Está en la página 1de 16

About the intuitive content of quantum theoretical Kinematics and mechanics. From $ V. Heisenberg in Copenhagen. With 2 illustrations.

(Received 23 4 [arch, 1927.) In the present work, first precise definitions of the words are: location, speed , energy, etc. set up (eg the electron), which remain valid also in quantum mech anics, and it is shown that canonically conjugate G1Oen simultaneously with a cha racteristic Accuracy can be determined (1 g). This inaccuracy is iler real reaso n for the occurrence of statistical correlations in quantum mechanics. Their mat hematical Eorinulierung achieved by means of the Dirac-Jordan's theory (g 2). Th us obtained by the principles starting shows how transactions from the macroscop ic Quantenmechauik out understood knnen (f 3). 7UR explanation of the theory are discussed in some special Gedankenexperirnente (S 4). A physical theory we believe to be clearly understood if we consider the experim ental implications of this theory are qualitatively in all simple cases, and if we have recognized same time that the application of the theory never contains i nternal contradictions. For example, we believe the Einstcinsche idea of the close d three-dimensional space graphically understand, because the experimental conse quences of this idea are consistent conceivable for us. Of course, these conclus ions contradict our usual ideological space "time terms. But we can convince the m that the possibility of applying this familiar space "time terms can be inferr ed on very large spaces neither laws nor our thinking from the experience. The p hilosophical interpretation of quantum mechanics is yet full of internal contrad ictions, the conflict of opinions in order discontinuum and continuum theory. Co rpuscles and waves affect. This alone would conclude that an interpretation of q uantum mechanics with the usual kinematic and mechanical terms is not possible i n any case. Quantum mechanics was precisely the result of an attempt to break wi th those usual kinematic concepts and to replace them with concrete relationship s between experimentally given numbers. Danish seems to have accomplished this, the mathematical scheme of quantum mechanics, on the other hand, require no revi sion. Neither a revision of the room "time geometry will be necessary for small spaces and times, as we through the choice sufficiently heavy masses the quantum mechanical laws can approximate any classic, even if it is still so small spaces and times. But the fact that a revision of the kinematic and mechanical terms is necessary seem s to follow from the basic equations of quantum mechanics directly. If a certain mass m is given, it has an easily understood meaning in our usual intuition, th e location and the velocity of the center of mass n>. to speak. In the Quantenh but mechanics is a relation pq "= qp. between mass, location 2 xi and speed are made. So Xvir have good reason to draw against the uncritical appl ication of those words "place" and, speed "suspicion. While admitting that disco ntinuities are somehow typical of processes in very small spaces and times, such a failure is just the terms, Location "and" speed "even immediately

g 1 The terms: place, train, speed, power. In order to follow the quantum mechan ical behavior of any object, one must know the mass of this object and the inter action forces with some fields and other objects. Only then is the Hamiltonian o f the quantum system can be set up. [The following considerations should in gene ral relate to the non-relativistic quantum mechanics, since the laws of quantum electrodynamics theory are still very incompletely known] '). In the 'shape' of the subject any further statement is unnecessary, most appropriately refers to t he totality ienna interaction forces with the% character locations. If you want to be clear about what is meant by the words "place of the object",

such as the electron is (relative to a given reference system) to understand, on e must specify certain experiments that allow you to, position of the electron " intends to measure, otherwise this word has no meaning. In such experiments, in principle, the "place of the electron" even allow to determine as precisely as d esired, is not a defect, such as:. Illuminating the electron Man and examine it under a microscope, the highest achievable accuracy of the location determinatio n is here essentially given by the wavelength of the light But it will basically build around a I'-beam microscope and with this the location is determined as a ccurately perform knnezb as you want, it is, however, this provision is a seconda ry factor significantly.. Compton effect. Each observation of the coming of the electron scattered light is a photoelectric effect (in the eye, on the photograp hic plate, the photocell) ahead, so it can also be interpreted to mean that a ph oton hits the electron, reflected on this or is flexed and then through the of t he micro-lenses work on the basic postulates of quantum theory Iz. B. CA. f Phys. 18, 117, 1923) and Einstein's discussion of the relationship between wave field and light quan ta. Most clearly the problems are discussed here in recent times and discussed t he issues arising been partially answered by W. Pauli (quantum theory, Handb d P hys, Vol XXIII, still cited as 1st c..) By quantum mechanics has in the formulat ion of these problems by Pauli little changed. It gives me a great pleasure to t ake this opportunity to thank Mr. W. Pauli for the various suggestions that I ha ve received from common oral and written discussions, and has contributed signif icantly to this work. >) But great progress has been achieved by the work of P. Dirac [Proc recently i n this area. Roy. Soc. (A) 114, 913, 1927 and later appearing llntersuchungen]. About the intuitive content of quantentheoretischeu kinematics etc. 1 r 5 Skops again deflected the photoelectric effect triggers. At the moment of locali zation, ALSN the moment. in which the photon is diffracted by the electron, the electron changes its momentum discontinuous. This change is all the greater, the smaller the wavelength of the light used, ie the more accurate the localization . In the moment in which the position of the electron is known, therefore, be pu lse can be only up to sizes corresponding to those discontinuous change is known , so the closer the location is determined, the less accurate the pulse is known and vice versa. we see this as a direct ideological He1> explanation of the relation pq "qp = " Let r1, the accuracy with xxi is the value of r> r known (ry is about rler average error of ri)> So here is th e wavelength of light, 1>, the accuracy with which the value of @ can be determi ned. So here the discontinuous change of f> the Compton effect, sn stand for ele mentary formulas of t'nmptoneBekts and r1, in the relationship (1) p> 1 ~ lr That this relationship (1) a direct mathematical connection lt with the commutation relation pq "qp = " ". stands were later ge2 7TI be displayed. It should be noted that equation (I) of the Ausdrucl for this is p recisely the facts that you previously sought to be described by dividing the ph ase space into cells of size 1>. To determine the electron can also place other experiments, for example, attempt s to make impact. A detailed AIessung of the place requires collisions with fast particles, since the slow electron diffraction phenomena that are a consequence of the de BroglieXVellen by Einstein (see, eg, Ramsauer effect) prevent an accu rate determination of the place. In a pulse measurement, the precise location of the electron thus changes again discontinuous and easy estimation of the accura

