted that all of these relations are valid only {or homogencous substances. ‘From the fiat law we obtained an), 2% (er), +l, 1 thie oral combined witha. [25 thon we obtain exactly Bq, (12.7) Pathe, we deduce trom the sonad a Gilera, ‘Rquation (12.7 ean be Further transformed mathematically ‘Becaase of the existence of the equation of state, we ean ‘assume that V(p, 7) is given, Then, av= (3) ap (Sp) ar ‘At constant volume we have (@V/0M),—0. Therefore, (ex),~@. G8) +), an: “ ()-- gran iT) ~ eV ‘Pherefore, Hq, (127) can be written Levene (Viepe 10, cinass FUNCTION. 4, Definition or the froo energy ¥ the independent variables are 24 24y-0) Bay 2. On the other hand, the Gibbs function ‘Dis to’ a thermodynamic function whose independent var bles a6 8h, Yr Yor Yay 7. Therefore, we apply the fol- lowing Legendre transformation to _ o-?- $48), Pi Sana 2—ne+ Sen, i aonarpaws Sinan ~sars$aene (Gp) ar+Zean. sa) ‘Te terms containing ds, cancel and, therefore, we have OO Io doyonr der). Furthermore, fi Aplin tompotts etre OP pv and ab—— SAT +Vap. ‘Therefore, @= ip, 7). For quasistatic changes of state ‘at constant presture p and constant temperature 7 (so therial isobaric process), the Gibbs function is constant. urthermare, In the previous section we obtained for the sped heat at constant volume [4-5] Analogously, we obtain for the specific heat at constant&, Application to ideal gas ‘With the help of the second law, the three axiomatic prop ‘ertes of ideal gases can be reduced to two, At the stme time the agreement ofthe thermodynamle temperatare seale vith the gas scale can be demonstrated, 1, Since (@B/2¥}_~0, it follows that T(@p/2 P=AV}E. Theretore, pis 4 linear function of 7. 2, The laotherms are pV constant. por From 1 and 2 follows p¥'— (constant) Re. teal It we have a mixture of both gases in volume Voy which ix connected with a higher volume V; by an ascending tube, ‘then the light gas rises to F; and the Beary gas stayt Delow. Te is assumed that Vy and ¥; are 20 chosen that x ote [H62) Tein to be noted that Vy and ¥; can be so chosen that Bq, (16.2) is satiated, since Bq, [16.1] implies Since exjgm has a temperature 27=7-+A7 diferent from that of the first box. For ideal gases we have 8 ‘From this follows the adiabatic equation of state, stant: = eslog?— Rlogp + constant. CAB Pog SE In the following, for the sake of simplicity, we neglect quantities of order higher than A. Also, we use the wame notation as in the fist experiment. 1, We introduce 2¢ moles of Cl, and 2¢ moles of 1,0 into the box whichis at temperature 7, and we remove de ‘moles of HCL and ¢ moles of O,. The work done during the process is W,= RreSy,—eRr, 2, We then bring the gases HC! and O, adiabatically and lnothermally trom (py, 1) t0 (iy). (@) Adiabaticaly trom (pe, 2) to (Pfs) A= —c(seh + eP 7. For each individual gaa, we have Tr gAt melee 5 ~aF (@) Teothermally from (95,7) to (hy T") a ate Phe —enr (s10e He ~ ene (4196 TS + tog where the last term is approximately equal to wee + ear to be ‘Therefore, we find "(410g 28 Wyatt Mm =e (410g P+ oe + CA) 4 (eT 8. We introduce de moles of HCI and ¢ moles of O, into the box, which is at temperature 7’, and we remove 2e roles of Cl, and 2e moles of 11,0. We have Wy—~ eh" (reverse of 1, with 1° instead of 7). “f. We bring the guses Cl, and H,0 adiabatically and iso- thermally from (py) back $0 (Ps) . (a) Taothermally trom (p\, 7) to (piy 1"), where pi is 40 ‘chosen that the adiabatic curve from (Psy 2) passes through nts: won na fut enr (-2 we Mate) Pho (0) Adiabatically trom (pi, 1") to (Pos 7) Wy=- OB =~ e324) ‘For eset individual gas we again have igi? Bog wa‘Therelore, we nd W, to be ~ere (210g Epes Hla eh 4 ete — gear. ‘he total work done is Wa We We tiem een (2 Fe 4 2tog og 4. 2g tog 2 Po, HERTZ rtog lt = + ek og )—tog X(T) me Since we have carried out « Camat eytle, we have or) _auny_w a ar At constant pressure, QU) reduces to the heat of reaction: GTI=0, (@, per each » motes), fle KU") — toe a a [Ng AO ~ er Soe R(Ty, OD. By using the Carnot eye, we need not diseus the thermo- ‘dynamic function of the gas mixtare in the box. Only properties of ideal gases in relation to semipermeable men Dranes and gus mixtures are needed. Toi the concentration of gas inthe mixta a sure p, then = tree p We thon have Py where Bey og (P) = Enlogpa= Flowers loge Eo. oat | vane murs ous « ‘With the reactions tHe PHO (with y= 1 y= Twa) and Beat (with Hate 2y there are complications, becouse the reaction products de Compose im the van't Hof box. In these cases the dir ulty is often removed by the introduction of antieatalysts trhich slow down the reactions. However, this would seem {o be arbiteary and unsatisfactory (A-11]. The eyclical pro- eoces uaed by van't Hoff can, by suitable application of Ckternal force Aelds which act differently on the diferent Components, be so generuized that the introduction of Cinatable states or antiatalyst is superfiuoas ‘Let us consider, for example, the already mentioned dis: sociation of iodine: sel Tn this cage iis useful to introduce into the box a magnetic fleld of strength #, in which the para Tare magnetic jodino atoms I behave differently from the dia Imagneticfadine molecules T,, Indeed, the former have an additional free energy per gramatom of FU) FO) =~ Ato sven | Chg. 2 in which x is generally temperatare dependent, ‘This fl: lows from the property of the free energy that ar. san—pav—man ‘when the magnetization is proportional to the feld strength, Mann, ‘This means that the Gibbs function, O— ao. + PY, satisfies sar+Vap— man Jn a magnetic fed If the sources of the magnetic feld are included sem which is considered, then the work done is siven by 317 }I1dM and the thermodynamic potentials ‘which appear are PoPpM ad O=04MI Instead of P and @, If the thermodynamic functions and @ are nevertheless used, then tis unnootssary to con sider the soaroes of the magnetic eld. Tn the following Aiseusion a eyelical process will be considered in which there is constant feld strength (but variable nomber of free iodine atoms). ‘The various amounts of work HAA which appear in Uhis process cancel one another. In the left’and box (zero field strength, pressure py volume ¥, temperature 7) let there be a total of gram: toms of iodine