A. G. Stoletov’s Work on the Photoelectric Effect

N. S. Khlebnikov, Moscow

I. Among the names of the first researchers of photoelectric phenomena—the names of those people whose works have been embodied in our own day in numerous technical achievements—the name of A. G. Stoletov occupies one of the most honorable places.

A. G. Stoletov’s works devoted to the photoelectric effect belong to that period in the development of this field which is characterized by the initial accumulation of experimental material and the search for the basic general directions of systematic research. Indeed, Stoletov’s first work was begun in 1888—the year following the publication by H. Hertz of his observations on the influence of ultraviolet light on the discharge of an inductor—and the last is dated 1891.

In this short interval of time A. G. Stoletov established a whole series of important facts characterizing the basic features of the phenomenon, and developed the classical methodology for its experimental investigation. His premature death interrupted these important works.

In order to clarify the role of Stoletov’s investigations, it is necessary to be aware of the level of knowledge about phenomena connected with the external photoelectric effect that existed at the end of the 1880s and the beginning of the 1890s. Without this, i.e., without the correct historical perspective, it is impossible either to understand the significance that Stoletov’s works had at that time, or to assess the difficulties he had to encounter both in interpreting the results and in the very formulation and conduct of the experiments.

First of all it must be borne in mind that the end of the 1880s and the beginning of the 1890s were precisely the period of the triumph of the electromagnetic theory of Faraday–Maxwell–Hertz—a triumph created by Hertz’s experiments (1887). This meant the dominance of wave conceptions of radiant energy as applied to interaction with matter, on the basis of which (as became clear later) nothing can be understood in the external photoelectric effect. Still more essential at this stage were

the prevailing views on the nature of electricity, based on this theory. Following Faraday, Maxwell and especially Hertz shifted the center of gravity of electrical phenomena into the ether, renouncing all evidence for the corpuscular structure of electricity. Thus, for example, with respect to cathode rays, which behaved quite manifestly as streams of particles of electricity, being deflected in magnetic and electric fields, Hertz held that they too were disturbances in the ether, arising as a result of impulsive discharges at the surface of the cathode.^1 This circumstance is clearly felt in Stoletov’s works devoted to investigations of the photoelectric effect at atmospheric pressure. In establishing the law of the saturation current, he says that the magnitude of the current is proportional to the charge density on the surfaces of the electrodes:

\[ i=\varphi\left(\frac{E}{\delta}\right)\varphi(\sigma) \]

(\(E\)—the potential difference between the electrodes, \(\delta\)—the distance between them, \(\sigma\)—the charge density), but immediately rejects this “fiction” and returns to the “electric force acting at the surface of the plates.”

Nevertheless, the idea of charge as something material runs through Stoletov’s reasoning, both here (there is no need whatever to pass from the potential gradient \(\frac{E}{\delta}\) to the electric force through the charge density \(\sigma\), since \(\frac{E}{\delta}\) precisely gives the field strength) and elsewhere, especially when he discusses the nature of photoelectric (“actino-electric”) currents, emphasizing that they should in all probability be regarded as convection currents.

Stoletov conducted all his experiments either in an atmosphere of air at ordinary pressure or in gases not too strongly rarefied. This made understanding the phenomena difficult, since the first theory of gas discharge was created only in 1900 (by Townsend), and the impetus for its creation was again provided by Stoletov’s experiments,^17 in which currents greater than the saturation currents were discovered. When working in air at normal pressure, when saturation currents were obtained, there was no reason to distinguish between primary and secondary processes, i.e. between the photoelectric effect and the gas-discharge phenomena accompanying it. But when Stoletov transferred his investigations to a rarefied gas,^2 he immediately pointed out that the complication of the picture (the absence of saturation) is caused by a secondary phenomenon. He believed that improving the vacuum would lead to final values of the photoelectric current independent of the potential. This proposition was confirmed by Righi’s experiments^3 in the same year, 1890, and fina-

positively established as a result of Lenard’s investigations (1899).

In his work Stoletov had, from the very beginning, to construct the experimental method and its technique. He had no predecessors in the field he had chosen, since the experiments of Hertz, Ebert, Wiedemann, and Hallwachs dealt with the action of ultraviolet light on conductors placed under very high potentials (sparks). A number of procedures first used by Stoletov have become widely adopted in the practice of scientific research—for example, the method of the emitting (now incandescent) probe for investigating the distribution of potential in the interelectrode space. None of them, however, bears his name.

II. Stoletov began his investigations with the aim of establishing whether the phenomena observed by Hertz and others would appear at lower potentials. Besides answering this question itself, he showed that, if the results proved successful, this would provide greater possibilities for the quantitative study of the phenomenon, since all measuring techniques for lower voltages are simpler and make it possible to obtain greater accuracy.