cy of the formulas of the de Broglie nomic V represents the relationship (I). Through this discussion, the term "position of the electron" seems defined clear ly enough and that it was only a NVort about the white bear "of the electron add ed. When two very fast particles in the very short time interval after the other Qt meet the electron. as defined by the two places of the electron particles ve ry close to each other at a distance Ql. From the laws that are observed at r> t -rays, we conclude that Ql up to sizes of order 10 cm re can depress if only z1t sufficiently small and the particles are chosen to be su fficiently fast. This is the sense, when we say that the electron is a corpuscle whose radius "cm is not greater than 10. Let us now turn to the term "path of the electron." Bottom web we understand a s et of points in space (in a given reference system), which in turn accepts the e lectron as "Places". Since we already know what is to be understood by "place at a certain time," come here, no new difficulties Nevertheless, it is easy to see that, for example, the often used A.usdruck. "1 S-orbit of the electron has in the hydrogen atom 'from our point of view does not make sense. these 1 S, train "to measure, namely the atom would have to be illuminated with light whose wavel ength is in any case much shorter than 10 " is s cm. But of such light a single l ight quantum is sufficient to the electron completely from his., Train "to throw (and therefore of such a path always only a single point in space can be define d), the word. Train" so asked here no rational sense. This can already be easily inferred from the experimental possibilities without knowledge of the newer the ories. In contrast, the imaginary position measurement can be performed at many atoms i n 1 SCheck the condition. (Atony in a given "stationary" condition can be, for e xample, by the Stern-Gerlach experiment in principle isolate.) It has for a spec ific condition, for example, one of the S atom to give a Vvahrscheinlichkeitsfun ktion for the locations of the electron, the the average of the classical path i s represented at all stages and is detectable by any exact measurements. Accordi ng Born ') this function is through tri, s (q) tri, s, where (g) when tri, s (if ) the condition for means S 1 corresponding Schrdinger wave function. using Dirac ') and Jordan') I would like with regard to possible future') The statistical significance of the de Broglie waves was first formulated by A. Einstein (d Sitzungsber. preuil. Academic of Sciences., 1925, p.6). This stat istical element in quantum mechanics then plays an essential role in M.. Born, W . Heisenberg nberg and P. Jordan, Quantum Mechanics II (Zs Phys. Way, 557, 1926) , particularly section 4, 6 g, and P. Jordan (Zs Phys. 87, 876 , 1926), it is a seminal work of 5f. Rome (Zs Phys. 88, BOB, 1926) mathematically analyzed and us ed to interpret the Stoi) phenomena. The grounds of probability approach from th e theory of transformation matrices can be found in the works by W. Heisenberg ( Zs Phys 40.501, 1926.), P. Jord.an (ibid. 40, 661, 1926), Paul W. i (note . to Z S. Phys. 41, Bl, 1927), p you ac (Proc. Roy. Soc. (A) 118, 621, 1926), P. Jordan (Zs Phys. 40, B09, 1926). In general, the statistical side of quantum mechanics is discussed in P. Jordan (Naturwiss. lo, 105, 1927) and Af. Born (Naturwiss. l o, 23B, 1927). generalizations say that the probability is given by S (1 S, q) S (1 S, q), wher e S (1 S, q) is that column of the transformation matrix S (E, g) from E to q to e = e, s belongs to (E = energy). In that the quantum theory in a particular state, such as 1 S, only the probability function of the electron location can be specified, on e may behold with Born and Jordan train a statistical characteristic of quantum theory, in contrast to the classical theory. But you can if you want, say with D irac that the statistics was brought in by our experiments. Because apparently o nly the probability of a particular locality electrons could be indicated in the classical theory, as long as we do not know the phases of the atom. The Ilnters chied between classical and quantum mechanics is rather: Classic we can by preli minary experiments always think determines the phase. In reality, this is imposs

ible, because each experiment to determine the phase destroyed or changed the at om. In a given stationary "state" of the atom, the phases are in principle indet erminate, what is known as a direct explanation of the equations can view. (J = effect variable, w = angular variable.) The word speed "of an object can be easily defined by 31essungen if it is forcefree motion. Klan, for example, the G-URPOSE light with red light and determined by the Dottlereffekt of the scattered light, the velocity of the particle. Dete rmining the rate becomes more accurate, the more boring the light used, because then the change in velocity of the particle dry quantum of light is the lower of Comttoneffekt. Position determination is accordingly uncertain, as it correspon ds to equation (1). If the velocity of the electron in atom is to be measured at a given moment, one is about at this moment the nuclear charge and the forces c an suddenly disappear from the other electrons, so that the movement takes from then force-free, and then perform the above mentioned provision. Again, If, as a bove, easily persuaded that a function p (t) for a given state of an atom, for e xample, S 1, can not be defined. contrast, there are again a Wabrscheinlichkeitsfunktion of p in this state, which has to Dirac and J ordan, the value S (1 S, p) S (1 S, y). S (1 S, y) is again the one column of th e transformation matrix S (E, y) from E to p, which belongs to E = E, s. Finally it should be pointed out the experiments, which allow to measure the ene rgy or the values of the action variable J, and such experiments are important bec ause we can only define with their help, what do we mean when we and the diskont inuierlirhen change in energy J speak. The Franck-Hertz puncture tests allow to measure the energy of the atoms due to the validity of the energy theorem in qua ntum theory due to the energy measurement linearly moving electrons. These measu rements let himself carry any right in principle, if only on the simultaneous de termination of the electron resort, ie the phase waived / S (see above, the determination of y), the relation Et "ti = - " Duck Egg speak ing. The Stern - Gerlach experiment allows the determination of the average elec tric or magnetic moments of the atom, ie the AIessung of quantities that depend only on the action variable J. The phases remain undetermined principle. Just as it makes sense to speak of the frequency of a light wave at a given moment, we can speak of the energy of an atom at a given moment. This corresponds to the St ern Gerlarhversuch the fact that the accuracy of the energy measurement is lower by sn, the shorter the time period in which the atoms are under the influence o f the deflecting force '). An upper limit for the deflecting force is in fact gi ven by the fact that the potential energy that may varying between two successiv e deflecting force amounts only to within the beam, which considerably smaller. than the energy differences between the stationary states, if a determination of the energy of the stationary states should be possible. Thus, let X be a power amount that satisfies the condition (E are at the same time the accuracy of that energy measurement), then E / d, the maximum value of the deflecting force when rl the width of the beam (measured by the length of the used aperture) It i ~ means. The angular deflection of the atomic beam is then where t dy 'denotes the length of time in which the atoms of deflecting under the influe nce of SCehen force the momentum of the atoms y in the beam direction. This deflection must be mindesCens same order of magnitude as the natural by diffraction at the aperture of the beam broadening brought forth, Damic a measurement is possible. The angular deflection by diffraction is roughly L / d when L is the de Broglie wavelength see called, so X, t, la or as (T dp 1:, t, h ~ (- ') This equation corresponds to equation (1), and shows how an accurate energy dete