present; of these, Ne are 1, molecules, and N20) are free T atoms (00, then, since the entropy is an increasing function (seeond law}, the change in entropy could occur spontancously; thet is the neighboring state ould be stable. Thus, we have the following law: A closed system at fixed volume and internal energy it in equilib rium when its entropy has the largest value consistent with the given volume and internal energy. If T and ¥, oF T and p, are held constant in a system Instead of B and YY, then we have the inequality sare te Banton, <0 (QPlap>0 OF (Ap, 20, respectively. ‘Therefore x closed system at fixed volume that ie held at ‘given temperature Is in equilibrium when ite tree energy ‘has the smallest value consistent, with this volume and temperature. Also, a cloved system at fxed pressure and ‘temperatare isin equilibrium when the Gibbs function has ‘the smallest value consistent with these conditions, For reversible changes of state, the equilibrium conditions are B3er= 0, BFer=0, BOK» =0 Whether these states are stable is determined By the second variation, We mutt have WBler<0, (BFne>0, BMIng>0 ‘That is, the second variations must be definite quadratic forme with the correct sign. APPLICATIONS (FIRST VARIATION) 4 Arbitrary substances nt the Gibbs function for an arbitrary substance be siven, Tf in the substance there are XN, moles of a basic” recrumny | Cup. 2 ubstance then o. 1, Pr Nay Naya Me ‘The possible variations at constant pressure and tempera ture are aM, corresponding to a reaction assumed to be possible. The fiat variation of @ is eo 0-3 (22), asieepms = OOM ee In called the hemieal potential. Because (3P)p5— (Fey flows immediately from (7, p)= FUT, ¥)-+p¥, we also GL by obtain the general equb- where Using Zrd.—0, we th ium condition rm », Homogencity proprtce “mhe tree energy and the Gibbs function are homogeneous fonctions of the frst degree in the variables V, Ny, and in the variables Ny, respectively F(T, AV, AN) = BBE, VSN O¢r, 2) = AT, PL Md ‘From this, ifthe temperature is held constant, fllows the ue elon tr F (er. Ble S018 | armucamone mar vio) ® with Pav + SandNe, we have lows us (0 sot PUL, VyN) = VHT), where m= S/F —eeiMe- (Me molecular weight of sub- stance k) For the chemical potential we then obtain m= (&).~ ln), At constant 1’ and p, the Euler relation for on Eels. = Fam, (On the one hand, and on the other hand, 80 = ZmdNes May ‘We therefore have the important relation (Gibb): Vap=ZN du (tor xed 1) 0, Heat gases Let ¥=EN, be the total molar number of a mixtare ofideal gases, Writing the partial prestare se we ott = pa (ertee for the Gibbs function of the mixture, Making use of the relation 20 mm ($2) --eatogt + Pog PBT + Be for the chemicel potential. ‘Therefore, yn p= BMA p+ MTs on, with m= NIV ~ Nh, = BT opm oD) Both f(T) and g4(Z) are functions of temperature only. 4. Semipermeable wate ‘Consider the case ofa semipermenble wall. In volume V” let there be {moles of substance 1, and in volume V" let there be 27 moles of eubstance & (k= 1,23, um) Let the wall between V’ and V" be permesble to substance 1. MN se a8 | arcamane anne vnis08) ” ‘The entire syste a to bein an infinitely large heat reservoir at temperature T. Since T and V are constant we use the tree encruy PUD, V,%), The possible variations at con stant volume and temperature are BP = ny 0,05 yO) ADEE aly Mey yy a) BNF ° Since the molecules leaving ¥" go to Y, we must have 4 N= 0. rom these two relations we obtain the condition Hm 05 0,5) = ey mb yn ‘The concept of partial pressure does not appear here. In this example the possible variation is an actual reaction. For ideal gases we obtain Jog) logs, or wins thats, xy" If we define plo REY,V" and p= RIN;/V", which ae: tually correspond to the partial presaures, we obtain rer 1, Boree fields ‘Consider now a mixture of arbitrary substances ina force field, and Tet fy Menus) De the free energy per unit volume, Let the potential energy 2, of a consi ‘uent substance be proportional to the mole number ‘The total free energy is fn 4 PB ntot ‘where (#) depends on position. With respect to va Of the mx), # i to have a minimum value; thus, ofl’.lows Since the substance can only be digplaced in the fore field, none of i i lost, ‘Therefore, Jomarno From the above two relations follows Mas ya) + Biel) ‘This isthe generalized barometer formula which wa orig- Inally derived by Gibbs for the gravitational field, J Osmeric presure ‘The Gibbs variational method can be quite generally ap- plied to solutions." Denote the solvent by 1, and the solute by 2 In ¥; let there be Xf moles of solvent, and in V let there be a solution consisting of 3, moles of solvent sand X, moles of solute, separated from Y, by a semi permeable membrane which is permeable only to the sot vent. The concentration in ¥ is then O=N YN, (aolute divided by solvent) ‘With semipermeable membranes the equilbriuin condition is tho 45 0)= pl 37) serie ria Goon" sd Wao 7. Wid ‘Since the Gibbs funetion ia « homogeneous fanetion of the frat degree, we can write OE. rK NI RIE Ho) amd oH 2,0) mo (een tl@) ot n-(B),- If» isthe oumotie pressure, then we have the equilibrium conaition Milbe+401= ploy 0), where p= Bex ‘This relation can be transformed into PsP 0) alte 0) = palBo+ 0) Ba +50). ecaute (20/2p}-—V snd (@O/2N,)ey day it follows that @ylePle= (@VIBNIng- For om that iy if contains ‘only tolvent, then v B¥ Therefore, alee 0)— les +259) “nina vt nism] ea It re eget he compro te wet sta "(4 niet 30-mvemci~(E)e=(2)a te deena oft sel tet esnee- Moe use tel ow eee 4B enh tne raf hr ie 1 the solute it volatil, then there is equilibrium between ‘the vapor ofthe solute and the solution. If pis the vaporn= (2) Because of the ideal gus laws, wo have pressure, then tre) MAP)=BTlogD-+{(E) (chemical potential ot gases) ar (@)=anvap ven ‘Using Henry's empiscal aw, sell bey AERP oge Rog Inegatin ge det he geo (e+) from this It follows that where Auy dyyny cam be functions of p and 7, We have i042, 0)—malbo0) v a ate(tZe%.) Pots.) fe 8 | arcane em van 8 Hore RTeN,|V— REN |¥ inthe pressure ofan deal gua whose volume is the tame at that of the solution and which has the same number of molecules as the solute, ‘A, Builibrivm between the solution and the wapor ofthe solvent Of course, there is also an equilibrium between the vapor of the solvent and the solution. Let j be the vapor pres: ture of the tolvent. We have 15086) = ol ‘Between the pure solvent and ie vapor there ie the relation P58, )— ol) ‘Therefore, oy fi) ap — rriog®. anion =|) ir Rok According to the definition of osmotic pressure, PAl0.6)= plo 0)- ‘Therefore, Hl, 2H, 0) = slp = 0), 0) = RE [Neglecting the comproasbility of the solution, we have mina 0--ninor-](2) erie, (f= rie sat me theta thin ann amor sure and the change in vapor pressure which comes about ‘because there ie solution instead of a pure solvent, From® socnsaen | cup. 8 ‘this formula, changes in boiling point and freezing point ‘an be caleulated at a function of osmotic prestre. 