INVESTIGATIONS ON THE THEORY OF THE BROWNIAN MOVEMENT ALBERT EINSTEIN, Px.D. woe 3 sone DOVER PUBLICATIONS, INC.“Tis Dove fon, pune in 1955, ean seabed aul ‘united repletion fhe ‘undaln ngnally abled THT pa {a troop special srangement wets nd Gas Lids ete ete of Abert Hain Standard Booh Number 486403019 Libary of Congres Catal Card umber 755 Manufictarl in the Une Stes f Ameren ‘Dor Paestion te 2 eat Bd Set Mio, NY, 1101 | | PREFACE A eee EINSTEIN was bom on March, 1879, in Ulm. When he ‘was only six weeks old his parents moved to Munich, where he spent his infancy, and went to schoo! until his fourteenth year. When fifteen he came to Switzerland, attended for another year the Gymnasium in Aarau, and took there his schoo! leaving examination. ‘Then he studied Mathematics and Physics in the Zurich Polytechnic, where Minkowski was fone of his teachers. In 1902 he came to Berne as Engineer in the Patent Office, and in addi- tion to his duties there, prepared himself for the examination for his Doctor's degree, which he took in the year 1905. At this time there appeared in rapid succession his first great ‘works on the foundations of molecular physics, of which those relasing tothe Brownian motions are collected in this little volume; as well as the well-known papers on the special PrincipleTHEORY OF BROWNIAN MOVEMENT of Relativity. In 1909 he accepted a call to a Professorship in the University of Zurich, 1911 a call toa full Professorship in the University of Prague; in 1912 he accepted a Chair in the Zurich Polytechnic. In 1914 hhe was invited to Berlin as successor to ‘Van't Hoff in the Royal Prussian Academy of Science, where in addition he undertook the duties of Director of the Kaiser Wilhelm In- stitute of Physics, To this period, up to the year 1915, belong his researches on the general ‘Theory of Relativity, as well as a number of fundamental studies on the Quantum Theory. CONTENTS Prance On me Movewenr oF Satan Pants Stem oa Spat Lat ‘Twzony or Hear {18 On the Ossie Freware o be Ascbed 19 ‘he Suspended Pericles 42. The Omote Presa om ty Stang ofthe Motecalrivte They of Heat A. Tcry ofthe Dion of Seal Spheres $4. On the Ineyuler Morement of Pate Suspended a's Light and te Rein of tis to Dion 45. Fomala fr the Meas Diplicerent of ‘Sepended Pavlos A Now Method of Determining the Rest Sse of the ‘toa 7 (Ow mae Tuuony oF ru BxowNIAN Moves 4 On a Cae of Thermodynamic Egulie 2. Kaamoper of Applton of tbe Haan ‘ined in $5. On the Caogenin he Pantin « og ‘bout by the Thermal Motion 9‘THEORY OF BROWNIAN MOVEMENT 44 Appian ofthe Equation derived oe" 415. On he Limits of Applicaton ofthe Formalin foeVa “ 1, A Naw DeroRumuarion oF MoLecetan Di twats 6 41. On the Eft the Motion of Liu ‘of very sal Sphere Sorpended in it 37 412 Caleslaon ofthe Vicoiy-neicent of Tig in whic a age uber of sll Spheres ave Supende in Iregaae Dis ‘baton » 5 On'the Volume ofa Disaved Sabaasce of ‘Molcslar Volame lage i. Compson trih that of the. Solvent se $4 On the Dillion ofan Undated Seb vance in Seton in Liga 6 5 Determination of Malacals Dimetsions With the help of the Ralaonsalexly Shed 0 meson 6 . unnearan THeony ore Browns MO fifi se fa Pt 2 Dilan Ineptar Motion of the Mae oe * f 5. Mower of Single Moller Brows INVESTIGATIONS ON THE THEORY OF THE BROWNIAN MOVEMENT I ON THE MOVEMENT OF SMALL PARTIC! SUSPENDED IN A STATIONARY LIQUID DEMANDED BY THE MOLECULAR: KINETIC THEORY OF HEAT NN this paper it will be shown that according to the molecular-kinetic theory of heat, bodies of microscopically-vsible size suspended in a liquid wall perform movements of such magnitude that they can be easly observed in a microscope, fon account of the molecular motions of heat. tis possible that the movements to be discussed liere are identical with the so-alled * Brownian molecular motion”; however, the information available to me regarding the Tater is £0 lacking in precision, that T ean form no judgment in the matter (f), If the movement discussed here can actually be observed (together with the laws relating to2 THEORY OF BROWNIAN MOVEMENT it that one would expect to find), then classical thermodynamics can no longer be looked upon 1s applicable with precision to bodies even of Gimensions distinguishable in a microscope: an exact determination of actual atomic dimensions fs then possible, On the other hand, had the prediction of this movement proved to be in- correct, a weighty argument would be provided ‘against the malecular-kinatic conception of heat. $1, Ow sur Oswortc Pressune To BB AscRIBED ‘ro THe SusPENDED Parrici3s Lot + gram-moleeules of a non-lectrolyte be dissolved in @ volume V® forming part of a quantity of liquid of total volume V. If the volume V* is separated from the pure solvent by a partition permeable for the solvent but {impermeable for the solute, a so-called "osmotic pressure,” , is exerted on this partition, which satisfies the equation PVeeRTs 2 2. @ when V*/2 is sufficiently great, On the other hand, if small suspended particles fare present in the fractional volume ¥* in place ofthe dissolved aubstance, which particles are also ‘unable to passthrough the partition permeable to the solvent: according to the classical theory of MOVEMENT OF SMALL PARTICLES 3 ‘hermodynamics—atleastwhen the force of gravity (hich does not interest us here) is jgnored—we ‘would not expect to find any force acting on the partition ; for according to ordinary conceptions the “free energy " of the system appears to be Independent of the position of the partition and of the suspended particles, but dependent only ‘on the total mass and qualities of the suspended ‘material, the liquid and the paetition, and on the pressure and temperature, Actually, forthe eal- cation of the free energy the energy and entropy of the boundary-surface (surface-tension forces) should also be considered ; these can be excluded if the size and condition of the surfaces of contact do not alter with the changes in position of the partition and of the suspended particles under consideration. But a different conception is reached from. the standpoint of the molecular-kinetic theory of heat, According to this theory a dissolved mole- cule is differentiated from a suspended body solely by its dimensions, and it is not apparent why a number of suspended particles should not [produce the same osmotic pressure as the same umber of molocoles. We rust assume that the ‘suspended pacticles perform an irregular move ment—even if a very slow one—in the liquid, on4. THEORY OF BROWNIAN MOVEMENT account ofthe moletlar movement of the Hg if they ate prevented from leaving the volume V* by the partition, they wil exert a pressure on the partition jst ike molecules in solution. Then, Jf there are » suspended particles presen in the volume *, and therefore n/V* = » in a unit of Yolume, and if neighbouring particles are sufi- cently far separated, there will ea corresponding ‘osmotic pressure f of magnitude given by Rn RT. ?=ANO WN where signifies the actual number of molecules Contained in a grammolecule. 1 will be shown in the next paragraph that the molecular-kinetic theory of eat actully leads to this wider con- ception of osmotic presure. §2, Osworic PRESSURE FROM THE STANDPOINT oF tie Mouecutar-Kineric Turory oF ear (*) If py Bars + Pe are the variables of state of (+) In tis paragraph the papers ofthe author on the “Founlaions of Therinodynamios "azo exumed to be fare to the reader (nn. Phye, 8, P4¥7, 1902 4 p-170, 1003)_-An uadarstanding of the conlasions feached ta the prevent poper is ot dependent on & Iknowieage of the former prpers ov ofthis paragraph ct ‘he preset pape. MOVEMENT OF SMALL PARTICLES 5 1 physical system which completely define the instantaneous condition of the system (for ex- ample, the co-ordinates and velocity components ‘of all atoms of the system), and if the complete system ofthe equations of change of these variables of state is given in the form » Po bib. pd = Ea, whence oe 7 ‘then the entropy of the system is given by the expression Sips. db (3) Font where Tis the absolute temperature, Z the energy of the system, F the energy as a function of f,, ‘The integral is extended over all possible values of #, consistent with the conditions of the prob- Jem, 2 is connected with the constant IV referred to before by the relation aN = R. We obtain hhenoe forthe free energy F, — Rr fe tap, ... dp = — Shige.6 THEORY OF BROWNIAN MOVEMENT Now lots consider a quantity of lguid enclosed in volume V; let there be w solute molcales (or suspended partes respectively) in the por- tion 7 ofthis volume V, which are retained in te volume V* by a semi-permeable partition; the integration limits of the integral B obtained in the expressions for $ and F will be affected accordingly. The combined volume of the solute molecules (or suspended particles) is taken as ssnall compared with V*. ‘This system will be completly defined according tothe theory under Alisossion by the variables a condition fx. «« Px 1 the molecular picture were extended to deal with every single unit, the calelation of the integral B would offer such difficulties that an exact caleuation of F could be scarcely contem- plated. Accordingly, we need here only to know hhow F depends on the magnitude of the volame ‘Vin whic al the solute molecules, or suspended bodies (hereinafter termed. brefiy“*paticls”) ate contained. ‘We ill cll 9, 4 the rectangular co-ordinates cf the centro of gravity of the first particle, a Jo Ba Chose ofthe second, et, yy JH toe of the ast pastcle, and allocate for the exntres of gravity of the particles the indefinitely small domains of parallelopiped form dy dy, d,; by MOVEMENT OF SMALL PARTICLES 7 Ay deg. ay Ay, eg, lying wholly within V*, ‘The value of the integral appearing in the ‘expression for F will be sought, with the limita- tion that the contres of gravity of the particles lie within a domain defined in this manner. The AB = try dyy-- deg. J, where J is independent of de, dy, ete, as well as of ¥*, ie, of the postion of the semi-permeable patition, But J is also independent of any special choice of the position of the domains of the centres of gravity and of the magnitude of V*, as will be shown immediately, For if a second system were given, of indefinitely small ‘domains of the centres of gravity of the particles, and the latter designated diy'dy'de'; diy'dy dey ej dy.de, which domains differ from those originally given in their position but not in their ‘magnitude, and are similarly all contained in V%, aan analogous expression holds AB = aaay, diy J Whence dye dig = dy “Therefore ay a8 THEORY OF BROWNIAN MOVEMENT ‘But from the molecular theory of Heat given inthe paper quoted i iseasly deduced that 48/2 () {or aB"/8 respectively) sequal to the probity that at any arbitrary moment