rmination may be achieved by a corresponding UngenauigkeiC in time. g'2. The Dirac-Jordan theory. The results of the previous section we would like to summarize and generalize this statement: All terms that are used in the class ical theory for the description of a mechanical system can be defined precisely for atomic operations analogous to the classical concepts. The experiments, whic h serve such definition, but is purely according to experience an uncertainty in when we require them to simultaneously BesCinrmung konjugierCen two canonical v ariables. The degree of uncertainty is given by the (extended to any canonical k onjugierCen sizes) relation (1). It is tempting to compare this with the quantum theory of special relativity. According to the theory of relativity can be the word "same" is not defined otherwise than by experiments in which the propagatio n speed of light enters significantly. Would there be any, sharper "definition o f simultaneity, eg signals that propagate infinitely fast so the theory of relat ivity would be impossible. But since there are no such signals, because rather i n the definition of simultaneity already found the speed of light, space is crea ted for the postulate of konstantenliichtgeschwindigkeit, why is this postulate with the appropriate designated use of the WrCer "location, speed, time," not in a contradiction. Similarly with the definition of the terms, Elektronenort, spee d "'in quantum theory. All experiments that we can use to define these words con tain the necessary by equation (1) given inaccuracy if they allow to define each term p, y exactly. If there were experiments simultaneously., Sharper "Determin ation of p and q allowing, when it corresponds to the equation (1), the quantum mechanics would b e impossible. This inaccuracy is defined by equation (1), so only creates space for the validity of the relationships in the quantum mechanical VerCauschungsrel aCionen 1 ' (Q "q ') 2 2 a (q " q') 2 p '( q "q') S (2l, q) prope Qiq ", ie SSprope qi2. (B) Then for the probability amplitude corresponding to y S (ql, y) = J S (ql, q) S (q, f)) dq. (4) For S (q, y) can be set to Jordan 2 'q tip S (q, z) = e () '-' ~ (F> "> ') q, of y for which Then by (4) S (ql, y) only for values Ii, is not substantially greater than l, differ significantly from zero. In particular, (B) applies if: 2 <2 (p "p ') q (q' " q) S (ql, y) prop j e <2qi2 dq that is, (P "p ') 2 2 xi (p " p') 2 + Q '(p "p') S (2I, y) prope 2 2", so SSprope where I2 Iqi q (= " 2m find their distinctive expression, this equation it allows, without the physical meaning of the variables y and q are changed mCe. For those physical phenomena which (luantentheoretische formulation is still unk nown (eg, electrodynamics) means equation (1) a requirement that may be nCzlich t o find the new laws. For quantum mechanics lC equation (1) by a herleiCen slight g eneralization of the Dirac-Jordan's formulation. If we take the acidified q of t

he electron to q 'determined for the bestimmCen value 2I any parameter with an a ccuracy q "so we can use this fact by a Wahrscheinlichl-eitsamplitude S (21, q) to express the q only in an area about the size to q 'is significantly different from zero. particular, one can set eg The A.nnahme (6) S (q, q) corresponds to the experimental fact that the value of y 'for p, the value q' q [with the accuracy limit (6)] was measured. Purely mathematical characteristic of the Dirac Jordan formulation of quantum me chanics, that the relations between p, q, E can be written as equations between very general matrices, etc., such that any given quantum-theoretical value appea rs as a diagonal matrix. The possibility of such notation is clear, if you clear the matrices (eg inertia) suggests itself as tensors in multidimensional spaces , between which there are mathematical relationships. You can move the axes of t he coordinate system in which we express these mathematical relationships, alway s place it in the Hauptacbsen of these tensors. Finally, one can always characte rize the mathematical relationship between two tensors A and B by the transforma tion formula, which convert an oriented to the major axes of the coordinate syst em A to another, which is oriented to the major axes of B. The latter correspond s to the formulation of Schrdinger's theory. Than the actually "invariant", indep endent of all coordinate systems formulation of quantum mechanics, on the other hand, the Dirac notation of q-numbers view., If we want to derive physical resul ts from that mathematical scheme, we need the quantum theory sizes so the matric es (or "assign tensors' in multidimensional space) numbers. This is to be unders tood that in multidimensional space that a particular direction is defined arbit rarily (namely, is set by the type of experiment employed) and is asked to choos e which of the "value" of the matrix (for example, in that picture of the value of the A.ber in this case there is an exact answer to the question even if, mome nt of inertia) is given in this direction This question has an unambiguous meani ng only when the predetermined direction coincides with the direction of the pri ncipal axes of that matrix.. the predetermined direction is little different fro m the one of the main axes of the matrix, so one can speak "value" of the matrix in said predetermined direction from said still to a certain given by the relat ive pitch inaccuracy with a certain probable error. You can say so: each quantum theoretical size or matrix can be assigned a number indicating their "value", w ith a specific probable error, the probable error depends on the Coordinate system from, for each quantum theoretical size there is ever a coordi nate system in which the probable error for this size disappears. A particular e xperiment can give accurate information on all quantum theory sizes so never, bu t in a way characteristic of the experiment, it shares the physical quantities i n the "known" and unknown "(or, more accurately, and less known quantities). A. The results of two Experirvente can only be derived exactly apart when the two e xperiments divide the physical quantities in the same manner. Known 'and' unknow n '(ie w hen the tensors in that repeatedly used to illustrate multi-dimensional space in both experiments be "seen" from the same direction). Cause two differe nt experiments organizations in "known" and "unknown", as can be the connection between the results of those experiments, fitly specifying only statistically. For a more precise discussion of this statistical relationship, a thought experi ment is carried out. A star-G erlach shear atomic beam will first sent through a field E, which is highly inhomogeneous in the beam direction, it considerably a lot (bergnge! Causes by "Schttelwirkung." Then the atomic beam run some time off, in a certain distance from E, but starting a second field F ~, similar inhomoge neous as e. Between E> and E ~ and hinter1 ', it is possible the number of atoms in the different steady states by a magnetic field applied to any measure. The Strablungskrfte the atoms are set to zero., if we know that an atom in a state of energy E 'was before e happened. experimental fact that we can thereby express> that we are the atomic wave functions "such as in the p-space "with the specifi c energy E and the indeterminate phase P" S n f n (((+ la) S (E ', p) = g (E', p) e