4. Vapor pressure above @ spherical eurface We consider a sphere of radias r which ie to contain ‘n moles of id per em? AL constant pressure we assume that mi constant, Then Asta, dA =Sxrdr, oe wens, a= nterdr, 3 ‘Lat y be the sartuce tenslon, Then aw = pav—yaa wits O 474, wo obtain A =— sath vap+uan +yaa, (inept =f 43], BD HAD) Ne. ws , we have He Boo [nny 2 mio] N= 0 and eet mie 7Z = m0 Here pi to denote the vapor pressure above the spherical surface, Tie formula ean also be derived from WA +R 0 with AVE VK) <0. ‘Taking the Limit r-+co, we obtain, for « plane surface (rapor pressure p), (P= Ht) ‘Tons, P= le) 01k) md or, in terms of the molar volume rdefiu-even ‘This equation is exact ‘Parthermore, fr an ideal gus and an incompresibe Suid, Tog r =nip'—P) ‘Thus, the vapor pressure p' depends on the radius r. ‘The smaller + is, the larger p” is at constant p (because the derivative with respeet to p’ of the righthand aide of the formula ia RT}p!— >0 20. COMMENTS ON THE SECOND VARIATION Let @ homogeneous substance in a volume ¥ be separ ted into two fxed parte of volumes Y; and Y; by an imaginary partition, We want to find the conditions for ‘table equiibriam. We have ¥, MEMeN, Band 48,08. Totwith the subsidiary conditions AB 4AB,=0 and aN,4aN,=0. ‘Therefore, the equilxium conditions are (ce) tom (202) Since Shas the homogeneity” property, SUAV, 2, AY) =AS\V, BLN), wo have sor (28) sa(3) (es)—l6-¥ ie)-#(@3) For homogeneous substances, 48 (QE 4 pd¥)/Z; that is, asjevapi? snd aso —a/r. ‘Therefore, Consequently, from Ka. (202, DINER ONT, oF ON = OI, that i, nam ‘The above relations are necessary for stable equilibrium, fn 28 | oman sar econ ante « Dut they are not sufficient, ‘Therefore, we sil must inves: tigate the second variation. We must have or=ponaos BC + (iS, Janan3 2%) ax <0, ith te ssn condtns (unre By, aBaN a a8,AN, and (N= NE Sine inthe equirum ste the ques any 8, a8, Mapax,, Nat are independent of j, wo ean introduce the following quan- tities: internal energy. por mole, Ns molar volume, = V/No= 8; and entropy per mole, s(,*) = S[N=SN,. Then we obtain asaar{l(@)aors (2 seor In what follows we shall again write 8, E, and V3 re: fading that, we must bear in mind that these quantities refer to one mole. With the new notation, the stability condition ie 32 6( 25) ans (8,)anars jaro Me wt ete i on ee fe i Gi, *-7-*(), GRer-(Se)ev ana —2 ars |) —2( 2% arUsing the relations a(33)=(S)0" 7 a(3)~(spaw)ae+(Sen)a"> vona(ansa($)ar = a er] F)47—Fe[()- *erav)|**) EEG sSe)ele- sor} From this we obtain the following equiibeiam conditions {for V= constant, 50; and, for P—constant, @p/2V),<0. se, | memeoeece mle ° 2, THERMOELECTRIC PROBLEMS Because of irreversible proceses, these effects are par- tially removed from the scope of thermodynamics, Aa tistical explanation ie possbles however, t goes beyond the framework of phenomenological thermodynamics. 4, Thomson's fleet, ‘tan electric current J flows through a wire, between the fends of which there isa temperature diference A, then fan amount of heat Q i produced per second, given by gaan ‘Depending on the material, + ean be positive or negative ‘Tals proces is reversible, b, Paton cfect "At the boundary between two metals through which a current is passing, the heat Q—AJ is produced per see: fond, The Peltier onstant 2 depends on the temper (Thermal emf In the solder joint between two metals, an emf, is produced, A 7 Y} A~ ToastsLet bro different wires be soldered to one another at thelr ends And B. Joint 4 isto beat temperature T-=47, snd joint Bat temperature 7, In A the emt By-+dl, is produced, and in # the emf Ey ls produced, In'A the Peltier heat ~-+-dx is absorbed per unit current, and in B, ‘is given up per unit current. ‘The amount of heat pro- ‘Guoed per unit current in the two wires because of the ‘Thomson effect is (z~~r))47. Sinee the “internal energy” ‘of the system does not change, therefore, according vo the first law, the amount of work done per unit eurrent mut ‘qual the amount of heat absorbed per unit eurent. Since ‘the amount of work done per unit current per second oquals the potential diflerence between 4 and 2, we have AB dr h(a aT, aE _ ax ar ar eon Tntograting from 7, to 7, we obtain AT) — aT) + [irre a7. It we view the entire process as reversible, which is not ‘exactly cortect, we can apply the seoond law and obtain ah) AT), f, oe Ae +f {this follows from wry) — 21) ar, wie 5001 Since 7, and 7, are nearly equal temperatures, we may Aliferentiate and obtain age ° a . (9)+3hE-]-* ~(z)+2li)-° ‘This yields preetely the relation au 7 tes dae to omson A, sn #)--2 (tr) he drat of hs ran at 4 mater of fmt sera nortan hse iar on taafes mine on the bass of statistical considerations.