of time the centres of grevity of the parts ate included nthe Glomains (ey ds) or (dy. diy) reac: tively, Now, if the movements of single parties tre independent of one another to a silent degree of epproximation, ifthe liga is homo ggncous and exerts no free on the partes, then for equal size of domains the probability ofeach of the two sytem wl be equal so thatthe follow ingholds a3 _ ae Bo But from this andthe last equation obtained it follows that J ‘We have thus proved that J 8 independent both of V* and of 3 Ys tye By integration we obtain Bow [Hse dg JV ad thence Rr, 5 Fane em V9 (YA. Bintan, Amd Phys, 2p. 170,903 MOVEMENT OF SMALL PARTICLES 9 Tehas been shown by this analysis that the exist- ence of an osmotic pressure can be deduced from ‘the molecolar-kinetic theory of Heat ; and that as far as osmotic pressure is concerned, solute ‘molecules and suspended particles are, according to this theory, identical in their behaviour at «great dilution, §3. THnony oF 1m Dirrusiox o Swat. ‘Spmenes ny Susrestos Suppose there be suspended particles iregularly ispersed in a liquid. We will consider their state of dynamic equilibrium, on the assumption ‘hat a force K acts on the single particles, which force depends on the positon, but not onthe time. Tt will be assumed for the sake of simplicity that ‘the fore is exerted everywhere inthe direction of the » axis, Lot » be the number of suspended particles per unit volume; then in the condition of dynamic equilibrium »is such a function of «that the varia- tion of the free energy vanishes for an arbitrary virtual displacement 8x of the suspended sub- stance, We have, therefore, aF = a8 — Tasto THEORY OF BROWNIAN MOVEMENT 1 will be assumed that the liquid has unit area of cross-section perpendicular to the x axis and is bounded by the planes v=o and x=1. We have, then, 8E — [Kobus and ‘The requiced condition of equilibrium is there fore RI we @ — Kg BP KoBeo. 2 2) “The last equation states that equilibrium with the force K is brought about by osmotic pressure forces. “equation (1) can be used to find the coeficient of diffusion of the suspended substance. We can Took upon the dynamic equilibrium condition con- sidered here as a superposition of two processes ‘proceeding in opposite directions, namely : 2. A movement of the suspended substance under the influence of the force I acting on each single suspended particle MOVEMENT OF SMALL PARTICLES 11 2. A process of diflusion, which is to be looked ‘upon asa result of the irregular movement of the particles produced by the thermal molocolar movement. Tf the suspended particles have spherical form (radius of the sphere =P), and if the liquid has 4 coefficient of viscosity &, then the force K im parts to the single particles a velocity (*) K pO ‘and there will pass unit area per unit of time Ba oiP particles. I, further, D signifies the coefficient of diffusion of the suspended substance, and j the mass of a particle, as the reault of difusion there will pass ‘across unit area in a unit of time, — DX gras (OL eg. G. Kuchhod, "Lectures on Mechanes,” List 26,5412 THEORY OF BROWNIAN MOVEME! Since there must be dynamic equilibrium, we ust have py ep We can calculate the cooficient of diffusion from the two conditions (2) and (2) found for the dynamic equilibrium, We get Rr x i DN Gk? 7 ‘The cotficiont of difusion of the suspended sub- stance therefore depends (except for universal constants and the absolute temperature) only on ‘the coeficient of viscosity ofthe liquid and on the ize of the suspended particles $4 Ow THE Innecutar Movawenr oF Parrici.es SUSPENDED 1Y A LiguiD aND THE RELATION or ta1s ro DisFusion ‘We will turn now to a closer consideration of the irregular movements which arise fom thermal ‘molecular movement, and give rise to the difssion investigate in the last parageaph. vidently it must be assumed that each single particle executes a movement which is indepen dent of the movement ofall other particles; the ‘movements of one and the same particle after MOVEMENT OF SMALL PARTICLES 13 different intervals of time must be considered as ‘mutually independent processes, so long as we ‘hink of these intervals of time'as being chosen. not too small We will introduce a timeinterval + in our dis- ‘cussion, which isto be very small compared with the observed interval of time, but, nevertheless, cof such a magnitude that the movements executed by a particle in two conseeutive intervals of time + are to be considered as mutually independent henomens (8). Suppose there are altogether # suspended par- ticles in liguid. Th an interval of time + the co-ordinates of the single particles will increase by 4, where d has a different value (positive or negative) for each particle, For the value of 4 ‘certain probability-law will hold ; the number dn of the particles which experience in the time- interval + a displacement which lies between 4 and 4 + 4d, will be expressed by an equation of the form, n= nf, where ee laayaa = and ¢ only lifer from zero for very small values ‘of d and fulfils the eondition 42) = 8-4).12 THEORY OF BROWNIAN MOVEMENT Since there must be dynamic equilibrium, we rust have 2K ® xk Pie We can calculate the coeficient of diffusion from the two conditions (2) and (2) found for the dynamic equilibrium, We get GP o ‘The cocficient of diffusion of the suspended sub- stance therefore depends (except for universal constants and the absolute temperature) only on the coefcient of viscosity ofthe liquid and on the size of the suspended particles, $4. Ow me Inzecutan Moveunwr oF PanricLes SusPENDED IY 4 LiguID AND THE RELATION (oF amis to Divrusion We will turn now to a closer consideration of ‘he iregular movements which arise from thermal ‘molecular movement, and give rise tothe diffusion investigated in the last paragraph, Evidently it must be assumed that each single particle executes a movement which is indepen dent of the movement ofall other particles ; the movements of one and the same particle after << MOVEMENT OF SMALL PARTICLES 13 diferent intervals of time must be considered as ‘mutually independent processes, so long as we think of these intervals of time’as being chosen. not too small ‘We will introduce a timeinterval + in our dise cussion, which is to be very small compared with the observed interval of time, but, nevertheless, of such a magnitude that the movements executed by a particle in two consecutive intervals of time ‘rare to be considered as mutually independent Phenomena (8). Suppose there are altogether # suspended par ticles in a liquid, Ta an interval of time + the co-ordinates of the single particles will increase by 4, where 4 has a different value (positive or negative) for each particle. For the value of 4 4 certain probability Jaw will hold ; the number dn of the particles which experience in the time- interval + a displacement which lies between 4 and 4 4 dd, will be expressed by an equation of the form ano npc, ficeayaa and only difers fom zero for very small values ‘of d and fulfils the condition Ha) = #4). where14 THEORY OF BROWNIAN MOVEMENT We will investigate now how the coeficient of ‘aitfusion depends on §, confining ourselves again to the ease when the number » of the particles por unit volume is dependent only on 2 and Potting for the number of particles per unit volume » = lz, 1, we will caloulate the distri bution of the particles at a time #4 + from the distribution at the time & From the definition of the function (4), there is easily obtained the number of the particles which are locsted at the time f + + betwoon to planes perpondicular to the r-axis, with abscisse x and x-+ dx. We get Sot aiem de. [Fes ayaa, —— fatto rte, we can expend r+ din powers ae Retain pepe BleD 4A MED atin very small, we can put ‘We can bring this expansion under th infagral sign since only very svall values of 4 contribute anything to the latter. We obtain Pons Kaus saan Uf Seon MOVEMENT OF SMALL PARTICLES 15 On the right-hand side the second, fourth, ete, terms vanish since g(x) = §(—); whilst of the first, third, fifth, etc, terms, every succeeding, term is very small compared with the preceding, ‘Bearing in mind that Joana =, ana puting 5 Bows, ane taking into consideration only the fist and ‘third terms on the right-hand side, we get from this equation &) yale This is the well-known diferential equation for — dp _ 8. ~~ B ishere defined by the condition that ule) must be the energy corresponding to the velacty of change f. In the case, therefore, of translational movement of thesphere fi, for example, the Kinetic energy of the liquid carted with it. Tt follows by integration 8 ‘We conclude from this result that the formula (Holy hots for intervals of time which are large ‘compared with uB. (22) For small bodies of 1 diameter and unit density fn water at room-temperature, the lower limit ‘of availabilty of the formula (1) is about 10-7 seconds ; this lower limit for the interval of time increases in proportion tothe square of the radius of the body. Both hold for the translational as well as forthe rotational motion ofthe particle, Berne, Decomber, x905. (Received, 19 December, 1905.)m1 ‘A NEW DETERMINATION OF MOLECULAR DIMENSIONS (Brom the Annalon der Physib (4), 19, 1906, pp. 28-306. Corrections, ibid, 84, 1011, pp. s9r-sg2) (23) IHE kinetic theory of gases made possible the earliest determinations of the actual dimensions of the molecules, whilst physical phenomena observable in liquids have not, up to ‘the present, served for the calculation of molecular dimensions, The explanation of this doubtloss Ties in the dificuties, hitherto unsurpassable, which discourage the development of a molecular kinetic theory of fiquids that will extend to details. Te will be shown now in this paper that the size ‘of the molecules of the solute in an undissociated dilute solution ean be found from the viscosity of the solution and of the pure solvent, and from the rate of diffusion ofthe solute into the solvent, if the volume of a molecule of the solute is large % MOLECULAR DIMENSIONS 37, compared with the volume of a molecule of the solvent. For such a solute molecule will behave approximately, with respect to its mobility in ‘the solvent, and in respect to its influence on the viscosity of the latter, a8 a solid body suspended in the solvent, and it will be allowable to apply to the motion of the solvent in the immediate neighbourhood of a molecule the hydrodynamic ‘equations, in whieh the liquid is considered homo- encons, and, accordingly, its molecular structure ignored. We will choose for the shape of the solid bodies, which shal represent the solute mole- cules, the spherical form. $1. On rmx Erruct ox raz Moniox oF & Lxgu1D OF A Very SwALE SPHERE SUSPENDED 18 1F ‘As the subject of our discussion, Jet us take an ‘incompressible homogeneous liquid with viscosity 2, whose velocity-components ux, will be given as functions of the co-ondinates , , z, and of the time. Taking an arbitrary point 9, Jy te We will imagine that the functions #, », w are de- veloped according to Taylor's theorem as fone- tions of x — xy, — yy — Zand that a domain G is marked out around this point so small that within it only the linear terms in this expansion38 THEORY OF BROWNIAN MOVEMENT hhave to be considered. ‘The motion of the liquid contained in G can then be looked upon in the familiar manner as the result of the superposition of three motions, namely, 1. A parallel displacement of all the particles of the liguid without change of their relative postion. 