assign. After crossing the field E, this function is have turned into) ani ~ e (a + P ~) i) See P. Dirac, Proc. Roy. Soc. (A) 112, 661, 1926, M. Born, ZH. f Phys. 40, 167, 1926. Herein are the P ", somehow arbitrarily fixed, so that c '~ are uniquely determi ned by F. The matrix c "~ transformed values of the energy passing through E to wh ich, after the passage through E. We take you behind E, a determination of the s tationary states, for example, through an inhomogeneous magnetic field, so we wi ll with probability c "," c "~ find that the atom has changed from state m to st ate m. If we find experimentally that the atom is really gone up in the state m, we will have ~ assign it to calculate all that follows is not the function + c "~ S" "but rather the S function with indeterminate phase, the experimental stat ing that "state m," we choose from the plethora of different ways (c '~) certain : m, but destroy the same time, as afterwards explained everything to phase rela tionships even in the sizes c "contain ~ had. While passing the atomic beam by S ~ repeats the same as e. K. Let d "", the coefficients of the transformation ma trix ~ transform the energies E ~ the <to e., if ~ between E and E no provision of the state is made, then the eigenfunction transforms according to the followi ng scheme: S (p ) Zc. S (+ p) Z Zc <LS (<t p) (S) m m It should be + c "d = - e" set. % Ith the stationary state of the atom found beh ind F ~, one is with a probability e "<e '<the state / find. In contrast, when b etween E, and E ~ the statement: "state nz" was made, then the probability of "l " behind e be <given by d ~ <d ~. In several repetitions of the whole experiment ('ll each time determined between E, and E, the state), so you will behind e ~ the state / with the relative frequency Z '< "Q c" ~ c "d ~, d ~ <beobWl eighth. This expression does not match e "<e '<. Jo rd to (1 c.) Has therefore s peak of an "interference of probabilities." But to him I do not want to join. Be cause the two experiments that lead to ~ e "e" <or Z ',' are indeed physically r eally different., in a case suffers the atom between E, and E ~ no fault, in oth ers it is disturbed by the cameras, enable the determination of the steady state . These devices means that the "phase" to the atom to basically uncontrolled bare amounts will change, as well as at a determination of the location of the e lectron momentum changes (see g 1). The ihlagnetfeld to determine the condition between F, and Fs is detune the eigenvalues. in observing the path of the atomic beam are (I think of Wilsonaufnabmen) the atoms statistically different and unk ontrol'ierbar brakes, etc. This has the consequence that the final Transformaons of the energy before entering F, on ch is given to exiting matrix e "~ (the values Fs) not by Q c '~ d ~ ~, but ru each term of the sum has still an unknown phase factor. We '-nnen therefore only be expected that the average value of e' <e "<e ntuellen of all these phase changes is equal to Z ~ ~. A simple calculation give s "ila this is the case. "We can say after some statistical Ri" a Iessen of an ex periment on the possible outcomes of another '>. The other experiment itself cho oses from the plethora of Possible ~ nits a very specific and limited by the pos sibilities for all subsequent experiments . Such an interpretation of the equati on for the transformation matrix S or the S see chr dinger wave equation is only possible because the sum of solutions is a solution again Therein we see the pr ofound meaning of the linearity of Schrdinger's equations,. therefore they can on ly equations for waves to be understood in the phase space and therefore we want every attempt, for example, in the relativistic case to replace these equations (with multiple electrons) by non-linear, consider hopeless. g B. The transition from micro to macro mechanics. By carried out in the previou s paragraphs analyzing the words "Elektronenort", "speed", "energy", etc., the t erms of the quantum theoretical kinematics and mechanics seem to me sufficiently clarified, so that a clear understanding of even the macroscopic processes from the standpoint of quantum mechanics possible must be the transition from micro