* Following Boltzmann we can, with thermodynamies, rigorously de- Olav A=heat conductivity and o: own 2h where lectrieal conductivity sR hin i ee fitChapter 4. Nemnst’s Heat Theorem 22 NERNST'S MEAT THEOHES ‘The Nernst heat theorem is concerned with the behavior ‘of the thermodynamic functian in the neighborhood of the ‘absolute zero. Itis well known that the second law deter nines the entropy only to within an additive constant. ‘This constant ie eenalby normalized by means of the Nernst heat sheoren ‘Let us first consider the free energy rower (3) (AF) ar @), ar-anse ‘From thi follow the equivalent relations ear) 109) Baer = Bor sand farther, igs —0 "This is the more retroted formulation of the Nerast the a8 | wane nar renee ” orem it-aayt that all entropy changes are zero at the solute zero, A generalization of the Nernst theorem due to Planck slates that in the relation lim AS=0, the & exo bbe dropped; ,= Jims should be finite and univers ‘A ressonable normalization of the entropy constant is eh that ims that is, he entropy of all pate substances has the value zero at the absolute zero, T'=0. From this requirement it follows that the entropy ofall substances must be zero at the absolute zero. ‘This is beoause the entropy change AS ‘which oceure when substance is made from pure sub: ‘ances at absolute zero i zer0 according to Nernst i follows that From this it follows that ar wn(ei)-0 oe Likewise,follows suf Faro s-[ gar mn its goneraized formulation the Nerust heat theorem states thatthe above two integrals exist. Tf lime, exists, then, for T=0, we must hare = ¢=0. Since S=—(@r~M, and s=— eoRn,, 4 follows that P. forfotParsrun and r, vf Maa erm ‘The requirement of the unatiainabilty of the absolute zeros contained in Nerast’s theorem; itis, however, weaker ‘han Nernst's theorem. Tn onder to prove the wnattin ability ti sufficient to show that it impossible to reach ‘the absolute zero by adiabatic changes of state, because every process can be decomposed into adiabetie and iao- thermal processes. According to Nernit’s theorem, there does not exist an adiabatic curve along which one could reach the absolute zero. Since at the absolute zero the tudiabatic curve 8=0 js the same at the fsotherm 7=0, therefore no adiabatic curve 8 constant 0 ean intersect ‘he adiabatle curve 8=0, which correponds to the ito therm T=0. Thus, itis not possible to reach the absolute zero by means of adiabatic changes of state, For ideal gaes the entropy per mole i iog(v 2-0 At 7a, 6 does nog thvugh seo for fie change ‘of state.” Tt would appear that Nernet's theorem is sat ised, However, bocause of Sm Glog + Rog + ig (50, 1)— sc 73~ Ri) 20, this isnot true. In spite ofthe fact that in this case the absolute zero is not attainable, Nemst's theorem is not satisfied. One mast not doubt the Nernst theorem because of this, Tather, one must accept it ua correct and at the ‘tame time, assume that the above equation for ideal gases in no longer valid at very low temperatures (in the neigh Dortiood of the absolute zero); that is, the ideal gases be come degenerate. ‘The quantam theory confirms this st sumption,Chapter 5. Kinetic Theory of Gases 1m connection with the atomic structure of matter, it ‘mast be asramed that the emallet components of matter, that is, tome and molecules, are not at rest but are in motion. Of course, this motion cannot be seen directly however, its existence is apparent from Brownian motion. ‘Becoure of this, first Kronig, and later Maxwell and Claw js, formed the following hypothesl: Heat energy is iden- tical with the kinetle energy of the molecales or atom, ‘Therefore, one can aay that heat is a disordered form of energy. Boltamann introduced the cancept of the *prob- bility of 4 state” and connected it with entropy. ‘The lar that the entropy can only increase is identical with the statement that a systom which changes its state can ‘only go to a more probable state. "The various states of matter are difrentiated as folloy Gases: Except for collisions, the molecular motion is foree fre (noglecting external force field) Solide; The atoms oscillate about an equlibrium con ‘guration. “Liquide: Because there are always a large number of ‘molecules which interact, neither uniform motion hor an equilibrium configuration is a good ap- proximation. ‘Today the theory of gases and the theory of the struc: tre of the solid state are very advanced, in contrast with the theory of liquids, whieh is not yet well founded, ‘With ideal gusee we may make the following assumptions 1. The average soparation of two free molecules is very large compared with the molecular dimensions ‘2 Tho average potential energy of the molecules is very mall compared with the kinetic energy and, therefore, we fan neglet it. ‘Wo shall aasume that a gas consists of hard spheres. T¢ ‘tums out, however, that the majority of the results are independent of the specie model 14. CALCULATION OF THE PRESSURE ‘We want to caleulate the pressure which the molecules of an ideal gus exert on a completely elastic wall. Let Fent,sldedndr, be that fraction of the molecules for ‘hie the velocity (6,29) ies in the interval 6, t0 40, ty to t1tdta, to tides. et f be £0 normalized that fens, desde, where a ‘The pressure is given by the momentum delivered 10 the wall per unit time snd per unit area. The motecles for Which the velocity lies im the interval r, to dn, 10 trtdny, 15 to ty-Fdry, and which hit 1om* of the wall in a second, are in a cylinder of volume n, (Gee Pig. 24.1) ‘Therefore, the numberof these molecules is nfs), ‘where w is the number of molecules per eubie centimeter ‘Momentum 2mn, is delivered to the wall by one molecle; ‘therefore, momentum 2mefnf(e, est) is delivered by all of the moleeles in the eslinder. ‘Thus, the pressure p‘mmo wip oF ota | Cap. das the value p= 2mnfitfiare ordre. b— If we assume that / is an even funetion in oy, that is, Ht 9) = flr ay ta, then = mnfetfra rs mare, Wo deine , the average aloe of 1, as A=[eleenmar ‘With this definition, we have p= mn ‘This result ie independent of the nature of the wall, If an arbitrary wall is given, then, by adding an imaginary nue face, we make it into a closed surfuce’ We ean define © ‘momentum teneor Tam nny, ‘Tho ith component of the momentum transported per unit time across the imaginary surface Fis then fa teonsuat ‘where NV is the outward normal to af An analogous expression, in which the pressure appears ‘sa tensor pay is valld for the wall W. Considering the Dalance of momentum within the volume bounded by P and , we obtain Tacosih, 841 +[¥ pucosiN, 3.44 F th a +d fos It we substitate A= ap ( constant) Into Gaus Joap-focasr,src mons ov cam | Cha. & then, wing divA—a-grady, we obtain Jorar=a fora gav , Zaefrom Nanay Zoof Lav, focosw.suar[ 2 ar Substituting 2, for ¢ and letting F represent a losed sur- face, we obtain from Eq. (24.1) fpitsarsd fea he Hence “ Se =-— 322+ nsnnie atin or the stationary case, from this equation as well as from iq, [24.1], we obtain [4-12] Pa=Ta independent of postion) In the cate of an isotropic