2. A rotation of the liquid without change of the relative position of the particles of ‘he liquid 3. Amovertent of dilatation in threo directions At right angles to one another (the prin- cipal axes of dilatation). We will imagine now a spherical rigid body in the domain G, whose centre ies at the point ‘du and whose dimensions are very small com- ‘pared with those of the domain G. We will further assume that the motion under considera- tion is s0 slow that the kinetic energy of the sphere is negligible as well as that of the liquid, 1e-wil be further assumed that the velocity com- ponents of an element of surface of the sphere show agreement with the corresponding velocity components of the particles of the ligula in the ‘mmeaiate neighbourhood, that i, that thecontact- layer (thought of 3 continuous) also exhibits —e MOLECULAR DIMENSIONS 39 everywhere a viscosty-coeficient that is not vanishingly small. It is clear without further discussion that the sphere simply shares in the partial motions rand 2, without modifying the motion of the neighbouring Tiguid, since the liquid moves as a rigid body in ‘hese partial motions ; and that we have ignored the effects of inertia. ‘But the motion 3 will be modified by the pres- ‘ence of the sphere, and our next problem will be ‘to investigate the influence of the sphere on this ‘motion of the liquid. We will further refer the ‘motion 3 to a covordinate system whose axes are parallel to the principal axes of dilatation, and we will pat xm 6, yon roast ‘then the motion ean be expressed by the equations un AB, « 4 = By, w=, in the case when the sphere is not present, A, B, C are constants which, on account of the Incompressibiity of the liguid, must fulfl the condition ® A+B+C=0 + (a)40. THEORY OF BROWNIAN MOVEMENT Now, if the rigid sphere with radius P is intro- duced at the point x,y % the motions of the liquid in its neighbourhood are modified. In the following discussion we will, for the sake of eon- venience, speak of P as "finite"; whilst the values of 9, { for which the motions of the liquid are no longer appreciably influenced by the sphere, we will speak of as infinitely great.” Firstly, it is clear from the symmetry of the motions of the liguid under consideration that ‘there can be neither @ translation nor a rotation cof the sphere accompanying the motion in ques- ‘tion, and we obtain the limiting conditions « owhen p= P ‘where we have put p= JET REED. Here u, », » are the velecty-components of the sotion now under consideration (modified by the sphere). Ifwe put wnat ay @ vein te, Cot, since the motion defined by equation (3) must be ‘transformed into that defined by equations (r) in the “infinite” region, the velocities u, oy, will vanish in the later region. MOLECULAR DIMENSIONS 4r ‘The fonctions w, v, © must satisfy the hydro- aynamic equations with due reference to the viscosity, and ignoring inertia. Accordingly, the following equations will hold :— (*) where d stands for the operator wt aa ta tag and p forthe hydrostatic presire. Since the equations (2) are solutions of the equations (4) and he latter ar linear, aozording {0 () the quantities wy», must also satisfy the equations (@ Thave determined ya, according to a method given in the lectre of Kirco quoted in § 4 (f), and find (7G. Miedo, * Ltr on Mech” Lect. {h"From the ountons (It fllms at ap =o cps chore i atortane th tha cad, tad ‘ction ie deterained wich sles te option ava, ‘en tho equations (4) a ated we pat wae, 0 Mae, Wau, on pe,42° THEORY OF BROWNIAN MOVEMENT pp 2Q ae Pa +0") const, jum ag —Spral, 2, 9) fr} -2 w= Of — Sch —% Nowitwept a the constants 0, 8, « canbe chosen so that when p= , wao=se=o, By superpestion of three similar Selatons we obtain the solution glvan in the equations (Gh and (a) MOLECULAR DIMENSIONS 43 pa agate”) alge ad G) Spi 4 tpele +ofgrte sy i ae}. eis easy to see that the equations (5) are solu tions of the equations (4). Then, since where Memo, Bm o, dpm? and Spy need +p Bh ten} Bt the lst expression obtaina i, acconing to ‘the first of the equations (5), identical with dp/dg. In similar manner, we can show that the second Sp, Pa44 THEORY OF BROWNIAN MOVEMENT and third of the equations (4) are satisfied. We obtain further bu | be, aw eth ty sea?) 4 sola eee ad Bit since, according to equation (5a), one it follows that the last of the equations (4) is Salised, As for the boundary conditions, our ‘equations for u, », w are transformed into the ‘equations (r) only when p is indefinitely large. By inserting the value of D from the equation (ss) in the second of the equations (s) we get ©) wm ae SZ eae + Boe + CoH + EF eae + Bot + 0m) — Fae ts) We know that u vanishes when p=, On the rounds of symmetry the same hols fr» and w. ‘We have now demonstrated that in the equations (5) @ solution has been obtained to satisfy bott “4 MOLECULAR DIMENSIONS 45, ‘the equations (4) and the boundary conditions of the problem. tan also be shown that the equations (3) are the only solutions of the equations (4) consistent with the boundary conditions of the problem. ‘The proof will only be indicated here. Suppose ‘that, in a finite space, the velocity-components of aliquid u,v, w satisfy the equations (4). Now, if another sotution U,V, W of the equations (4) can ‘exist, in which on the boundaries of the sphere in question U=u, Vv, Wa, then (U—, V —2, W—w) will be a solution of the equa- tions (4), in which the velocity-components vanish at the boundaries of the space. Accordingly, no ‘mechanical work can be done on the liquid con- tainod in the space in question. Since we have Jgnored the kinetic energy of the liquid, it follows ‘that the work transformed into heat in the space in question is likewise equal to zero. Hence we infer that in the whole space we must have w= 1, v=o, w=, if the space is bounded, at least in part, by stationary walls. By crossing the boundaries, this result can also be extended to ‘the case when the space in question is infinite, as in the cage considered above, We can show thus ‘that the solution obtained above is the sole solution of the problem,46 THEORY OF BROWNIAN MOVEMENT ‘We will now place around the point sphere of radius R, where R is indefini compared with P, and will caleulte the enerey ‘hich is transformed into heat (per unit of time) in the liquid lying within the sphere. This energy W is equal to the mechanical work done on the iquid. "If we call the components of the pressure exerted on the surface of the sphere of radiue R, Xo, Ye, Zn then We f (Xu t Yo + Zyw)ds ‘where the integration is extended over the surface of the sphere of radius R. Here aan (mbt xtead), Yn (¥f +¥d4 vd), ~(ab+ 22428), where Kempo, Yen yn (EE) Yeah aXe a(R Foon ast Men + 3) MOLECULAR DIMENSIONS 47 The expression for u,v, are simplified when we note that for p= R the terms with the factor Pip! vanish We have to put pfs bt or) By — Spd Bot + CB) 2 wn ot ile bys For p we obtain from the frst ofthe equations (5) by coresponding onisions pa sipB t BP $C | const We obtain frat Kyo nakd tron xy sll aghpnbte Net gh LOEE ayy Ae BoC) ‘and from this AE agpantin ce) Bye) With the aid of the expressions for Ys and Zy, obtained by eyelic exchange, we get, ignoring all 3) . (23)48 THEORY OF BROWNIAN MOVEMENT tems which involve the ratio Pip rused to any power higher than the third, au Yooh Za = Rane + Bh + OH) TAME Bet 4 C1) gh (AGH Boh COP (29) [If we integrate over the sphere and bear in mind that Sds = 4, S Grds mS gids = § (ds = $e, S elds = $ ntds = § Leds = teks, S oHthds = § Eds = § Byhis = Yo, SALE + CES Ha B EC, (25) wwe obtain () We beRovae + xP where we pat Rea BEC, R= V and teP* a” x wav +2), (a) If the suspended sphere were not present (D = 0), ‘then we should get for the energy used up in the volume ¥, ys) W = aR, On account of the presence of the sphere, the energy used up is therefore diminished by 8%. (26) a MOLECULAR DIMENSIONS 49 $2. CALCULATION oF tie Viscosrry.CoRFFICIENT oF A Liguip 1x waren a Lance Newsex. or SwAtE Sritemes ARE SusPENDED 1 IR- REGULAR DistRIBuTION In the proceding discussion we have considered ‘the case when there is suspended in a domain G, of the order of magnitude defined above, a sphere ‘hat is very small compared with this domain, and have investigated how this influenced the ‘motion of the liquid, We will now assume that fan indefinitely large mumber of spheres are dis- tributed in the domain G, of similar radius and actually so small that the volume of all the spheres together is very small compared with the domain G. Let the number of spheres present in tunit volume be m, where 1 is sensibly constant ‘everywhere in the liquid. ‘We will now start once more from the motion of a homogeneous liguid, without suspended spheres, and consider again the most general sotion of dilatation. If no spheres are present, by suitable choice of the co-ordinate system we ‘ean express the velocity components ty a, in the arbitrary. chosen point x, 9, #in the domain G, by the equationsso. THEORY OF BROWNIAN MOVEMENT where APBHC= Now a sphere suspended at the points, J» 2 will affect this motion in a manner evident from ‘the equation (6). Since we have assumed that ‘the average distance between neighbouring spheres is very great compared with their radius, and ‘consequently the additional velocity-components originating from all the suspended spheres to- gether are very small compared with 1 5, wwe get for the velocity-components u,v, w in the liquid, talking into account the suspended spheres and neglecting terms of higher orders— | ae — 2fB SLAG Bt Ct | _ 5 P Blass canbe MOLECULAR DIMENSIONS 51, where the summation fs extended over all spheres fn the domain G, and we put baron BAIR ETE bats 4 Jp 4 ane the co-ordinates of the centre of the sphere. Further, we conclude from the equa tions (7) and (7a) that the presence of each of the spheres has a result (neglecting indefinitely small ‘quantities of « higher order) (23) in an increase ‘of the heat production per unit volume, and that ‘the energy per unit volume transformed into