to macro mechanics has already been treated by Schrdinger '), but I do not think that the Schrdinger consideration the nature of the problem hits, and for the fol lowing reasons:. According to Schrdinger is in high Anregungszustndeu a sum of nat ural vibrations of a not too large wave packet may result, in turn, under period ic changes of its size, the periodic motions of the classical "electron" r) E. Schrdinger, aturwiss. 14, 664, 1926. executes. Against this end is to enter the following: If the wave packet would h ave such properties, as described here, the radiation emitted from the atom in a Fourier series would be developable, in which the frequencies of the harmonics are integer multiples of a fundamental frequency. The frequencies of the spectra l lines emitted by the atom according to quantum mechanics, however, are never i nteger multiples of a fundamental frequency "except for the special case of the harmonic oscillator. Schrdinger's thought is thus treated by him only for the sim ple harmonic oscillator, in all other cases spreads over time a wave packet over the entire Rat.w in the vicinity of the atom. The higher the excitation state o f the At s, the slower that scattering of the wave packet. But at one waits long enough, it will happen. The above-urte argument about the radiation emitted by n uclear radiation can be z chst apply against any attempts to seek a direct transi tion of <& distinctive in the classical mechanics for high quantum numbers. ian has therefore previously tried that argument to escape by reference to the natur al radiation width of stationary states; certainly wrong, because firstly this w ay is already locked in high states of the hydrogen atom due to the low radiatio n, secondly must the transition of quantum mechanics in the classical be underst andable without bond in the electrodynamics. In this well-known difficulties tha t are a direct connection with the quantum theory in the classical way, Bohr ') has already pointed several times. We have only so again as explained in detail, because they seem of late to be forgotten. I believe that you can pragnant as the emergence of the classical "path" formula tion: the "path" is created only by the fact that we observe.. Be given, for exa mple, an atom in the excited state in 1000, the track dimensions here are alread y relatively large , so that it suffices in terms of g 1, to make the determinat ion of the electron situation of relatively long wavelength light. When the dete rmination of the location should not be too inaccurate, the Compton recoil will result in that the 4.tom after collision in any state between, say, the 950 and 1050 is,. simultaneously the momentum of the electron can be closed with a deter minable from (1) The accuracy of the Doppler effect as given ex. ~) N. Boring, basic postulates of quantum theory, 1 c. ) dH (9 p (9) are closed. The web can but, as already stated, only statistically calculated fr om the initial conditions, which can be regarded as a consequence of the fundame ntal inaccuracy of the initial conditions. The statistical laws are different fo r the quantum mechanics and the classical theory, which can lead to gross macros copic differences between classical and quantum theory under certain conditions. Before I discuss an example of this, I would like to take a simple mechanical S ystein: force-free motion of a lilassenpunktes, show how the above discussed tra nsition to the classical theory experimental fact can be due to a wave packet "better chance Package " in the q-s pace of a given by the wavelength of the light used size, composed essentially o f eigenfunctions between the 950 and 1050. Eigenfunction, and characterized by a corresponding packet in the y space. After some time a new location determinati on with the same accuracy will run. Your result can specify only statistically a fter 2 g, as likely places all within the koinmen now been broadened wave packet

with calculable probability into account. This would be no different in the cla ssical theory, because even in the classical theory the result of the second Ort sbestiinmung because of the uncertainty of the first term were only statisticall y be specified: the system tracks the classical theory would be similar spread a s the wave packet. However, the statistical laws even in quantum mechanics and i n classical theory are different. The second localization chooses from the pleth ora of options from a certain "g" and limited for all of the following provision s the possibilities. According to the second location determination, the results of later measurements can only be calculated by the electron back a "smaller" w ave packet of size L (wavelength of the light used for observation) maps. So any localization reduces the wave packet back to its original size, the "values" of the variables p and q are during all experiments known with a certain accuracy. Fact that the values of p and qinnerhalbdieser G enauigkeitsgrenzen afford the cl assical equations of motion sequence can be directly from the quantum - mechanic al laws About the intuitive content of quantum theoretical kinematics, etc.] 87 is to formulate mathematically. The equations of motion are (at dimensional movement) 1 1 p = 0 (10) Since the time can be treated as a parameter (as Pay-se) when occur not dependent on external forces of time, this is the Solution of these equations: 1 ~ = "s.t + e c "p. (11) m [T lt 6 [M ~ frl Iqp J q, [S (q, q) 2 rtict = QS (q "q) (2: 1) S (q., q) (V ~) d ~ o S (qp, q) = const. e (1 3) SS is therefore independent of qp, ie, when the time t = 0 qp is exactly known, of q equally likely, that is, the probability that q for any time t) 0, all values is within a finite range, is even zero. This is so vividly clear without furthe r notice. For the exact determination of qp leads to an infinitely large Compton recoil. The same v ould of course apply to any mechanical system. 6 But hen q, at time t = 0 with an accuracy qr uild pp with the accuracy yi was known [see Eq uation (3)] (Pc "p ') 2 S (pp, qp) = const. e 2Pt ' Journal of Physics. Itu. XLIII. 2 rri P '(Pc "V') 13 where p and q and location pulse at time t = mean 0th At time t = 0 wertie [see equation (3) to (4)], the value qp = q 'with the accuracy of q, yp = y' measured with the precision p. To close out the cWertens of pp and qp on those values "of q at time t, the one transformation function must be found by Dirac and Jordan, which, in such transformed all matrices, where qp appears as a diagonal matrix, where q is the Diagonal matrix appears. p i, in the% 1atrixschema where qp appea rs as Diagonalmatix by h d the operator .. be replaced. After Dirac [l c. Equation (11)] 22rt 'DQP

then for the desired transformation amplitude S (q, q) of the differential equat ion: so for the YVahrsrheinlichkeitsfunktion q according to the formula S (g, e = f B (q,% j S (V "9) dq, to be calculated. The result is 2 + in [(t) qg "I ot ' " v "1 I,. (, -. U) - "] S (f / ~ q) = const. J th e [(m,) 2] 2 vt' ' Leads to the abbreviation (] 4) t It ~ ~ 7E. ' i'tt qj a, the exponent in (14) (15) qo '(+ ") " q ~ q' + "~ q " "f ~ ~ + q The term with q's, in the constant (independent of q factor) be included and the integration yields [~ + "'(v " "-' p) j ' Q i> t 1 + " S (q, q) = const. e (1 5 (- = "I> ')' (-)) 2 gt2 (1 ~ tt2) = Const. s It follows 2 (Dd " " 't' 'e') S (t1, q) S (q, q) = const. e vi '<' + t '> (17) t The electron is thus at time t at the point "p '- + q' m with an accuracy q, $ 1 + P. The "wave packet" or better, Wtahrscheinlichkeits paket "has increased by a factor of (1 + ps enlarged. P by (15) is proportional to the time t is inversely proportional to the iNasse " this is immediately evide nt "and qi are inversely proportional '. A too great accuracy in q ~ has a large uncertainty in the y ~ result and therefore leads to a large uncertainty in q. g The parameter that we introduced above for technical reasons, could be omitted in all formulas here since it is not included in the bill. As an example, the difference of classical statistical laws w on the quantum the ory may lead to gross macroscopic differences between the results of both theori es, the reflection of an electron current in a lattice is briefly discussed. Whe n the lattice constant of the order of the de B roglie shy wavelength of the electrons, the reflection occurs in certain di screte directions in space, like the reflection of light on a grid. The classica l theory is here roughly once-roskopisch something else. Nevertheless, we can no t determine the trajectory of a single electron to a contradiction to the classi cal theory. We could, if we could steer the electron about a particular point in a grid stroke and then realize that the reflection is there unclassically. If w e want but as accurately determine the location of the electron, we can tell whi ch job a grid stroke it is true, as the electron receives through this localizat ion a high speed, which is de Broglie wavelength of the electron is so much smal ler that now the reflection really can be done in this approximation in the clas sically prescribed direction and is, without contradicting the laws of quantum t heory. g 4 Discussion of some particular thought experiments. After trying this ideological interpretation of quantum theory