velocity distribution, we have po and HFA, Pa therefore, poi ‘The kinetic energy ofa tingle molecule it By foe, and ‘the kinetic energy of all of the molecules ie. T= {a0 therefore, p=208. If lathe molar volume and Z it Avogadro's number, then Tm ne, and we obtain po 41k, According to Boyle's law, pe 0; fom thin i follows that where = /Z is Boltzmann's conttant, ‘The average ki- netic energy is thus a function of temperature, Here we fan make the following important new assumption: In ‘thermodynamic equilibrium the average kinetic energies of ‘bro molecules are equal to one another. Under the as: sumption that the Kinetic energy accounts for the entice heat energy, we obtain 2 oa aF Ste for the epee heat, The second assumption isnot ge evaly alld Tt io vt only for monatomic gases (aable tes, metal vapor). For polyatomic gases, e,> 3/2. Tis Means thatthe Lintic energy dovs not account for the total heat energy. Strely speaking, Wee statement are outside the realm of the Linlic theory of gach Since there are three degrees of freodom associated with the translational motion, and tee the kine exergy in this case (monatomic gases) t Ey, —SET}2 It aanumed that tae kinetic energy per degree of fredom i Ea. Tn the cave of polyatomie gases the rotational energy mist be considered long with the tratiaiona energy Monatomie gases: * diatomic gases: a1 5 triatomie gases n= 183. ‘The quantity » ie calealated from the folowing relations: pee 1 =t 6w ones rune or game | an ‘The diiutticg are associated with the determination of the Hhumber of degrees of from, (or details see the letures fn statiatieal mechanics) "Phe value of the average speed 0 V' of gus mol: ‘cule is of the order of magaitude of the peed of sound Vs VeiVar From ito save Yate 28 VELOCITY DISTRIBUTION ‘the exact derivation of the velocity distribution prooeeds trom « consideration ofthe collisions of the individual mol foules or atoms. In the case of an isotropic velocity dit: taibation, we have flew ravenidton lr)dte = feyerdead, where Pael-pehtefy and where 4 Is the diferential Told angle. (In polar coordinates d20=orded®.) Maxwell postulated that the aistribation of one component of the Yelocity is Independent of the other components, There fore, he at fo)-—=ateateae()- Introducing the functions ‘ye and gfe) such that logaten = vin) = 252,3) and logslet) = ate" ‘obtain the fanetional equation et) = ofot bet Fo = pled tyted + 9) A solution existe omly if g(et) and ye) are ui ext ‘v(0) are linear fane- vet) =—att+— oF gi.) = constant xem! ‘From this i folows that i ‘The constants determined from the n thew rom norman ffs wwe obtain fo)= (2) e~ ‘Now, can be determined from the mean square a, Sorseuabede square velocity #, Jevm_ iho. em hence oer Pally, we obtain ferwe= (aq) lB for the isotrople velocity distribution fu bution function. For he ‘probability that a molecule has speed e, we obtain seen fea ~eG!‘This yields the velovity distsibution shown in Fig. 20. ‘The most probable speed is obtained trom the relation aio) o= 0: 1 oe wie) 2a), we ob since the mean square speed is P= 3% (@msy. The average speed ® in defined by fennel i ee ery, There [Aer two partial integrations we obta fore, once V/ER a= Sau Pea “We now want to calealate the molecular eurent J which potas acros a unit eurface area in the direction. With an frbitrary isotropic velolty istibution, the numberof mol- ‘eeales with speed © which erote the unit surface area per second i ae, {(e) de = ne, floytava, where 40 2minbad, Thereore, a [ee ees "The Maxwell distribution can be experimentally veried ‘with the experiment of Stern. Stern produced « molecular ‘beam in a highly evacuated tube. ‘This beam hite «screen Asie tt pont P. If the tube is now allowed to rotste about fn axis perpendicular to the molecular beam, then the ‘beam in defected due to the Coriolis force and strikes the sereen at point P” instead of at point P. ‘The velocity Aisteibution of the molecular beam ean be determined from the defection PP. 26 MEAN FREE PATH AND COLLISIONS ‘Weimagine that thogas consists of hard spheres. Then two rmolecales will collide when the center of one comes within ‘the interotion sphere of the other. ‘The Interaction sphere of a molecnle is the sphere whose radius is o=2, twice ‘that ofthe molecule. Ifwe take foree center instead of hard sphores, a quantity o can be defined in a ike manner; thequantity is velocity dependent in that cate, Tet w be the ative velocity of two colliding molecules with velocities vPand v'y and Tet ¥ be the eenter of mass velocity of the system consisting ofthese two molecules. ‘Then, under the ‘efhamption that the two molecules have equal Masses sec 23) and Tem plato ‘phe number of collisions Z per unit time (one second) 8 ecanse all molecules which are in the eylinder drawn in Fig, 26.2 collide witha given molecule; # is the averase ‘yale of the relative epgod w. The probability that to Tnotecules with velocities » and w', respectively, hit one another is sey») area perae ener (2) emrenaay If we introduce the veoct tm ‘the velocities we and F instead of w and e', Vit bie, ecanse Meryare We then obtain Sine frewe iene wwe obtain Vie by making the via tanfomation v= V3 : mation = VE he a. erage speed is 8, the number of collisions is ws naviateTn case the masses of the molecules are diferent, we in- troduce the reduoed mass Bene eee (febtert ana cal t i avaw, 2° ate = constant x¢"® ane, ie — poets free path we obtain distance '= Jambar of eolisions per distance ‘The quantity not equale the sum of the collision eroas Sections of all molecules per eubie centimeter Satherland diseused for the iret time the case in which the moleeules are force centers which attract at large sep rations and repel at small separations. Te found A(+4), ‘ee €50 tr nitinol, Se S. CuarwaN and T.. CowuiNe, The Mathematie Theory of Nonuniform Gaser (Cambridge, 1989),) 2B. TRANSPORT PHENOMENA (>) Included among transport phenomena are internal friction, heat conduction, and diffusion. They are used in the deter ‘mination of the mean free path 1. With all of these phe ‘homens the gradient of some quantity generates current, Tm the cate of internal fretion, the gradient of the macro copie veloity generates a momentum current. Tn the ease of heat conduction, a heat curren in produced by the tem: erature gradient. ‘The concentration gradient generates Particle current in the cate of difusion. All of these phor ‘omen are connected with an integral equation. (See the lectures on statistical meckanies) ‘4 Internal friction (simple special ease) [Between two parallel plates there isa gas, and one of thea omc moe op cama | Chap. 5 plates is moved parallel to the other with speed 2 Let 2) ‘be the macrotcopi velocity of the gas at a distance = trom the lower plate, ‘The phenomenological theory predicts that (varies linearly with =, Thus, le) = A that is, ‘This theory further states that a shearing stress which acts against the motion of the upper plate, appears along that plate’ In the general cate, there is a stress tensor? 