heat in the domain G has the value W = ath + nono, (es) ) where ¢ denotes the fraction of the volume oocu- pled by the spheres. ‘From the equation (7b) the viseosty-coeficient ‘can be calculated of the heterogeneous mixture of Tiguid end suspended spheres (hereafter termed briefly “mixture") under discussion ; but we ust bear in mind that 4, B, C are not the values ‘of the principal dilatations in the motion of the liquid defined by the equations (8), (23); wevwill all + 3)52 THEORY OF BROWNIAN MOVEMENT ‘the principal dilatations of the mixture 4, 2, C*. On the grounds of symmetry it follows that ‘the principal directions of dilatation of the mix- ‘ure are parallel to the directions of the principal dilatations 4, B, C, and therefore to the co-ondi- nate axes, If we write the equations (8) in the oom By wa Cr+ Eu, If we exclude from our discussion the immediate neighbourhood of the single spheres, we can omit ‘the second and thied terms of the expressions for 4,9, , and obtain when x= y= z= 0:— 5Pix(dnt + By + Os, | pm SB uae att C2, pee SPH Bet 02 where we pat 12 SERED 0. MOLECULAR DIMENSIONS 53, We extend the summation throughout the velume of a sphere K of very large radius R, whose centre lies at the origin of the co-ordinate system. If wwe assume further that the iregularly distributed spheres are now evenly distributed and introduce an integral in place of the summation, we obtain 7 a, AtaAnn i) Feed, a aaafit de. tan) ‘where the last integration is to be extended over the surface of the sphere K. Having regard to (0) we fina ard SEs fate + Bat + Cos An a(fen) dee) By analogy Bt = Be 9), cmc 9). We will put 3 AML BOM, then neglecting indefinitely small quantities of Digher order, be = BYE — 29)54 THEORY OF BROWNIAN MOVEMENT We have found for the development of heat per unit of time and volume wired). . a Let us call the viscosity-coefficient of the mixture 2s, then we We aes, From the ast three equations we obtain (neglecting indefinitely small quantities of higher order) Re +258) ) We reach, therefore, the result — Tf very small rigid spheres are suspended in a liquid, the cooficent of internal friction is thereby {increased by a fraction which is equal to 2°5 times, the total volume of the spheres suspended in a ‘nit volume, provided that this total volume is very smal §)3. Ov Tun Voruue or 4 Dissorven Sunstaxce oF Mouucuan Vorvser Lanox x Comparison \WITH THAT OF THE SOLVENT Consider a dilute solution of a substance which does not dissociate in the solution, Suppose that a molecale ofthe dissolved substance is large com- pared with a molecule of the solvent ; and ean be thought of asa rigid sphere of radius P, We can ‘then apply the result obtained in Paragraph 2, MOLECULAR DIMENSIONS 55, 1 be the viscosity of the solution, & that of the pure solvent, then Martasé.. +. 23) where 4 is the total volume of the molecules present in the solution per uit volume. ‘We wil calculate 4 for a x percent. aqueous sugar solution, According tothe observations of Burkhard (Landolt and Bomstein Tables) Je h-= rons at 20° C) for ax percent. aqueous sugat solution; therefore 2460-024 for (approxi rately) oor gm. of sugar. A gram of sugi dis solved in water has therefore the same eflect on the viscosity a8 small suspended rigid spheres of total volume 098 e. (3) We smostreolect here that x goof slid sugar tas the volume 06: ce, We shall find the same value forthe specific volume ofthe sugar present {solution ifthe sugar solution is looked upon as a mixture of water and sugar in a dstlved form ‘The speiic gravity ofa x percent aqueous sugar solution (referred to water at the same tempera ture) at 175° is 100388. We have then (nglect- ing the dltfeence in the density of water at 4* and ot 27'5)— saligg~ 909+ 00, ‘Therefore $= 061.56 THEORY OF BROWNIAN MOVEMENT While, therefore, the sugar solution behaves, as to its density, like a mixture of water and solid ‘sugar, the effect onthe viscosity is one and one-half times greater than would have resulted feom the ‘sospension of an equal mass of sugar. Tt appears to me that this result can hardly be explained in the light of the molecular theory, in any other ‘manner than by assuming that the sugar mole- cles present in solution limit the mobility of the water immediately adjacent, so that @ quantity fof water, whose volume is approximately one- half (23) the volume ofthe sugar-molecue, is bound ‘on to the sugar-molece. We can say, therefore, that a dissolved sugar molecule (or the molecule together with the water hheld bound by it respectively) behaves in hydro- lished at the end of § 2, equilibrium with which must be maintained by the force K. (6) p. 21—This expression for the resistance experienced by a sphere in a uniform movement of translation through a viscous liquid was first deduced by Stokes hydrodynamically, with the assumption that the liquid adheres completely to ‘the surface ofthe sphere and its velocity becomes vanishingly small: so that the velocity of motion does not exceed a certain value, ‘There is no doubt that when the above condi- tions are fulfilled Stokes" formula really gives the ‘motion of the sphere accurately, but it i a ques- tion whether the conditions are'really fulled in ‘the case of the Brownian motion of very small spherical particles. ‘Then, on account of its derivation, the formula is only valid for the case when the hydrodynamic ‘equations still hold, which from the Atomic point ‘The expression Ky — NoTEs 93 of view can only approximately be the case 80 Jong as the radii of the spheres are large compared to the free paths of the liquid molecules, This condition is actually fulilled by the particles of visible size in liquids, but notin gases, so that it Is necessary in the latter ease to apply certain corrections to Stokes’ formula, which can be derived by consideration of the Kinetics of gases. ‘The frst correction of this type wos given by E. Cunningham for the ease when P is compar- able with the free path of the gas molecule. It appears that the expression for the velocity ofthe particle must be multiplied by the factor TEAL, Where 4 is a constant which can have values between 815 and 1-63, according as to whether all impacts of the molecules against the particles are elastic or inelastic in nature. According +o F, Zerner, however, these limits must be corrected to r4oand 13754 Experimental tests of the law of resistance in ‘gases have been carried out by diferent investiga tors: by M. Knudsen and S. Weber by variation of the gas-pressure whilst employing a fixed size of sphere, and by L, W. MeKeehan * also for Cunningham, Prot. Roy. See. (A), 88,357 VE, Lerner, Phys. Zeit 2, $96, 1909, 1 Koadaen and Weber, At de Phy 8, 981, 1910 ‘exeehan, Phys Zeit, 12, 707, 14094 THEORY OF BROWNIAN MOVEMENT different sized spheres. They obtained the fol- lowing empirical formla for the correction-factor 968 function of the quantity Pil:-— P. A= 068 4 0356 ‘This formula also holds for the case when the radius of the sphere is small compared with the free path, The researches of B. Meyer and W. Gerlach,’ and of J. Parankiewiea* are in agreement with this, whilst J. Rowx® obtained values between 1-23 and 163. That there can bbe no agreement with Cunningham's law is also ‘evident from the researches of R, Furth * on the ‘determination of mobility from the Brownian movement, Te appears, further, that it might well be assumed that the velocities of the particles in- ‘volved in the Brownian movement remain below ‘the limits for which the Stokes formula is valid It can be taken, from an investigation of H. D. ‘Amnold,* that the Stokes formula holds below a 2, Meyer and W. Gerlach, Eltr Geil asl, Views, PP. 196, te Sy Pamalewice, Phys. Zit 1, 286, 1918. 85, Roux, dion. de Chin ef Phy, wi, 29, 69, 913 CR Barth, don d. Phys, 60, 77, 3919) 6, 52, enol, Phi Meg, 28, 735, 2908 Notes % vey V whichis th eq onion PW cos, ‘where « indicates the specific gravity ofthe liquid. As is shown by an approximate calculation, it i, in general, scarcely to be expected that the veloc ‘es resulting from the Brownian motion could reach these upper limit, Finally, it must be borne in mind that the Stokes formula is deduced for constant motions of translation and established experimentally for ‘these conditions ; so that itis not impossible that considerable divergencies may occur with ac- celerated motions. For certain special cases of accelerated motion, the form of the law of resist- tance has also been determined theoretically, eg. for the case of small pendulum vibrations and can also be established experimentally in a satise factory manner. Tt is then a question whether it is permissible to apply the Stokes law to the Brownian motion, which in reality exhibits no regolar translation, but hasan iregulsr character. ‘Again, if the formula cannot be applied with cer. tainty to the single zig-age of the Brownian ‘motion, ican be still assumed that, on account of ‘the irregular character of the motion, the deper- ‘tres from the Stokes law cancel out on an average Einstein's deduction given here corresponds 1096 THEORY OF BROWNIAN MOVEMENT this thought: in which there is assumed to be 8 statistical equilibrium between the process of difsion and a fictitious constant force K. Tn addition, the Stokes law has been proved for accelerated motions of a regular character at high frequencies in gases, and is known to hold swith very close approximation up to periods of about 60 per second. ‘A detailed discussion of all the problems sug- ‘gested here isto be found in a paper by J. Weysen- To? (7) p. 22-—It is notable that the result for D nno longer contains the applied force K. This ‘ust, however, be the case, if the method intro- duced here is successful in its aim, since K is a completely fictitious force which has nothing to do with the process of diffusion itself. This eir- ccumstance indicates that it must also be possible ‘to obtain the result without the introduction of thetictitionsforce, Such deductions haveactually Dean eartied out ; amongst others might be par- ‘ienlarly mentioned, on account of its special simplicity, the dedsction by Ph. Prank’ with EN AL Shower, Pips, Ree. (2), 8425, 19075 RB, Abate Phy fen, 4, 38%, 1918; A Suothlag, Vora. K, Abad. v, Welsh. Amst, 25, 3873, 1937; A Fith, dn. Phys, 68, 528, 1920 1), Weyeeala dn d, Phy By 2, 1929. 