to the timing of transitions, the "quantum leaps" 'as be specifically detected by means of measurements, such as the energy in stationary states. The accuracy with which such a time h is determined will be given by equation (2) by) when ZE DI "means the change of energy in the quantum leap. Avir to think about the following experiment: An atom at time t = 0 iru State 2, may pass through radiation to normal l. The atom can be roughly analogous to equation (7) the eigenfunction 2 1ttE2t 2 mtEt l be assigned, if we assume that the radiation damping is ~ 'expresses a factor of the form e in the eigenfunctions (the real addiction is perhaps not so easy). T his atom will be sent to measure its energy through an inhomogeneous magnetic fi eld, as is usual in the Stern-Gerlach experiment, but should follow the atomic b eam path a long piece of the inhomogeneous field. The corresponding acceleration is to measure about the fact that the whole route, which traverses the atomic b eam in the magnetic field, in small ') See W. Pauli, L. c. P 12 Dividing sections at the end of which determines the deflection of the respectiv e beam. Depending on the speed of the atomic beam corresponds to the division in to sections on nuclear division into small time intervals Qt. After $ 1, equatio n (2) corresponds to the interval Qt / T accuracy in the power of "The probability ~. At to measure a certain energy E, so let itn interval directly inferred from S (y, e) and by not to (n + l) At is calculated by: (N + 1) 2 ct t e n t t S (p, E) = S (y, t) e dt not ~ (n +1) times If at time (n +1) Bt the statement: "Condition 2" is made, the atom for all is n o longer later the eigenfunction (3S) to you, but one that can be seen from (18) when t by "replaced (n + 1) Qt If one hand firmly.:" t state 1 ', then the atom from then on the eigenfunction 9mtx <t. g (e "y) e assigned. ivan will be first transformed into a series of intervals observed Qt "state 2", then continuously "state 1". In order to distinguish between the two states is still possible, Qt must not h be depressed below. With this accuracy, ie the date of transition is determined. An experiment of the just described way we think in the spirit of the old by Pl anck, Einstein and Bohr reasoned view of quantum theory, when we speak of the di scontinuous change of energy, since such an experiment is carried out in princip le, have an agreement on its A. usgang be possible W ear in basic postulates of quantum theory, the energy of an atom as the values of the action variables J in front of other pieces of determination (position of the electron, etc.) has also the advantage that its numerical value can always s pecify. This privileged position takes up the energy to other quantum mechanical quantities over, she owes, however, only the fact that it represents an integra l of the equations of motion in closed systems (for Knergiematrix E = const), wh ile not closed systems, however, the energy is in front No other quantenmechauis chen Size excel. In particular, w ill you can specify experiments in which the phases of the atom w are precisely measurable, but for these the energy principle rema

ins undetermined, a Relah tion J [tt "[dd J =. or t, te, ~ h accordingly. Such 2xt Experiment is, for example represents the fluorescence resonance irradiation of an atom with a natural frequency, say p, s "', the atom oscillates in phase with the external radiation, in principle it makes no sense to ask to what state X, or X ~ the atom vibrates. The phase relationship between the atom and external r adiation can be, for example, by the phase relationship with each other to deter mine many atoms (Woods tests). If you want to refrain from experiments with radi ation rather, one can measure the phase relationship also so that one precise lo cation provisions referred to in 1 g of the electron at different times relative to the phase of the incident light makes (many atoms). The single atom is about the "wave function" 2 mi (Et t y tt) S (g, t) = e ~ ttt ~ (X ~, g) e> +) 1 "c ~ tttt (xy tt) e" (19) can be assigned; h erein c ~ depends on the strength and from the phase of the incident light from. the probability of a particular situation therefore is g Tttt S (q, t), S (d, t) = e ~ ttt ttt ~ ~ + (1 " " c ~) I 2 i 2% 7l + E, 11 "e ~ 4.4, e ", (,, " " [(Ea egg) t! ti) "+ " [(Et "egg) t + fJ Q + Q e tt (2O) The periodic term in (20) is separable from the non-periodic experimentally beca use Ortsbestirnmungen can be performed at different phases of the irradiated lig ht. In a famous thought experiment of drilling indicated the atoms of a Stern-Gerlac h's atomic beam are initially excited at a specific location by irradiated light for resonance fluorescence. Xach a part of the way they go through an inhomogen eous magnetic field, the radiation emitted by the atoms can the whole way, befor e and after the magnetic field is observed. Before the atoms are in the magnetic field, resonance fluorescence is usual, that is similar to the theory of disper sion must be assumed that all of the atoms emit in phase with the incident light , spherical wave. This last view is in contrast to the first, what a rough application of the light quantum theory or quantum-theoretical rule s follows: for from it one would conclude that only a few atoms are lifted into the "upper state" by taking a light-quants, so the entire resonance radiation wa s coming from a few intense radiant excited centers ., it seemed obvious early t o say that the light quanta considered here may be used only for the energy bala nce of momentum, "in reality" radiate all the atoms in the lower state of weak a nd coherent spherical waves. After the atoms have passed through the magnetic fi eld, but can hardly be a doubt that the atomic beam has been split into two beam s, one of which the atoms in the upper, the other corresponds to the atoms in th e lower state. Now, if the atoms radiate in the lower state, so here would be a gross violation against the energy record, because the entire excitation energy is in the atomic beam with atoms in the upper state. Rather, there can be no dou bt that behind the magnetic field only emits a beam of atoms with the upper stat es Lichta "namely unkohrentes light " the few intensely radiating atoms in the upp er state. As drilling has shown particularly clearly makes this thought experime nt, which caution in applying the concept sometimes "steady state" is necessary. From here, the developed view of quantum theory from a discussion of Bohr's exp eriment can be carried out without difficulty. In the outer radiation] ungsfelde the phases of the atoms are determined, so it makes no sense to speak of the en ergy of the atom. Even after the atom has left the radiation field, you can not say that it would be in a certain steady state, if one asks for the coherence pr operties of the radiation. But you can do experiments to check what state the at om is, the result of this experiment can be given only statistically. Such an ex periment is actually performed by the inhomogeneous magnetic field. Behind the m