334 k=1,2,3) 1 ia easily seen that the trace of this tensor is zero [Although the rigorous theory ix rery complicated, there fn a very simple elementary theory. We shall concern ou selves here only with the simplest special case. [or the rigorous theory of transport prooeses, see S. CHAPMAN and 17.6, CowLINO, Mathematical Theory of Non-aniform Ganee (Cambridge, 1980),] We decompose the motion of the mol- feoules into ‘an ordered (macroscopic) motion and a disor re xpumatl eak fe samo ae 18 the extn SA7Eian 2a ia To a ‘ered motion, Let v be the velocity of « molecale, the maeroseopie veloity of the gus, and w the velocity of the Aisordered motion of & molecule, We have vate or , and {In our special coordinate aystem we have Let P be a surface cloment of unit area inthe plane «=, We want to ealeulate the amount of momentum trans: ‘LLZZEZZEZEZEZ EEL une (LLLZZEZZZZZZZZZ ported across F per second. The molecular eurrent J trans: ported acroes F (Section 25) First, we consider thote molecules which cross Fin the ++ diveetion, or our ealesation it i suficlent £0 assume ‘that these molecules all come from a layer which i at a dis- tance al from F, and that they have acquired a eorrespond- ing macroscopic velocity. (For us is an undetermined factor whose determination roquites an exact caleuation of the distorted distribution of velocities in the moleealar stream.) The macroscopic velocity in the layer 36 (mma al with (dae‘Thus, the molecular current caries an amount of momentum, Il ccaa= From an analogous consideration we obtain the molecular ccarent in the —= direction, Therefore, te Ba fs the total momentum carsied across P in the +++ die tion, However, this equals the shearing stre, "gaat 802» Be Pa 9 mit SS, which is exerted on 7; from this it follows that mua. ‘Thus, the kinetic theory of gases ean explain the phenom- enon of internal frietion. Since a rigorous theory exists, the above result has only qualitative value, to be sure. “However, it does turn out that 1 can be determined from y, Because the free path is inversely proportional to n, 1 fe independent of m. TE wo introduce the quantities Wa 2ilra}, T= A/(nvEno!), and a= mj@eN), ve obtain (A= molecular weight) Vin? ,_ VEE, Tor aT se 21 |msmone renown (>) m Since o depends on the temperature, y~ VF cannot be ‘taken Literally. For the cage of hard spheres, «0.995; that is, a is nearly 2. However, these formulas are not restricted to the cate of hard spheres >, Heat conduction There is a gas between bro plates which have tempera- tures 1, and Zy, ‘The resulting heat current ae is propor ‘onal to the temperature gradient ar_n—n, aoe According to the phenomenologial theor ar a ‘where x isthe heat conductivity. Otherwise the treatment in the same as before, We obtain, sine e=0, ©=w, (B= energy per molecule). If we introduce the quantity 6, ‘8 the specific heat per unit mass, then Bae TE SE mop E‘rom this it fllows that m= bnmbatey. ‘Comparing the expressions for x and 7, we obtain cone Tn the case of hard spheres, a'/a= 2.5. &. Diftasion ‘The particle eurrent per unit area ts given by mn sav, where D is the difusion constant, In a way analogous to ‘the previous considerations, we obtain D= jou ‘For gas mixtures, D. where n= mm. "All ofthese transport phenomena are independent of the density e. However, the phenomendlogieal theory is invalid for very low densities, In our previous considerations we tacitly asromed d'>I. However, in the neighborhood of the plate, there is always a layer, whose thickness is of the fonder of 1, im which ele) 2d is no longer vali, Like- wise, the caleulation of the average velocity is incorrect for that layer. ‘The distribution of velocities js then as shown in Pig. 274. A better approximation of the velocity gradient is then where al and a is & numerical eoeficient. ‘Thus 7 ‘proportional to the free path. Th the case of internal ive ton the shearing stestes are then Puma Por air at one atmosphere and 18°C, al~10~ em, Anal ‘ogously, in the ease of heat conduction we must write ar_n-1, aoa" If 1-4, wo arrive at very simple results, Then there are practically no collisions of the particles with one another to take into account; there are only the callisions withthe walt. Such a gus behaves like radiation, Heat exchange comes about only through collisions with the wa ‘is no longer a temperature gradient, Then, "The heat current is proportional tothe density, but is inde-pendent of the separation of the plates, Knudsen esrried fut such experiments, which require « very high vacuum and very fine capillary tubes 28. THANSPORT PHENOMENA (1-4) ‘8 Diferion through holee and pores et gasfiled volume be divided into two parts by a wall, In the wall let there be an opening whose linear ah dimension is of the order of d, When dl and p= Psy there is no mass current through the opening. However, Md <1, then there is no mate current under the cond! tion that (8h eh Because #~VF and n~p this condition ean be rewritten as eVR=evF: or, Because p~o7, it ean also be expressed as Pe (condition for no mass current) 2), Heat conduction at lowe pressures Let us again consider the ease of two plates wt diferent temperatures T and 7’, now for the case d-) a e. Flow throwh a tube at low preeures Again let 12a, in order that we can regard the particle 20 ‘motion as radiation. ‘The surface element do radiates Apoosbdedo toward @ [A13], where 2 sin@dddy, ‘Thus, de radiates upwar: sefifiadontation Ante ‘This is A,r per unit area of do, which must