1S Ph, Fran Wier Bor, 196 (aa, 2173, 1915; Am 4. Phy, BB 323, 1937. NOTES 7 the help of the conception of the Viril introduced by Clausius @) p. 14—The introduction of this time- interval + forms a weak point in Einstein's argument, since i is not previously established ‘that such a time-interval ean be assumed at all. For it might well be the case that, in the observed interval of time, there was a definite dependence of the motion of the particle on the initial state ‘A deduction of the formula for the Brownian ‘motion, which does not involve this presupposi- tion, has been given by L. S. Ornstein * according to a method suggested by Frau de Haas-Lorents,* as well as by R. Furth” (in agreement with the former) by another method. In contrast with the formula of Einstein given. ‘on page 17 of the tex, this formula runs at mB 4%) where B= s/60AP indicates the " Mobility” of ‘the partele and ym its mass, For a silently large time-interval, the formula actually comes {nto line with Einstein's, whist for very short 1S, Orme, Pro: Amst, 81, 96,1918, "1. de Haat-Lovnts, “The Brownian Movereat and ome’ Related Phencmens,” Die Wissel, B52, Veewet 903 SR Prth, Zi J Phys, 2,244, 1920.98 THEORY OF BROWNIAN MOVEMENT times it indicates a sectiinear and. uniform rotion ‘AS an approximate calculation shows, the instein formula holds for particles of a size that can be observed, under al circumstances. (0) p. 26-—There Js sought here a socalled “ Source-Integral” (" Quellenntegral ") of a di ferential equation of the second order, that i, @ solution for which the boundary condition is ‘assumed as a definite value for the integral of ‘the diffusion-steam over the source. The corre- sponding diftesion problem is: if at time t= 0 the concentration of the diffusing substance is everywhere zero with the exception of am indefi- italy narrow space around the plane x =o, but Such that the whole mass of the substance is ven af ll times by [ene ‘then the formula given for lt, will be the ex- pression for the distribution of the concentration ofthe substance at some later time # and at any point 2! ‘Similar methods have been applied to diferent problems of the Brownian motion by Smoluchow- 2 Video B. Riemann- Weber," The Partial Dies. tial Equations of Mathematical Phyicg” 48h edt Book 2 p90 NOTES 90 ski. Schrodinger," and Furth, by the solution ‘of the diffusion’ equation under corresponding voundary conditions. (10) p. 16—The meaning of the probable dis- tribution found J¢s wen is as follows: one imagines a lange number of similar particles accumulated at the time ¢= 0 in the immediate neighbourhood of the plane =o, and then left to themselves ; now, after fa time f such a distribution of the particles is spontaneously established that the relative num- ber of particles between the planes » and x-+ dx is given by (fds. Here we assume that the particles exert no forces on each other. Such a summation of systems may be called in statistical mechanics a “‘spacesunmation.” If we now ook upon this space-summation as. single system and imagine that a very large aumber of exactly similar systems are set up, and the same exper ‘ment carried out with these, i ie asserted that the Sis.8) = +The reader wil find further information on thie subject in the next volume of the Ostwali's” Kanaer, ‘which comprises Setlachowsi's paper on the Brownian 'E Sehrdinger, Phys, Zot, 48,289, 295. 836 Furth, Aim d Phys, 88, 277,160.to9 THEORY OF BROWNIAN MOVEMENT ‘mean value of the distribution g(x, ) obtaining in all these systems at the time ¢ will correspond exactly tothe function fs, ) in the above formula, ‘The purely imaginative summation used here is called a " virtual summation.” For an approxi ate realization of this one can proceed in such fa manner (asthe experimenter is generally careful ‘0-do) that we make use only of one and the same spece-summation, and after carrying out an vestigation, this is always brought back to the criginal condition by artificial means. But another important meaning can be given to the formula if we consider as our system not, as before, the whole space-summation, but the single particle in this epacesummation, Then f(s #) Genotes the probability that the particle has ‘been displaced in the time f to a region between sand xf dx, If one observes the movement of 4 single particle and notices the displacements experienced in suocessive intervals of time, the relative frequency of these displacements. will likewise be given by our formula, in the limiting case of an indefinitely large number of observa~ tions. This is what is indicated in statistical ‘mechanics by " time-summation.” Both methods of observation are actually carried out in connection with the Brownian ‘movement, and both lead to the establishment of Einstein's formla, Tt would appear also a8 if NOTES ror in thie case the mutual substitution of the two kkinds of summations were permissible without any further conditions, which is most decidedly not self-evident from first principles. The ques- ton a8 to the exchangeability of virtual and time- summations belongs to some controversial points in the foundations of statistical mechanics. Tt can be shown that this exchange can always be carried out when a so-called“ ergodisch " system fs in question ; yot it has not been possible up to now to give a singlo example of such an “ ergo- dlisch" system, The reader will ind a compre- hensive discussion of related probloms in the article on “The Conceptual Foundations of the Statistical Treatment of Mechanics,” by P. and ‘T, Bhrenfest A paper of, v, Mises? intro- ices a new view-point (G2) p. 17—Acconding to the definition of the ream value the mean square displacement is ‘obtained from the expression YT 000 = 7p Brew tlre + ire al = alerw 2D. att, pat 6. SAL te, Pye, Zoi, 24,295, 1920.to2 THEORY OF BROWNIAN MOVEMENT (02) p. 17 Iris the total daplaoement ofthe patil, ten reverts therfore a RABE PEA and since on account of the homogeneity of the liquid these are all equal— Paw, JR Je. (13) p. 28.—The wish expressed ere by Binstein was very soon full, resulting in & complete confirmation of his theory. Amongst the numer- cous experimental investigations there will only be tentioned here those which have given a direct confirmation of Binstein's formula in its original meaning. The fist of these investigations was carried ont by Sedlig who took two photographs ‘of an aqueous suspension of cinnabar on the same plate at an interval of or second, and measured ‘the distance of corresponding images on the plate. ‘He found that on an average the displacements at diffrent temperatures were inversely proportional to the viscosities, as the theory demanded. Henri found similarly with the aid of cinemato- AR Soltis, Phys Zah, 9,463, 1908 Zot f. Blake rch, 1,36, 1912 "Vr Hens Comper Rendur, 148, 2024, 1908; 187, 2, 1908 NOTES 103 ‘graph records of the mean dieplacement of par- ticles of eaoutchoue that the time law, 24 propor- tonal to , was followed. ‘The establishment of the first complete and absolute proof of the formula lies to the credit of Perrin’ and his pupils Chaudessigues, Dabrow- ski, and Bjerrum, who followed the movements of single particles of gamboge or mastic under a microscope and recorded their postions at equi- distant time intervals by means of an indicating apparatus. In this manner they could also use ‘the formula to determine the Losehmidt number A in a new way, and found values between s6 and ‘88-10% They could also confirm the distribution law for the probability of different displacements ‘fe, in a quite unexceptionable manner. Swedberg? and Inouye made their measure ‘ments in similar manner in metal sols of appre- ciably smaller particle size, and found a good agreement with the formula with large particles, Dut systematic departures from it with small particles: this is most probably to be ascribed to a breakdown of Stoker law in. connection with very small particles. They found the time- law well confirmed, and obtained approximately “PerimLottermoser, "The Atom,” Lelpdlg end Dresdea, 1944 "Th. Seedborg, “The Existence of the Malesia" Lalprig, 112.tog THEORY OF BROWNIAN MOVEMENT 6x30" for Loschmide’s number. Similar results were oblained by Nordlund with an autometicr registering photographic arangement. nally, a paper of K. Sedlis* should be men- tioned, which forms a continuation of Sedi’s ‘work and amplifies it suitably ‘The Brownian motion was first desribed in gases by F, Ebrenhalt# who has also shown that the onder of magnitude corresponds with the Einstein formula. A ditect confirmation of Ein- stein's formsla is out of the question here, since the determination of the size of the particles ‘cannot yet be camied out with accuracy ; never theless, it appears to be established ffom the former investigations that the formola can be appli with accuracy inthis cas too (G4) p. 23-—The law of Boltzmann mentioned here by Einstein is that known by the name of the etetheorem, which plays a great part in Statistical Mechanes. Te leads to the inference {hata system which i in statistical equilibrium vith another of indefinitely large energy i actually subject to Muctnations, whose relative frequency fs ven by the law (2), Wherever, therefore, *K, Soa, Zeit f Phys, Chem, 8, 682, 1904 E Ebronbate, Wiener Ber, 148, (Oa), 1139 1997 "Butter deals will be found in the eeport of Ta, Sredburg, Jahbuck der Rad. sw. Ellr, 10, 467, rong and Feet, i, 16, 319, 1920. NOTES 105 similar equilibria exist this lw should be applied for the calculation of the magnitude of the Buc- tuations. The reader will Gnd a detail dis sion of all phenomena of this type in Physics in Firth’s paper, “Fluctuation Phenomena in Physics.” * (G3) p. 2—M. v. Smaluchowski* has given @ detsiled theory ofthe Brownian movement under the influence of an elastic force, and, in this particular ease, has set forth in a very pleasing manner the points of agreement and diferences between the statistical and purely thermodynamic conceptions of natural proceses, especially con- cerning the apparent contradiction betveen the Principal reversible mechavical and the irever- Sible thermodynamic processes, He pointed out also that this ease can be verified experimentally by observation of Brownian torsional vibrations of a small mizrr fastened on a thin thread, or of the vibrations of the free end of a thin eastic quartz fibre, The last suggestion was experi- ‘mentally veried quite recently by P. Zeeman, though finaly satisfactory results have not been obtained up to now. © Forth, Viewos Callestion, No. 48, Brasowick, ono: and Phy, Zeit, 20, 1919. "ik. Smelochowi, Katower Bor, 418, 1913: Lectura on "Tho Kinelle Theory of Matter and Bloc teety a.106 THEORY OF BROWNIAN MOVEMENT (46) p, 25.