agnetic field, the energies of the atoms, the phases are determined, so indeterm inate. The radiation is incoherent and here only of atoms in the upper state. Th e magnetic field defines the energy and therefore destroy the phase relationship . Bohr's thought experiment is a very nice explanation of the fact that the ener gy of the atom "in reality" not a number but a matrix., The conservation law is as accurate as this each measured for the energy matrix and, therefore, for the value of energy will. Computationally this can be the repeal of the phase relationship as follow this: Be Q the coordinates of the atomic center of gravit y, it will be the atom instead of (19) the Kigenfunktion S (Q, t) S (tf, t) = S (Q, tf, t) (21) assign, where S (Q> t) is a function that [as S (rf,> f) (delete )] is different in only a small area of a point of zero in the Q-space, and propagates at the speed of the atoms in the beam direction. Th e probability of a relative amplitude value for any Q tf is given by the integra l of S (Q, tf, t) S (Q, tf, t) Q, ie (20). The Kigenfunktion (21) is calculated in the magnetic field, but also change and transforms because of the different deflection of the atoms in the state after t he upper and lower magnetic field have s i (t It + p) S (Q, If, t) = c, Ss (Q, t) tt> s (It, tf) e 2 t E> t + 01 "c'S, (Q t) It, (V e) c '" (22) S (Q, t) and Ss (Q> t) be functions of the Q-space, which are different only in a small neighborhood of a point of zero, but this point is S, a different one th an for Ss. Sr S "is therefore zero everywhere. The probability of a relative amp litude d and a given value Q is therefore S (Q> tf> t) S (Q> g> t) "cs Ss Ss fs + gs (1 cs) Sy St 'ihy Qy (23) I) as perio dic limb from (20) has disappeared, and thus wretchedness to measure a phase rel ationship. The result of the statistical l) rtsbestimmung will always be the sam e, no matter at v-eicher phase of the incident light, they would be arranged. We may assume that experiments with radiation, the theory has not yet been carried out, the same results on the phase relationships of the A.tome to the incident light will arise. Habens still was the end nnt a Problemkornplex studied the relationship of equat ion (2) Et tr ~ Jt, K the lead firm and other researchers') on the basis of the B oh r shy Eorrespondenzprinzips zv in an important work discussed). Ehrenfest a nd Tolman speak of "weak quantization" when a quantized periodic motion by quant um leaps or other disturbances under) P. Ehrenfest and G. Breit, CA. f Phys. 9, 207, 1922, and P. Honour fixed and R C Tolman, Phys. Rev. 24, 287, 1924, see also the discussion in X. Boring, basic postulates of quantum theory I. c. s) In this context, I am Mr. W. Paul has pointed out. is interrupted at time intervals, which is very long relative to the period of t he system can not be considered. It should in this case not only happen the exac t quantum moderate energy values, but with a lower quality of specifiable a prio ri probability and energy values that do not deviate too far from the quantum mode rate rates. In quantum mechanics, this behavior can be interpreted as follows: S ince the energy of the external disturbances or the quantum leaps is really chan ged, so must each energy measurement if it is to be unique, that plays in a time between the two disorders. This is an upper limit for t given in terms of g 1st So the energy value Ei ~ a quantized state we measure only l ~ with a precision e, ~ "this is the question of whether the system l Such energy values E that differ from E, "really" accepting the correspondingly sm aller statistical rumor, or whether their experimental determination lies only t o the inaccuracy of the measurement principle no sense. If t is less than the pe riod of the system, so it has no sense to speak of discrete stationary states or