equal the ‘molecular current per unit area In the » direction, "That is, Agra” Becamea omc mony oF wate | Chap. 6 ‘Here, as in photometry [4-13), Agcos ded = Agcon0! do’ 4a” applies to the mutual irradiation of two surface elements, However, agit Aa de WO and, because Q0,= FUeosd = reato, ‘Therefore, the radiation from @ toward P is [4-18] ay Ascos0 4080 = Apcos0da dar? aoa tar ine = [tend ansaoay—row oan ag jae Integration over © yields (4-18) aera, ee far: integration over the eross tection yieide for the current. If the eros section is epherically apm: rmetzic, we obtain afore, ‘and therefore du, set es a8 | enerone aman > 4) m For Jay the mass current, we obtain 4 Baty = Vie Comparing tefl with Palas oma _leaete 7 ve hain Fe contant x! 29. VIRAL CONCERT Let force K, act on a particle of mass m,. or an srbic trary number of particles the virial ie defined as wo Dark, 1 EK, ‘the origin of the coordinate system. Since K,= mi and , then It does not matter which point is chosen het sen gm aks yma Because the time average as well asthe statistical average of an exact diferential vanishes, we have 7 Gert =o, ‘6 Long as x and & are always finite, ‘Therefor wen Smat 4X mel=0 (virial theorem)2, APPLICATIONS a. Heal gases In the case of an ideal gas in a volume V, « and & are always finite. Per mole, Smet 2h. =3KT ‘This expression is also correct if forces are assumed to act Detween the molecules. ‘he virial of the pressure forces ‘Thierelation Is valid for volumes of all shapes. Hora cube the forve on one side is Km~pot. However, since for three of the sides «:K—aK, and for the other three aides =-K=0, therefore w= Spt Spe For a rolume of arbitrary shape the virial can be cal- 28 | amsnone a culated by means of Gauss law: spfa won rfenat=—pftaivna It no other forces need be considered, then per mole, 3p —3p04SRT=0 or p= RT; ‘that is, we obtain exactly the ideal gas equation, 1D. Real gases (Corection to the ideal gas epuation) Let there be forees botween the molecules. Let the po tential of the central forees be U(r), and let i be x0 nor malized that’ U(co)=0. ‘The forces K, between two mol coules are where rele] and zen, for > 0, where is again the interaction sphere, For r< wwe can set U(y|=co, ‘he virial for the pair of molecales then" In order to obtain the viral fo the interaction of one mol cule with all others, we must integrate fom o to co, The volume element is 4x70r and there are Liv molecules per cubic centimeter. (J; is Avogadro's number, and e is the ‘molar volume.) ‘The result i In obtaining this result it was assumed that the particles Interact only pairwise; that is, there are to be no many-ody forces, Since there sre Z molecules per mole, we obtain 2 fav E [2 aarar fa the viral for a mole. The factor } appears because we Ihave counted each pair of molecules twice. By partial integration we obtain aa afomners a. Introducing the abbreviations 2 yy) aad an —tanrfeyear, swe then obtain (0, Virial forthe coivion freer ‘The change in momentum of the two moleeues contrib ‘utes Zome, tothe virial per colision, where wis the relative ‘elouty of the colliding molecules, Since, for given Yeo tor x, 30%wy-n molecules collide per tecond with a given so a0 | amcor us ‘molecule, the virial is 2omoyscatty per molecule, The time average of this is Since mis 247, the virial per mule is Egat =a? (yi of leas Substituting {into this relation, we have wate Lar = nr, where b=20b/3 equals half of the total volume of all Interaction spheres in a molar volume, 4. Generalized barometer formula Lat the potential Uir) be represented as toate ‘Then, ants [aur “yar ar Performing a partial integration, we obtain BALK fog fe 1]a eros more ow ext | Chews Defining . aay = ante | fe Jar, es or the equation of state we thus obtain Air) ART ap0—3 [ret 1) has the form shown in Pig. 80.3, with °, pr, pa BP AD, (o=moleoular diameter) the doundary conditions Uinjmco and TUy=0. ‘Thee tonaltions imply = “ “yn (Py! werse, ana (Pajr-1 twree fe 8 | arcane = We have defined ‘whieh isan approximation to the van der Waals equation, rege BB (ete(lf +) ‘mpiricaly one nds «>0; that ie, for r2>a, we have U(r}<0, whieh implies attraction between the molecules, ‘The redius of the interaction sphere ¢ can be estimsted from these formulas and from the formulas forthe free path,Bibliography ‘Thermodynamiee ©. Sucxtm, ebrbch der Themochnie und Phrmadynamil (Be. in, 1025, ©. W.zwit sad M.Ravoass, Thermodynamic (revited by K. 6 Piaer and L Brower) (McGraw il, New York, 16), J.-D. vas bu Wists Lebbuch der Themadnamih inter Arwernng das Geicgeail! von Sytanen mi anirmiy [arin Phanew (Lecture aoten edited by By Rebun ‘Amsterdam, 1008-1013). W.H. Boworner, hermodgnamié (Beri, 19) 4.W, Gunns, The Caleted Works of J. Willard Gis, Vol. 1: ‘Thermodynamics (Yale Univeraty Prom, New Have, 1028) M. Puasex, Vorlenngen iter Thermadgnams (Verlag von Vell © Comp. Leip, 108), P. B. Brora, Terthot of Phermadynamic( Tac, New York, 187) RICE, Chavatos The Mechanical Theory of Heat Oscsan and ‘o,, London, 1870. actura on theoretical physlce given by HA. Lows, Hux inet Theory of Ganon ©. E, Maren, Die Kndsche Theorie der Gaue Maruiche vod Borenat, rans, 1677; mond eon, 160), 1, Borrenasx, Verirungen aber Gastheore (Verlag vox Jobs Ambrosian Barth, Leip, 1695. A Kapiic, Grondsge ciner Thoie der Gave (Beri, 1856. BS. B, Guava, raion works F.C. Maxon, Try of Het (Longmans, Gren, ao Co, Lon on, 1902), Wiley & Sons, J. Sans, dw Tntradtion tothe Kine Theory of Gee (Cai ‘erty Pres Cambri, 100, ETL Riwwann, Kivatie Thory of Guser (MeGraw:Hil, New Yoru, 103), HW, A'Lonpere, Letres on Thorlo Ply, Vol 1: Kintial Probleme (cman & Co, Lid, London, 182) 8. Guaraas aod T. G. Cowrix, Hathematice! Thory ef Now swijrm Gate (Univer Drom, Cambridge, 1853) ‘See to works by Motuze-Poviss aed Haaren,Appendix. Comments by the Editor [At] (p. 2). What is meant here is thermodyzamle equi- rium in contact with another eystem at fized temperature (reservoir. [42] (pp. % 5, 25, 28, 94 Actually, thee are w rel tions (equations of state) which may also be written as YeeWsirenae) Herein addition t=5,— constant, An ‘example ith n—2 ie given on p. 2: Solving the equation ‘of slate of each sueytem for ¢ yields F(9, 7, , P) VI, P=. [45] (PP. 8, 21). Observation of