—If it be assumed that the energy E of the system is continuously divisible, from the Gefinition of the mean value of a function we ‘obtain for the mean energy per degree of freedom, regarding the energy itself as the parameter , from the expression (1)— Accordingly the mean enargy of a linear osit- lator is also equal to this quantity. On the other ‘hand, Planck has shown in the paper quoted that the mean energy of such an oscillator, whieh is in dynamic equilibriam with the radiation in a hhllow body, is given by Lb Bom er, where L indicates the velocity of light, v the fre quency, and pray the energy of that part of the radiation per unit volume whose frequency Ties NOTES tor Detween y and »-+de, By equating the two expressions, i¢ follows that aT_ yD Wo a" nd thence the expression given in the text for pv ‘Bat ifit be assumed that the energy F isnot dis- tributed to the oscillator in a continuous manner, Dut only in multiples of an elementary quantum div, where his universal constant, as Planck and “Einstein have assumed, there is obtained for E a ‘sammation of the form ‘As can be seen by expanding the denominator, for small values of » and for high temperatures respectively this formula becomes B = RT, which agrees with the former expression. In general, however, by equating the value of E with Planck’ value given above, there is obtained by108 THEORY OF BROWNIAN MOVEMENT from which follows to a) ‘Which is in agreement with Planck's radiation law. Hence the " incompleteness of our physical conceptions ” perceived by Einstein is related to the necessity for introducing the quantum hy- pothesis, (27) p. 27-—This expression ean be interpreted again, in the meaning of statistical mechanics, in ‘evo ways (vide Note 10). [fone considers a very great mumber of particles similar to one another, it gives the relative mass of those particles ‘hich will be found on an average at a height 2... sds above the ground. As the form of the expression shows, the "space-summation " of the particles corresponds with the well-known vertical aerostatc distribution : which is implied by the nature of the case, since only a quantita- ‘ive, and not a qualitative, distinction exists between the gas built up of molecules and a sus- pension of microscopic particles, We can now vestigate whether a suspension of small, similar particles is actually arranged in accordance with ‘this formula, and, on the other hand, whether it agrees with the absolute figure, ie. whether by determining the other data Loschmidt's number can be calculated. This was the manner in NOTES 109 ‘which, forthe fist time, an exact confirmation of Einstein's theory was obtained, a result for which thanks are due to Perrin and his pupils (vide Note 13). Their method of procedure was first ‘to prepare, by the device of fractional centri= fuging” worked out by Perrin, « suspension of {gamboge or mastic with particles of exactly equal size, This was then enclosed in a microscopic chamber and the distribution in height of the particles determined, after equilibeium had been established, by counting in the microscope the particles in diferent layers above the bottom of the chamber, In order to facilitate the counting, a small sereen was introduced in the ocular of the ricroscope, so that at all times only a small number of particles were in the field of vision at the same moment ; these were made visible at regular intervals of time by intermittent illumina- tion of the preparation, and 20 a great number of observations were arranged for. Observations carried out on different sizes of particles and sspension-media showed throughout a very good. agreement between theory and experiment, and for Loschmnide’s number figures between 55 and Bo ro! errin’s investigations have been recently ex tended by Westgren to still smaller particles of 1A, Wentgren, Zeit f. Phys. Chom, 8B) 15h 1913 5 80,65, 1504! dich, fu Mat, Aut. och Fysihy 14 Now. 8110 THEORY OF BROWNIAN MOVEMENT colloidal metals, which were prepared of equal particle-size by the germ method of Zsigmondy. ‘The method of observation was somewhat similar to that of Perrin; the results of the extremely accurate investigation were in complete agreement with the exponential distribution in height ex- pected, and give, therefore, pretty well the most accurate direct determination of Loschmidt’s ‘umber, which was fixed thus as (606+. 20). 10%, If the concentration of the particles becomes so great that the forees (operative at a distance) acting between the particles exclude the assump- tion of the mutual independence of the particles, we must expect divergences from the aerostatic istrbution in height. Actually, Costantin* found such divergences in gamboge emulsions at very high concentrations, following Perrin’s ‘method : which can be accounted for, in the ‘manner indicated, by forces of repulsion between ‘the particles, which are evidently of an electrical nature But principally we can look upon our formula in quite a different manner, namely, asa picture of the timesummation for a single particle, if we imagine that a single particle is freely movable fad 4, 1916; Zeit fang, Chom, 98,238, 2015 95, 59, 106, SR Costantin, CR, 188, 1175, 19843 Am de py, chim. (0) 8 14,1915 NOTES mr above & floor to which it eannot adhere, Then, 1s a result of its Brownian motion, it will not simply sink to the bottom, as one might expect, ‘and remain lying there, but will always raise itself spontaneously and dance hither and thither in an irregular manner, The formula gives, then, simply the “ relative length of sojourn” of the particle in thediferent layers . . . x + deabove the bottom within a long period of observation. ‘This procedure was introduced by R. Furth,* and he was able to show that the formula applies ‘very well to the movement of the particle: here there is forther the advantage over the former Imethod that i is necessarily independent of iregulatity of particle sire and of forces acting at a distance between particles. Tt was also possible in this manner, by simultaneous determination of the size of the particle according to Stokes’ law, to make a fresh determination of Loschmidt’s number; which gave N= 64. 10% We see established hore in the most pleasing manner Einstein's assertion thet the particles can be ‘suspended in the liquid if they be fine enough. (G5) p. 30.—On account of the small magnitude of 4, we can put the lower limit of the integral in tig equal to zero, and develop the fonction F in the following manner — Foy i A) = Fla) + 4. F() 1, rh, nd. Pye, 89, 37,2987112: THEORY OF BROWNIAN MOVEMENT ‘Thence it follows further that aF eal eaef, Halo 1 we exchange here the oder of the integrations ‘the limits for ¢ willbe o and 4, and the limits for 4,0, and co, from which i immediately follows ‘tat E =m = — Fad 429d Fle arpaad ‘on account of the condition that WA) = H-4), ig) p. 31.—In the original paper there is given hhere in error 4. 108: actually the value of Loschmidt’s number 6 «108, according to themost aocurate measurements that we possess up to the moment. (20) p. 33.—The formula for the rotary Brownian movement was established in 1909 by Perrin, by suspending spheres of mastic of about 12, diameter in water and following their rota- tional movement, as & function of the time, by observation of small, differently coloured inserts in the particles. The formula could thus be closely confirmed, as well as the absolute dimen- sions: for Loschmidt's number was obtained 05.108. NOTES 13 (21) p. 33—If we call ¢ the quantity of elec- tricity that is displaced across any given cross section of the conductor in time f, and identify « with the quantity of electricity which has flowed across this cross-section since the time =o, then 4 =, and 2 will be the current i. The potential energy corresponding to the displace ment 4 is evidently identical with the electrical potential difference, and hence the fictitious force of the EME. £. According to its definition, therefore, Bash = r/o if a indicates the resistance of the closed circuit We obtain therefore A number of similar questions which are form- ally related in the closest manner with the Brownian movement have been dealt with by Frau de Haas-Lorentz* according to the method of Einstein and Hopf The list of possible in- vestigations given by her could be considerably increased: however, it has not yet been possible to discover these phenomena experimentally, 6.1. de Hass Lorente “The Brownian Movement tnd Related Phenomena,” Sanming Wisenzhf 32, ieweg, 10, Sho lneein and Hopf, Aven & Phy, 98,1105, 1016,114 THEORY OF BROWNIAN MOVEMENT since on account of their minuteness they escape ‘our measuring instruments.? (2a) p. 35-—For the lower limits of validity of Binstein’s formola and its substitution by a more accurate one valid for any desired small time- interval, refer to Note (6), R. Firth has also derived an estimate for the lower limits of validity, from the formula quoted above and from other considerations communicated in the seme paper, and arrives atthe conclusion that this time must be of the order ms, where m indicates the mass and B the mobility ofthe particle. Actually, one ‘obtains in this manner, i. in the order of magni- tude, an agreement with Einstein's estimate. (25) p. 36.—A correction of the following paper appeared a few years later with the ttle: A Eine stein, " Correction of My Paper, ‘A New Deter- rmination of Molecular Dimensions” (An. 4. Phys, 84, 592, 1977), in which some numerical terrors in the previous communication were rectified, which had also some influence on the results, In the reprint given here the resulting corrections are already introduced in the text with the aid of the paper mentioned, in order to facilitate the reader's task. The points of comrec- tion are indicated in the text by reference to thi note 'R. Furth, “Fiuctuatlon Phenomens fa Phys," ‘Sammiung, Views, No.8, Beussvic, 1920, NOTES 15 (24) p. 39~On account of the incompressibility ofthe liquid, the "divergence " ofthe liquid flow ‘must be, on the whole, equal to zero, Le By ye diva = ep AFBECmO. (25) ps 44-—Since ey ‘we obtain in (5) sl at a Be ce 4 B04 Maes oot cey ay G) ; Wasrseo SE gee 4 Bye 6 — fae anscen+ Sa} Since A + B+ C = 0, two of these terms vanish, tnd the remainder, put in (3), gives ag—Spud, — Spek age 4 Bet ig — Spread — Serb aes + Bit + C09) + 2Pinte ay 6x9 + Sees —Prae from which equation (6) follows immediately.116 THEORY OF BROWNIAN MOVEMENT (26) p. 48-—A remark is omitted here which refers {0 the amount of energy consumed, since this no longer involved after correction of an error of calculation (vide Note (2) (@7) p. 53—Fellowing from Gauss’ aw Saiv wdo = fuds: singe lies in the direction of the a-axis, and in sumctotal is equal to , then om oF ave and cos (9) = (28) p, 62—The values given here for the radius of the sugar molecule and Loschmidt's rmumber agree remarkably well with determina- tions of these quantities made in other ways ‘The most accurate value for Loschmidt’ num- ber at the moment is 6-06 1.08, determined from the "Faraday" of electrolysis and Millikan’s value for the elementary quantum. The values, from the Brownian movement, given in the pre~ ceding notes, agree, therefore, remarkably well ; as well a