discrete energy values. Make Ehrenfest and wide (lc) in a similar context to the following paradox atten tion: A rotator that we want to think about as we gear is provided with a device that just reverses the direction of rotation of f revolutions of the wheel. The grab some gear in a rack that is linearly movable in turn between two blocks, t he blocks zv-ingen after a certain number of turns the rod and thus the wheel to turn back. The true period T of the system is long in relation to the current t ime t of the wheel, the discrete energy levels are correspondingly tight, and th at the closer the larger T is. Since from the standpoint of consistent quantum t heory of all stationary states have the same statistical weight, virtually all e nergy values for sufficiently large t occur with equal frequency "as opposed to wh at would be expected for the rotator. This paradox is zunch by consideration of o ur points of view. '/ Aggravated. Namely, to determine whether the system will a ccept the rotator corresponding to the discrete energy values pure alone or very f requently, and if it is equally likely all possible A Values (that is, values that correspond to the small energy levels ")-to takes sufficient time t "is small relative to T (but)) t), ie although the great period for such measurements do not come into effect, it apparently manifests i tself in the fact that you may experience all kinds of energy values. We believe that such experiments to determine the total energy of the system would yield t he same probably really all possible energy values, and that this result is not large period T, but the linearly movable rod blame. Even if vstem even that) is in a Zdstand, corresponds to the energy of the Rotatorquantelung, so it can by e xterior forces acting on the rod, easily be converted into those which do not me et the iotatorquantelung '). The coupled system: rotator and bar just shows comp letely different Periodizitatseigenschaften as the rotator. The SOLUTION of the paradox lies rather in the following: If we want to measure the energy of the ro tator alone, we must first solve the coupling between iotator and rod. In the cl assical theory could be done to solve the Sopplung without energy change for suf ficiently small pale of the bar, so there could be the energy of the whole syste m of the rotator (with smaller mass of the rod) set equal to. In quantum mechani cs, the energy of interaction between rod and wheel at least of the same order o f magnitude as the energy level of the iotators (even with a small mass of the r od is for the elastic interaction between the wheel and rod a high zero-point en ergy!); Provide in solving the Sopplung for rod and wheel individually their qua ntum moderate energy values here. If we say the energy values of the rotator alone c an measure, we can always experiment with the limits set by the moderate accurac y, the quantum energy levels. Even with vanishing mass of the rod but the energy of the coupled system of the energy of the rotator is different, the energy of the coupled system can take all possible (approved by the Z-quantization) values equally likely. The quantum theoretical kinematics and mechanics is largely different from the o rdinary. The applicability of the classical kinematic and mechanical terms but c an be inferred either from our laws of thought nor of the experience, this concl usion ') This can not, or only very rarely done by forces acting on the bath after Ehr enfest and wide. gives us the relation (1) p, q, ~ b the right. Since pulse, place, power, etc. a re well defined terms of an electron, one does not need to push the fact that th e fundamental equation (1) contains only a qualitative statement. Since we can a lso consider the experimental implications of the theory in all cases simple qua litative us, you will not be more than not vivid and abstract) to view the quant um mechanics'. Of course, one might, if one admits this, the quantitative laws o f quantum mechanics directly from the philosophical foundations, ie essentially the relation can be derived (1). Jordan has tried therefore, the equation S (q q ') = J S (qq) S (g'q') dg ' be interpreted as a probability relation. This interpretation might differ but w

e do not agree with (g 2). Rather, we believe that the quantitative laws for the time being can be understood from the philosophical foundations out only on the principle of maximum simplicity. For example, if the X coordinate of the electr on does not "count" more, as in Equation (1) can be deduced experimentally, it i s the simplest assumption [which is not consistent with (1) in contradiction] th at this X coordinate is a member of a diagonal matrix whose non-diagonal element s are manifested in a lack of precision or with transformations in other ways (s ee, eg, 4 oz).'s statement that about the speed in the X direction "in reality" not a number but diagonal element of a matrix is, might not be abstract and unan schaulicher, -. os the finding that the electric field strength "in reality" the proportion of time an antisymmetric tensor of space-time world is the word "rea lly" is here as much and as be entitled to little, as with any mathematical desc ription of natural processes. Once it is admitted that all the quantum-theoretic al sizes, in reality 'are matrices that follow the quantitative laws without dif ficulty. r) denotes the Schrdiuger ttuautenmechauik as a formal theory of deterrent, yes a bstofieuder Uuanschaulichkeit and abstractness. Sure you will not be able overes timated the value of the mathematical (and thus philosophical) penetration of qu auteumechauischen laws that Schrdinger has done theory. Iu the basic, physical is sues but in my opinion has the popular vividness of wave mechanics removed from the straight path which was prefigured by the work of de Broglie and Einstc the one hand, by the work of Bohr and quantum mechanics on the other. If one assumes that the attempted here interpretation of quantum mechanics is co rrect already in essential points, so it may be possible to meet in a few words on their principal consequences. That quantum theory, in contrast to classical s tatistical theory is a much in the sense that only statistical conclusions could be drawn from the exact given data, we have not adopted. Speak against such ass umptions so for example, the well-known experiments of Geiger and Bothe. Rather apply in all cases where the classical theory relations existing between quantit ies that are actually all precisely measured, the corresponding exact Relatiouen in quantum theory (momentum and energy conservation). A.her at the sharp formul ation of the law of causality: "If we know the present exactly, we can calculate the future", not the Xachsatz, but the premise is false, we can the present in all its determinants in principle, to know not why everything is perceiving. sel ect from a wealth of possibilities and a limitation of future possibility. Since the statistical nature of quantum theory is so closely linked to the inaccuracy of all perception, it is tempting to suspect that behind the perceived statisti cal world another, real 'world hide in the law of causality applies But such spe culations seem to us that we must emphasize barren and meaningless Physics to de scribe the relationship of perceptions formally Rather, one can characterize the true facts much better as:... because all experiments are subject to the laws o f quantum mechanics, and thus the equation (1), the invalidity of the Causal Law is definitely determined by quantum mechanics. Xachtrag for the correction. Vach completion of this work have led to recent inv estigations of drilling points that allow a substantial deepening and refinement of the analysis attempted in this work, the quantum mechanical correlations. In this context, I was drilling attention <emacht that I had overlooked key points in some discussions of this work. Before. mainly due to uncertainty in the obse rvation not only on the presence of discontinuities, but is directly related to the requirement to meet the different experiences at the same time, on the one h and in the corpuscular the wave theory on the other hand are expressed. For example, to consider the us e of an imaginary I'-ray microscope, the required divergence of the beam into ac count, it has only the effect of the direction of the Compton recoil is known, i n the observation of the electron map with an inaccuracy. then the relation (1) leads. Furthermore, it is not enough emphasized that the simple theory of the Co mpton effect in rigor is applicable only to free electrons. The resulting cautio

n in applying the uncertainty relation is, as Professor Bohr has made clear, amo ng other things essential for all-round discussion of the transition from microto macro mechanics. Finally, the considerations about the resonance fluorescence are not entirely correct, because the relationship between the phase of the lig ht and the electron motion is not as easy as expected. For the fact that I get t o know and discuss emerging allowed the more recent studies mentioned ear B, whi ch will be published in a paper on the conceptual structure of quantum theory so on, I am obliged to Prof. Bohr hearty thanks. Copenhagen Institute for Theoretical. Physil University.

También podría gustarte