the variation of tem perature «by variable contact with two reservoir a di ferent temperature tnd ee the monsony ofthe sale between f, and, bat the sign oft (snot determined Indeed, this sgn defines the direction ofthe heat Now ae: cording tothe send law [4-4] (P11). Here Bg. 72] in meant toe taken at [45] (pp. 12, 88, $1). Tn the following « quantity of substance of 1 mole oF, asin Section 3, of 1g should be considered. [46) (pp. 21, 24, 83), ‘Tho thermodynamic temperature scale is determined only up to a factor of arbitrary sign ‘The choice of the positive sign is a convention. Within this convention negative temperatures are uncommon al- though not inconceivable, a demonstrated by the work of Pound and Parcel, Phys. Rev. #1, 279 (1951), on mag- netic population inversion, and that of Ramsey quoted on P. 2h [4-1] (p. 22). Actually, @=0 by construction, since the ‘coupling of the eyelo to the reservoir at 7, was only an frtfial (and even unnecessary) device, (43) (pp. 24, 81). Om p. 25 it was proved that the en- tropy of a closed system cannot decrease. But here alla sion is made to the arbitrariness of the sign of thermo. Aynamie temperature (se (4-6), [4.9] (p. 80). This should read quasi-static adiabatic, Ror ‘only in this caso are the y, in 3W welldefined functions of Siysoy Zee Note that the existence of atleast one quai static path between any two points 1, 2 in state apace it fasumed here a8 well as on p23, [4.20] (p. 45). Prom the fre energy the stable isotherms an be determined without making use of these unstable states (te, e, K. HUANG, Statistical Mechanior (Sohn Wiley & Sons, The, New York, 1963), Pig. 2.11], [At] (pp. 61-72). This part trom here to the end of ‘Section 17 isan insertion into the second German edition of a manuscript wrltten by Pauli in 1058 and which served ss bats for Paul's paper in honor of J. Ackert, Z.angew Math, Phys. 98, 400 (1958). This paper is noteworthy for being Pouls last published work [tee “Bibliography ‘Wolfgang Pauli" by . P. ENE in Theoretical Phytios im ‘he Twcwtich Century, a Memorial Volume to Wolfgang Pauli, edited by -M. Viers and V. BP. Weitskopf (Inter teionce Publishers, Ine, New York, 3960), and in Cot- lected Scien Papers by Wolfgang Paul, edited by R. Kronig and V. F. Weittkopf (John Wiley & Sons, Inc ‘New York, 1964)]. The footnotes on pp. 64, 78 have boon inserted at the same time,[A-12] (p. 98). Aetaally, this revut follows from Bq. (24.1) for a surface P chosen parallel and infinitesimally close to the wall Wand for stationary conditions [AA8) (pp. 119, 120), Here 4 and dO! are the elements of wold angle through which the surface elements do" and fig ate seen from P and Q, respectively, ie., (PQ)"a2 eos ade, (POA2'—cosde. [This is the same situa ton at in Section 4 (photometry) of the volume Optics and the Theory of Blectrons of this series.) ‘The radiation from @ toward P is by definition Aycos0’ dea. Inte. ration of this quantity on p. 120 is over the full solid Single Q, and not jurt over the upward half as on p. 19. ‘Therefore the term 4.0050 sin a0 dpe integrates to zero. Index ‘omic acarofmatr 98 hema, “et cement constant 4 ‘homie, 5 68 fore ve hm prce ‘Chal revo Ceqlisiam Socal 5 ‘minions 38, 7 Boing pins chngeeol 2 Choma sma 36 Beksmann (Cowon RB 2,31, 96 {hanson Clapeyron equation #2 {lore tem 1,31 Se al Bale voy 8 ation rove eto 102me Coton, 108 18,116 ‘embe 106 165 Cote pein 912 13,19, 10,1 Cyc proces, 7,2, 4, 55,61, 62, Dato’ In 45 Degree a reo, 93 Deny ty Sa 8 18 Pas diet al Yom 3.27 Dice fom, 5 2627 Bhrenet Aanssjows, 728 Ell conc 8 lee erent. Ene rate Kate, 9 ema 6 4616 e995 38 tet 785 Foto 16 ‘clan 9 ree 228 25,36 lod stem, 24 SD St pre nbs 98 Enropy conan 3391 eatin of ate 36,126 98 saat of ea 18 38 eto tno phe 1,1 lin dns 9.8, fer ain, 19 fee sae garam, 85 tm ant Holl ox, St ale retin 475 Evaporation | arena (of fonction), stv Flt, neal fre 1, 61,18, "et a of themodyaames 5. 4 it scan, 88, Be ll fcompreble, air condom fr 18 (eet en 3. Ina fore ld 7 eng pit changes of: 2 son tesa Is Fatal equton, I “tbo Thermodynamic fancton ‘ona i Wir 67 iio fenton, 38 310 228.31, Heat eevee 38 5-26 54 2, est ramen 1 Heat anpert, aerge of 16 ors mil Iw, 8° S30 Homogeneon ssn 1225 Hlomorencoas sytem. 2 Hsdentpmomi eetion 88 atonal en 210, 128 thermal changer 638, Tess ether 6 28,38, 645, 92,13 Letrepi lacy dara 8 trop very dito Iecshion 33 Kaden, ML 06 Legendre transformation fo the6 inde alt ulation proces gai Meena Moenttin 5 i Mao reduced, 06 Moo erent 10 Matera ol 96 Manel J.C. 4 100 ‘Mocellin 102 Mee fre pth Mechuneleqiaet of Mote wela 150 Motel crest, 122, 10,10, Moteaar dimensions, 9 Moleclr wel 3 Moment 35 9,109,110 9 Ser aa reste prove, Reply oeursngproemse oral frm of diere ‘om, Neuter ee Ordered eto, 18 Pant presare 5,54 65 6, 2 Pon * "ate, Indy tear. 28 Par bet nl mt ie fr king cond Kd Path Planck's generalization of Nernst thermodynomle, 6 Poti energy 1, 7.98 Pownd and Prcall experinen, 3 Pree arin $5 5862, 65,61, 72.76 Presure tensor, 7 Probab darbain of eer, Tor is Probability of at, 94 Pron ‘in, 29 ‘hele 3, 624.55, 61,62, 91 inert, 18387 athe 1 ‘quate soe 112124 ply eer. 23,24, 34 ever 45,87 unt thers, 9 ‘Qube sna ate Coreen) TIT 142224 2, ua ei proces 21,34 ‘Geeta ph Radio, 15,19, 120,12 Ramoey, oF 35 apy cig poe 823, ‘hema 3, 5661, 6, 65,68, 26 ato 52 Reston equation 56 Restart soap ln chm Testo Reverse proce 7,3 650, Reto emery. 99 Second lw of tarmedsmamie Tia 36.0 ‘Teomoe's fermen Stand ition 15 8-96 Semipermeshle memane (wally O48 ia Seed 5, 10 Sle egiieom, 8 raat, 45,6, Steal average. 21 Sita eth 48,9 Sts experiment 13 Sober theorem 9 Stren shearing 10, 10,122 Ste ema, 08 morenan, 12 35,91, 35-96 Seperated eed Sores sen § ‘Satherland, 107Sytem, led, 1,831. Sea modi 2,33 3,10 Tenor mem, 9 fee ‘etna, 3622 ‘Theomoeleie pons 8 Thomom We tlerd Rebla, 1, ‘Thomo fle, 87,88 Thomon raion, dealin 17 Tron “Transformation, nob rot ere. 9 Untainbiity of abel seo 92 456,181 Von der Wal equation, 127 Vie Hl bo 3 item condom 56,52 ieee 18 ‘sae 8 lent 8 or peril vray 8-83 anor preonre change Bewae Verne! method, Gi sverage (peed) 10,186, 112 Wierd ation, 19 reco, I 18 re 10 1 Velsey dtiaton, “ied 10 ra 15 eet coming 2 28,20 romipcrmenbla 40,16 Work fea done thy pen 5 0.51 ete Hof he Si, 0