that derived from the radiation of heat according to Planck's equation, 64.10". Further, from the Einstein'Smoluchowski* theory of A. Elateln, Ann. & Pye, 88,1294, 1920, #16, Smalichowall, Ann d Phys By 205, 1908, NOTES 7 density-luctuations in gases and Tiguid| mixtres, ‘a value can be obtained for Loschmidt’s number from measurements of critical opalescence,” in ‘gases in the nelghbourhood of the critical tem- perature, in liquids in the neighbourhood of the critical miscibility point. The former observa- tions were carried ost by Kamerlingh Onnes and Keesom,* and gave approximately 75.108, the later by R, Furth! gave 77.20%, and by F. Zernike* with more accurate equipment, 62 to 65.10", According to this theory, Loschmidt’s rnumber can also be determined from the extine- tion-coeficiont of air for sunlight, by which method Dember * obtained 64.10%, We see, therefore, that a very large number of completely independent methods exist which all lead to approximately agreeing values for this important constant, With regard to the size of the sugar molecule, ‘there is available for comparison the diameter of the fist electron ring of hydrogen, derived from Boh's theory of the hydrogen spectrum, about o'5 10“ em., whilst the sugar molecule would be about too times as large a6 this, the smallest of the atoms. According to the Kinetic theory of AW.A, Kees, Amd. Phy, 98 397, 290% R.Ph, Wien Ber, 124 (2, $7 1903 4, Zorn, Dissertation, Amntera, 1915 +H. Demat, Amd Phy, 8, 390, 1985118 THEORY OF BROWNIAN MOVEMENT gases, the diameters of gas molecules are of the ‘order of magnitude 10°* em, (29) p. 63—Th. Svodberg, “On the Spon- tancous Movements of Particles in Colloidal Solu- tions” Fist paper, Zell. f. Eletirchom., 12, 853-800, 1906. Second paper, Zeit. Elekirochem., 22, 900910, 1906 (69) p. 65—From ae tt — GokPo it fltows ehat m a ~ bake ~# and by integration — gliplos » + const ‘will equal 2 = gitplo from which fallow, for» = yo Pee SOLE” agg oP (9) p. Se, th rene Secon x of vote on pa. (2) p raCompate, eg, the ston, "Os note Tacry” in W. Neva," Thtetal Ekemisey” tg) ander Hochenty NOTES 19 (63) p. 75—Compare Section x ofthis volume, paragraph 3, pr. (64) p-76-—_With reference tothe Note on p75, the uncetainty inthe value of N indicated ere corresponds with the contemporary state of pro- ‘resin the enquiry. To-day this uncertainty ean beput at scarcely more than 2 percent. Vide the sugreement in the diffrent methods forthe deter mination of N, Note (28), p. x26.SUBJECT INDEX sro eminem omarion eco hea SR esac ct are ela stan rp 2,28 [= ESO sata ose perm 3 mveroomne eons Bietag a ea, a a aa Same” pomemae nce Telly, 94 8, e ‘THEORY OF BROWNIAN MOVEMENT eels uy ft cathe (oewrmanoe wanes, [iat Maieesnetnserions, 2,3, 60, |Spensioe of patilen, 37, nt 4 [ttl 4 Se eT aie = Shaye a Bh 6 thesis a Beta ene oso a as et 6808 None, 35 93h ea esse, ‘ve aa Lcd ‘Ramee 9.5% Yan of sion ven cave, 3% 4,455 eamen at yr0h ne, |= pea SE ee | Yetical di att Sova pat Soul iegral Spc eats” 5p, 50] — sua atti oy, | Seana oq | Volume a deed wane, ‘Stee ita seAUTHOR INDEX Bold aes ene 7 EARN E as FEES son evs A CATALOG OF SELECTED DOVER BOOKS IN SCIENCE AND MATHEMATICSA CATALOG OF SELECTED DOVER BOOKS IN SCIENCE AND MATHEMATICS Ser Ht ot lw {TWE EXTRATERRESTRIAL LIFE DEBATE, 150-10, Mica! |. Cre. Bet ‘k,l say In Enh the ay en tat deep ror 3 300 ng he hens flier rogaine of ATIISTORY OF ASTRONOMY, A Panne Wall bund cy ened ‘Aud coves sch pleating sok of Copsey Kes TaD wk mah oc Haas Ree ip {uATIUR ASTRONOMERS HANDBOOK B.S Tas ete carrera ores iti cope de, mer Somperapn more sins pen) SSIARS AND RELATIVITY, Ye. B. 2aforch an 1D. Nowihw. WL of ie py an at tin ene ay epee Chemistry {BALE ACHE Lena ak, ed on Rev hy Nin SS rene eran "HE DEVELOPMENT OF MODERN CHEMISTRY, Aaron J he Aux coer aor Chemis and te icy 20 al Sete nace Appel 0x ‘ceptor dor cover of ncaa for coy ea on a ‘no le poap ton ed ten me Bhp carazoG oF DovER BooKS SE HISTORIC, BACKGROUND OF CHEMISTY, Hy Mee ‘Blan fey ot nda Bogpty. Coenen ormlan fe s faced chee we pp. 8 ‘ios 4 glow HISTORY OF CHEMISIY | 8 een. Ch pation {isthe of vlenys nese am tenys ach move 2 Beate en) cont ENEHAL CHEMISTRY, Lins Paling Revd Se ion of le Sintyear ‘ery Nella A celia scan, uate ‘i cance hemody coed wh eign conse. Pbleme app a coy Engineering DERE METALLICA, Geog Apis Tefen Homer semitone nines 580° ALD orghal wos Cpr Oe ee FUNDAMENTALS OF ASTRODYNAMICS Roger tl: Man apy Seca UA Rac dey aged > soe Pree ‘Sine Nemes arate yp 9 ‘aera DYNAMICS OF FLUIDS 1N POROUS MEDIA Jac Bas Fo ave Sl gonna man py rs os ANALYTICAL MECHANICS OF GEARS, Ea Buckingham. adipose sre nr go mcr ne tt sah me om MECHANICS. F Den Haro. ce oad ato ar. Honda pplenine Bed deg pers tn saan of te, ened ees cabo st 08 ater proton “Opp. cet sa ah eo ar Som AppoiicIndoe ens agpp a3 80 Srmucrn OF MEA. Fa dea oof a ‘trial fenton, tron, beni, se) panna mtr engeing ‘eho spain 8h sweeper S298 aries ‘A MISTORY OF MECHANICS, Res Dug, amen ny of mei Uhm Rep ang any oer py Soe SDcaTaoG oF DovER BOOKS {en and walle volume adder al specs of he ate rem he Nar ‘aldvetpnen fund met ag concep fhe ech sree ‘thccuner chap lcaer appeninesS44pp a0 STATISTICAL MECHANICS: Mince apd Appltinn, Tere) Lil ‘nd tt cnc anenlsof san mec spins oon ‘heoy, impure gae dbaton cons more app 4x80 S500 ‘THE, VARIATIONAL PRINCIPLES OF MECHANICS, Coes Lance Gaon level cnenge fcr eins agen fmt, an ‘peta, so Fin expensive ppt ion cae wei na ‘Bisley pp 88 ‘san ‘A Chai Stn Centy Tht Teste Thang, Aging Rael Setting acy. sled ptr of is Pers sre or Se ee an ee es Spend Sis ilo [ROTARY WING AERODYNAMICS, W.Z Supe. Chas, con txt cov ‘Sr scredj paren of eo ley aine cope peo: ‘sone ction Ong prepared be NASA 56 ue lOpp. Ok IRRODUCHON 10 SACE DYNAMICS, Wen ak Tema, Cn {etn snc actin oe gingiva oped Sond Berta 3Syp 58 ‘stb HISTORY OF STRENGTH OF MATERIALS, SepenF.Tmobecko, Excl el may af be eng of mater wt many lees he eal ‘Sy and wrcuze 25 Bucs Lp. ‘alee [ANAIYTICAL FRACTURE MECHANICS, David. Ung. Seon text ‘Speman scarface echt ona oye mec ‘ingest ee Sp 8800 Mathematics HANDBOOK OF MATHEMATICAL FUNCTIONS WITH FORMULAS, GRAPHS AND MATHEMATICAL TABLES sdted bp Milton Abort and Jr Sia cer 9 mf voc sm hgh pn Dtegp. A caraz0G OF DOVER KOOKS FUNCTIONAL ANAIYSIS Gocond Coc Ft, Gor Bachman and Lvrence Nar cele esto ec eed vad ens wh ac oe ca agi ound as ys ogni Tt ne ‘Sonoma arte cpg ‘ney tbe led mjc 15 08 Slap “ais? aaa geht ta rete as oe es Hee oy Si Shay dean te eer {Soon sy eng te oe en pol ‘VECTOR AND TENSOR ANALYSIS WITH APPLICATIONS, A. Bono {8 Trp. Coheitoacton Wireou res, si ene “THE ABSOLUTE DIFFERENTIAL CALCULUS (CALCULUS OF TENSORS, Tele Lev Ci Gre 20-etry ated werk on anal ne (Suny fran gmp try ay pp SOKA GOT {AN INTRODUCTION 10 ORDINARY DIFFERENTIAL EQUATIONS, Ea ‘A cosngion A tough sod sytem nuns iemeny eet ‘eto fv snd matont nd sn, any errand ‘ates (anne Index Soap. ys Cogn Rtgs Sed per af he Toute metodo andro ‘ri Sm exremeAnter and ns 6p. SU an CCoMPUTABILETY AND UNSOLVABILITY, Matin Davis. Canc te Soren ny cmp aly fee tiny ce Aspromgytrions ny AMEN din Act coy ‘lpng eked examples in det En 2p, SU ‘sas SSA ON THE THEORY OF NUMAERS id Dein Too ae {tanting umber snd popes of atl sumer 1Spp 38x84. 2000Scaravoc oF vovER BOOKS APPLIED COMPLEX VARIABLES, Joo W Dtan Step by ep coverage of feameval of anc neon Gory ned expat fe pan ‘penton Ft! Ther: Ordinary Dike atone Fore om Ep Fao ype xp es Hci opr INTRODUCTION TO LINEAR ALGEBRA. AND DIFFERENTIAL EQUA, "TONS, oto Denn, Enon ent cover comple mb, derma, thon be, Lape, wars, sc er faeces son ‘aden ec lp BE cn [MATHEMATICAL METHODS IN PHYSICS AND ENGINEERING, Joho W. frayed tah Ale peeled neon onic anc Beary: are, anh 4p tt ‘6005 {AICULLS OF VARIATIONS WITH AFPUCATIONS, Go Ming ‘note andentanding of peed bola resem papers Sera for ated Selita opty, rey wp oe alg ar Fanless int So syd el Ste ew noma typ xB 3887 AN INTRODUCTION TO THE CALCULUS OF VARIATIONS, Cae Fox {Gado tt coven verte fo neg pene poems, et gmtsrRoni TORY FOR SCHNTITS AND ENGINEERS, Ra ‘foo near, RT cer mena Alo ingot spews Reems ki asada SA a. peepee een crete balogy, economics, Many ilusative examples; vet 280 problems, 2609p, aah ‘Hamming. Cla ext trees equncy aprons cova of lg, pay ‘mal ofvoxiatn, Foe spot, expoaeta appro, te {pon Ravel wl eneed Sod ton pg ec INTRODUCTION to NUMERICAL ANALYSIS nd Bion #8 te son, numeral ioeataon sept, rtp new pete. Eadpp x = me enon ST CATALOG OF DOVER BOOKS Gapchenive weutent of crthogal fljomias, hypergeme fins, Falfteqaton muck mow Bogs. do Mapp. xR SBD “THREE PEARIS OF NUMBER THEORY, A. ¥ Kinin. Tow comping ‘ony new rg eon cer Co oder Wardens theorem, be anda Sehr ype and an ‘Sete, and naomi Wes olen Saucon nde Sep heb CUCULUSREMESHERTOR EECA. OWI A ee Cee A Of en oo merino ge aa ‘Stas “THE PHILOSOPHY OF MATHEMATICS: An Indotiny Bay, San ‘ieper Survey thei of Pe, Ato Leis Kat oes re fd cso tel nd pre mate Inco wo open Inde Typ 8 ‘ped VODUCTONY REA ANALYSIS AN Klann ain. ‘lal aye Some 0 pre. py “amc APPLIED ANALYSIS, Coa Lancs. Clic wk om ay and sg of inte proven or pprosinning oo of muita preblen Algbre {Gey Ran tts ne th SP AVINTRODUCTION 10 ALGEERATCSTRUCTURES, sp Land. Spe ‘cote seco rac alge an umber ry a rope ‘yop sore Suro wl chon ane xara 9 Sea ‘0d SPECIAL FUNCTIONS XN Lebel by Richard Sivoo ‘Ruan wok eating more impotence hapa pee oben a pace sed engnering bigs dp. 38% 8 aus \UALIATIVE THEORY OF DIFFERENTIAL BQUATIONS VV. Ney SAYER rn ans gudntrlr tty te poset Sai no, Seal dict eqns val opp Sonne st [NUMBER THEORY AND ITS HISTORY, Osi Ov Un cla, sco [Ne iodacon coven atti proper of mumbo, pine Mnber mah roe bloga. Sp 4x ‘sand “TBORY OF MATRICES Sa Mi, Osta in coving eh nosing without the tr veton of dntesnanta indieCATALOG OF DOVER BOOKS INTRODUCTION 70 ANALYSIS, Mecnll Rowe. Unstay cascat cag tty ai en ci ng i, adele Baigent ca ee le eS sopuna Nonunean 2g, Tet Sty: eps pid ee a MATRICES AND LINEAI. ALGEBRA, Haus Snider and Googe Php Site Br tsaoak ceteris nd apa oi ‘treaties and rte oper me ae dormant, eevee! ad ileal ‘putone Nimeroreecce 1idpp 88 ‘sie [MATHEMATICS APPLIED TO CONTINUUM MECHANICS, Le A. Sep ‘Sales ms od ow ol ima. Fr ape evel mat aces ol cagorsing stents py. 8h “ins [BLEMENTS OF REAL ANALYSIS, Det. Specher Cli cove ind ‘etl cone eal me yop estan fl asi, eae ‘Sen mach nore Ove 90 kere tgp Fa 8 ‘sss MUINTRODUCTION TO MATRICES, SETS AND GROUPS FOR SCIENCE STUDENTS, G.Seheaoe Cane, eae txt ness op nd ‘ot party, mater owoderate aot of ye hemi ond ‘cag Pbleme ip. Se ‘home SET THEORY AND LOGIC Rebrt RS Lad ote tuned tay ‘rae omega Se ny ud ng oe cer concep nde Stang ore cum yt Sp 4B ‘san [TENSOR CALCULLIS, JL. Syage ad A. Schl, Wiely wed ited tnt ‘loon pu and teas ba reo Reman pce om Roan ‘pum be Sapp "acu? ORDINARY DIFFERENTIAL, EQUATIONS, Mocs Tinea and Hay fle, Extn ey a ray render anes ‘ubenaicy cnpoeeray, since Tooonge eur of renee’ Dagan ‘Bitlet Hn, Se 7 INTEGRAL EQUATIONS, ©. THcom.Auhorinive, wel tenement uy pl ut he pone ra Eg Eoraer Sdingery pe ae sas [FOURIER SERIES, Gurgaon by Rihad A. iver, Ava tie addon othe Lane 00 ho bj moving ely foe je one nd orem tore 17 problems veep. S88 nbs caTAL0G OF DOVER BOOKS POPULAR LECTURES ON MATHEMATICAL LOGIC, Ha Wang, Not Sard toe Sn ty a ey an {ier scotia ron cry, ore Sapp nip 1 teks 2kpp soa ‘ont {CALCULUS OF VARIATIONS, Rae Wind. a notin ee ‘eprint pron hry fst, mechan, econ Exttsteaghot Sapp, ‘so “THE CONTINUUM: A Cac emton of te Foundation of Ana, Henan Wh Ch of Sorcery anda! rch swith he Co ‘Sl po pon by cnanam Sepp. 94x84 ‘coz igsaascine, MarABNATICAL PROmLEMS WITH ELEMENTARY SOLEHONS. af Yeglom o 1- Sgn, Oner 0 calles pres Feb ay, oan a nea ome Pr Fonna oeap Soon oy Ba Rev ne A SURVEY OF NUMERICAL MATHEMATICS, Did M, Yoong and Kober “ha Gog Bod lectins coverage of empress neal ‘orice rien yr ml eens pen ty etal alsin momo A esa or cas INMRODUCTION 10 FARTIAL DIFFERENTIAL EQUATIONS WITH APHICAIIONS E.ahnangiagd ue We Hoeft ‘re oguion apie common probems tx engnenng nd he phy ‘es Meeme a eve gp 8 “as ‘THE THEORY OF GROUPS Hany. Zaeine alsin one ‘igre rer up Sree cs somos wef ‘tubemsdcs Aone clus of camper, homoorpic Msping Pon? (Ben, any pret trad oe ane tn i ne. DISTRIBUTION TITEORY AND TRANSFORM ANALYSIS: Aa notin {enenlerd Fenn wt Appeatonn AH. Zen Provieb at Ton tery dens eerie oer and Lapse anor: Nemes rene 1p 8 me Math-Decision Theory, Statistics, Probability [RLEMENTARY DECISION THEORY, Heme Chernoff and, Lint Moser Cl adnton waits ad ttn doy eves dln pees iio te, egypt, ch eB Siete ‘caelcAaTALOG OF DOVER BOOKS STATISTICS MANUAL, vin L Clow el Compechese, practi allen fsa and modern mets prepare by US Neel Oran Tet Sto Stes one nace se sane 2p 54 8 va So Tigo rN, a ae iy Sr fener, un et i rt acer oscars aur STATISTICAL ADJUSTMENT OF DATA, W. Baas Deng aod w eco of sen, crve ing, out sar slun ion wa an nig pen Shc wn Ppp. Sh LINEAR PROGRANANG AND ECONOMIC ANALYSIS, Reet Desian, al A Samacbon sod Hater Slo, Ht empha fen rogaming andar wore any Can tay, made were Snes Lease puto mae S259 ws DICTIONARY/OUTLINE OF BASICSTATISTICS JB. eal ad Fen. ‘ian chem doar aw OH ae wns an ano ‘esi mas ving cay, pane md moe \OBABILITY: Aa nt Snel Goldberg Eat bs tt covert ther probate cesar, Era ney och moe So probes MBbogpies, Spy sax He eas GAMES AND DECISIONS: troduction and Cites! Suey, R Da Lace nd owed Rail Spe sates! icon tp thy ay Sedo scl wine Uy they arom gem pren gen dec ‘ling mach mare Bogapy. Sp. 07 FTY CHALLENGING PROBLEMS IN PRORABILITY WITH SOLUTION Fee Mans Rome posers gen iy seated {nnd apc probly, Deed oan Rape BESSY PROBABILITY THEORY: A Cince Cone . A. Ross gy euble, Stconaiellaroduon coe cabintin if wenn, depeetet res ‘oui eas ce itpp a cas SIATSTICAL METHOD FROM THE VIEWHOINT OF QUALITY CON. TROL, Water Sheva para te oxplana eu of varie we etal coal to swe quay cnt ody, gsc, ae ae iiigp ‘si? THE COMPLEAT STRATEGYST: Being» Pane oth They of Ges of Sin 1. Wiss High entangle dest hay Sd saul ntti snes not tt eer pads (CATALOG OF DOVER BOOKS ‘Math-Geometry and Topology lve apc tptgy fom scr ont it ong: Row ae ‘Coplay aswel nt ad pce 25 te pp COMBINATORIAL. TOPOLOGY, PS, Alesana wen, elon Soe nt yy aig wih cen ce ny ot ag {he fooal weighing hy ad dace earl caer he sd peupe Nesom die ample otpp. 8 ‘Sino [EXPERIMENTS IN TOPOLOGY, Sep Bart Ci ely elation fone ‘te bya mathematics Ha bey Macs a sje pl np ‘Sling rte ote Keri inser dese ae cy ad Sitges opp aH ea CONFORMAL MAPPING ON RIEMANN SURFACES, Haney Cobo. Lac Sol tos pon al oer jt cra oa pt “THE GEOMETRY OF RENE DESCARTES, fee Decca Th get wk ‘tnd i ge, ign re ent ng fe "THE THIRTEEN BOOKS OF EUCLID' ELEMENTS raza wth due ou and commen ty Se Tha. Hoa Die eon, tal ad i ‘utc ot maleate 200 yer fie comment. asp Te SEN Tere 0082 Vo: E00800_ YLT: 0. OPOMETRY OF COMPLEX NUMBERS, tne Schwere. yng, Sey ps boo on asa sme of ec, he Nets tasomai Suttaimendonl non achdot rometen op 88 “tod DIFFERENTIAL GEOMETRY, Heirs W. Ouse, Leal ders [pornetysranopeaten oud as dar lea Carve ane Bem pups hens Ener 6 ares app AT CURVATURE AND HOMOLOGY. Fatuged Eaten, Samus 1. Cadory Sends cn in isle al cara Romy oC Rismanln amide mga ae goypn cmplee nfl carat, Soni fui ans Re i ew ped stp. se TOPOLOGY, join G Hosking and GS. Yoong. Super one er cour in ca ‘al toplos pagel pace cnet gy nh ae ‘Samples ol eben Boppy. nen Sig. 48 8 ‘saneCATALOG OF DOVER BOOKS LECTURES ON CLASSICAL DIFFERENTIAL GEOMETRY, Second Baton, Dik Sta sclon tr roto cover cave tery lute, nde Inenui squcons geomet on 9 wre cian! mapping ate pi Frio opps ‘sic Math-History of A SlIOKT ACCOUNT OP THE HISTORY OF MATHEMATICS W Wane Su Grre/ Son abun ys tom gprs ct theca ihc gee ch Gn, Cla, Ree ou co Sense 00 ‘i HISTORY OFTHE CALCULI AND Ts CONCITUAL DEYELOD. Ili eeu rndaten Hout 6 vata Bp SBR GOL {HE ISTORICAL ROOTS OF ELEMENTARY MATIIEMATICS, Lc NH ‘Ban ily 3 ones JD, Bat. Fen aenal wderoning modern tenho gdb, gomty apd mtn yen Grief ale ili oon app ova ‘aoe ‘A STORY OF MATHEMATICAL NOTATIONS, on Cs cn ion teacunted spend mong wins ren! conse io op {ints vet el or inate are Orga) to rune ea ‘rooted hve vols opp. 9 ost ‘GAMES, CODS & GAMBLING: A Hy o Pty a Sat a FN Der Ede om th Bs of Gall easter ats ie fect mca! of he ou ef nba eae sgt povalng le thee conc cel, gap ery. 192 eéson. S04. FE» atic 3 oye “o89 MEN AND NUMBERS: Th ey te rat Matera Ma Rectoning sgn of tenes ad scm of y's pele a ‘mel migra Desc ae Pnea Cate, ty toe Aneta, nminstng 50 ug Balipby 25bpp. 8 om FISTORY OF MATHEMATICS, Dail Suh Noschie survey fom {hc Grose and reso ht th cay eveaion of ates prey, SRE cane Somes Het ee foun SO [ACONCISE HISTORY OF MATHEMATICS, Dik Stk The bt bri hi yeh nhematen Sener orgie st save ery me are om me ‘Nek Eantoiscesy i one SpA ‘iss CATALOG OF DOVER BOOKS Physics OPTICAL RESONANCE AND TWOLEVE!. ATOMS 1. Allan. Hei ‘esonance phenomena. 5) rato. raced 3p. SBA BN ERIE UITRASONIC ABSORPTION: An lntodocon to the Theory of Sead “Abioion and Diperion Gates, id nd SA BBs, Sadar ‘fegrin the eld rove cla yt egeing ney ee of ‘toes concep adnan ste ar tars Nome tag Bop. pp. 0 pend (QUANTUM THEORY, Dai Bot. Thi avaned undergrad eve nt pe ‘Ser th gut eye a aire sd pve emp ele ‘pec lcs worked oot © mth Sa Pre Te Sp, [ATOMIC PHYSICS (8 etn, Mx Sor. Nobel ara’ ld wenn of ‘nec fg ene pry met ae puny sl ‘utr snd seal ne ach oe Ove 40 pend, aOphY an = eer ANINTRODUCTION TO HAMILTONIAN OFS... Bhd Pele tah dalled tc, Wado cise REA GENET ‘Geng Garon ac cml dao otal hay fey it Sac wen fe itp Tes el 1 Oy [eS i pee esses ng Sipe ci Fro fe Clee ote on nat fl ae ‘amicon mba compc Pens TRG den, S986 7 Se BYDRODYNAMIC AND HYDROMAGNETIC STABILITY, 8, Chandra: Let eenminson othe Rygh Sonn prod ea coverg of oe ta “Alt Ein, ve pp (108-8 vegaing dana of rowan neon {nt cavng mesa try. Noerby Fars py THE PHYSICS OF WAVES, Wan C Eire and Mask A Heald. Ungne ree faa wave thor Asus, op slosomnget aon mae adr crm eno lr selistdy ene pp HHL elcarazoG oF DOVER BOOKS [PHEYSICAL PRINCIPLES OF THE QUANTUM THEORY, Werner Heber [etl Lawes canes quasi ivy, sna, wae mechan, wok Sf Bins Schein, Capon, Wien, Exige x8 GN? OMG TEETA x aro SIRUCTURE, Gd Hey Oe ‘ee intctonsy pea or spect een pyc ‘aero meta aon 257 8 ‘oss [AN INTRODUCTION TO STATISTICAL THERMODYNANACS, er L Buen nc tn fo whenever untam satstalmes {hytong czas gun sy ne THEORETICAL PHYSICS, Gory. foos, wh ta M. Feewan Clase ote ‘iver smelt mics donne tan moda ga Sogdian ce pyc wpe fu pack a Hi PHYSICS, Chars. ‘ed ion In tverge of qunta mdm, we ec ‘oat moc esp sue 380 poems poe 59 pena Secon, typ ok sak ‘ch Mosunonl reece wot snes met ‘hey af un jen ep a ‘Sepa i aan See al Rs fata age tac’ eee TREATISE ON ELECTRICITY AND MAGNETISM Janes Gk Moe. "Zecca ad spony Sees irl ans offs ay. ‘hsipp kx Twove st SOT cobs Wor Tt os ‘OPTICKS, Soe Newton, Newtown experinent with speconopy cl lee reflection ecto, ong be yan ct fam Ford ‘ite Eis Sopp 08 ‘ns [HIEORY OF BLECTROMAGNETIC WAVE PROPAGATION, Ce eh True Gt veld) dca he Mane ed quan ution on ‘stemtnana be Doppler sod mare 24 Seb GS IYMRODUCTION TO, QUANTUM MECHANICS, Ya apis Fes oiler spice fem ens chemi spss ples ier sad ur ns be ext Chap bap, Append Inde pp St er) caraog oF DovER BooKS MetHODS OF THERMODYNAMICS, Howard ui. Oxeang ton ‘dite aw tarpon Pel 8 elton app stn ENSOR ANALYSIS HOR PHYSICIST JA Shan Cac xin of pv the try Ene ide Behoppay app 9888 aE [RELATIVITY 1N ILLUSTRATIONS, jas St ler orth a Peake ety more stn an er ttre: Ove dng a faeces more ceaty ‘han xt oe Only gh ee fey eed ore pp asx ‘THE ELECTROMAGNETIC FELD, Aer: Showin, Compenive ir [eae text cove ne of et nd mgt ele ao oem ‘icy: Aste ip cling cy ve ron eg GREAT RXPERIAENTS IN PHYSICS: Fond cms ite ‘eon, hadi dy ofthe eon, Henson eesromagontc waves, mre ‘tig seca del sonal pp 24> ste SUATIVITY, THBRMODYNAINCS AND COSMOLOGY, Rida C peso ef mas mache tea ono me LUoHT SCATTERING BY SMALL PARTICL2S, HC. vn de Hae, oye ‘Scotian tony «Bars Hp SHRYISICAL SIC, Gy Wn Cai cis my ‘ol pas Problems with slaions Boh SS2pp SMB SHOU. ‘ordoverpsbcaonsom, oy wg Dp Dre Ries ic Ea Sent Tc tn ge pc ‘ei ts ts op reineDOVER BOOKS ON PHYSICS “Tapers Nz Pr, ohn M a & Vicor F Welshop 668274) Lees cv Ge Ty, Ladi Boltzmann. (BSS) Gonna Misuse, Chace 4010-4) ioe Meson Pare so ces 8 Cles. (402200) ‘hvu Orns Orca Der AE Conrady (ST, GTS) 20 5. {eR W-Dtenbar, 66675) encucton rox Mas Ores, A Gerad. M- Buh (OA) Promnwe m Quoemae Mess, CL Gofdman and V.D.Reechenkov. (erat) Fug Mica, Rober A Granger (656-7) See Dean AC (69118) Cea, Prnsrs oa Sb Fs, .. Heavens. 66328) {ne Then Quon Seca, Volker Hai (S858). Conc or Font, Max are (406558) tom age Sears Chries Johnson, Je & Don A. Gabriel (éssze) Lace ov Ruts ToL. Landau & Ya. Soro (751340) ‘Tamoanustc sto Sten. Mei, Peter Landsbery, (6450.7) ‘ote Ess Laos NH. Mare & MCP. Tos (35984) ‘Tee oy Pros Queers MECN NH. March WH Youn “as sanpanthar (81540) ‘asta Memes HC Corben & Pal Stele, 680630) ‘Sroms Ores A A Micbeaon (37067) To Toor Hoe Rasa tx anc (581) “Trex 08 Taam, Max Planck (671%) Shxsou Psson Srtemsor Se no Ena, Robert D Reed TERR Roy (OEE) ‘rnc Tie SDE OF Von, Naso Ronchl (E80) rT An Meus, ME Rose (6800 terocucno 7 Sts Rare Jes Sth (8895) ume oe Maen. Pr, AN Tikhonow & A.A Samak (Ges) pod nn ese vai a nk dns ‘lie st wwredoverpablation om, oF by wets to Dep 25, Dovel ‘ito fe Tess nd Sect Mino, VT For urea [ie norton or frre catalog (pense cate el of teres), {ire to Dower Puetons or fg on to wedoverpableaons.com lin ace every Dover bok np Ech year Dower pte over #00 ‘ron fin art muse erate an eer, nques, ngs sr Bk chery, aterm Mancacted nthe USA