The arrangement of the principal experiment, which after several unsuccessful attempts Stoletov succeeded in carrying out

Fig. 1.

on March 9, 1888, is shown in Fig. 1. In front of a lantern with a voltaic arc, \(A\), from which all (glass) optics had been removed, there were placed two disks (22 cm in diameter). One of them, the one nearest the lantern, consisted of a metal (iron, brass, sometimes electroplated with another metal) mesh stretched on a ring made of thick wire. The second disk was a solid metal plate. These disks were connected to one another by a circuit consisting of a battery \(B\) and a galvanometer \(G\). The battery was assembled from Volta, Daniell, Beetz, Gassner, or L. Clark cells, and Stoletov most often used the last of these. The voltage between the plates was regulated by connecting the appropriate number of cells (usually varying within the limits from 1 to 200 V, with the exception of certain investigations that were carried out—

were made at the very lowest voltages—down to 0.01 V, or with no voltage source at all—and were measured by their number. Thus, for example, Stoletov says that the voltage between the electrodes was equal to 100 Clark (1 Clark = 1.43 V). As the measuring instrument he used a high-resistance (5212 Ω) Thomson astatical galvanometer, whose sensitivity without the upper astatizing magnet was \(6.7\cdot 10^{-10}\) A, and with this magnet present was \(2.7\cdot 10^{-11}\) A. Observation of the deflections was carried out “by the English method,” i.e., with a lamp and scale.

Discussing the properties of his apparatus, Stoletov makes the only remark with which it seems impossible to agree now. Pointing out that, when the polarity is changed to the reverse of that shown in Fig. 1, considerably smaller currents are observed than with the usual arrangement, he says: “Owing to the property of the front grid armature, the rear armature could be illuminated by the rays of the voltaic arc from the inner side, i.e. from the side where electric charge is predominantly accumulated. The other armature (the grid) was illuminated only from the disadvantageous (weakly charged) side, and from the inner side only by rays reflected from the solid disk.” In reality, however, the point here is, first, that the illuminated surface of the grid is much smaller than that of the solid plate (the investigation of the dependence of the photocurrent on the size of the illuminated surface later became the subject of a special study by Stoletov himself), and second, that the electric field in this case was substantially weakened by the shielding action of the grid. The quoted passage is of particular interest in another respect: it shows once again that, in Stoletov’s conceptions, charge was something more “material” than the geometrical location of the ends of the field lines.

The greatest difficulty in carrying out the experiments was caused for Stoletov by the voltaic arc, which was powered by a dynamo driven by an internal-combustion engine (Otto system), since electricity from the municipal grid was supplied to the laboratory only in 1889. Fluctuations in the number of revolutions of the engine, despite the presence of a regulator, led to such fluctuations in the intensity of the ultraviolet radiation of the arc that quantitative investigations could not be carried out. This forced Stoletov first to try to establish a connection between the electrical regimes of the arc and its radiation; but since this did not lead directly to the goal, he had to develop direct methods for monitoring the radiation—photoelectric methods.

These methods consisted in placing an additional photocell (which had a cathode in the form of a cross with crossbars measuring \(12 \times 1\) cm and a grid anode of the same shape and size—

…rays, fixed on ebonite spacers 2 mm from the cathode) in the path of the beam between the main photoelement and the lamp, which was supplied from a separate battery. In the first method, the photocurrent (of the first deflections) of the main and control photoelements was measured alternately on one and the same galvanometer. However, because of the large period of the galvanometer (17 sec.), this method did not always give the required results. Therefore it was modified in the sense that the control photoelement (the voltage on it, the distance from the source, and the distance between the electrodes remaining unchanged throughout the entire series, and sometimes even for a number of series of measurements) was connected to a separate galvanometer (the period of which was adjusted to the period of the main one); its readings were observed simultaneously with the readings of the main one by a second observer (I. F. Usagin). This method of control Stoletov found quite satisfactory, having tested it in a large number of special experiments. The only defect here was the different damping of the two galvanometers, which somewhat reduced the reliability of the control.

Assessing the photoelectric method of controlling the intensity of radiation, Stoletov writes: “The extraordinary sensitivity of the actino-electric current to any change in the arc not a little hinders quantitative observations... There is hardly any other way of so keenly following the constancy of electric light (or, more precisely, the intensity of a known category of radiation) as these actino-electric observations.” We know now how right Stoletov was; and what is especially remarkable is that both methods first implemented by him are used in photometry. Thus, in the book by Simon and Zuurman⁴, Photoelements and Their Applications, in the section “Methods for Eliminating Fluctuations in Light Intensity” (pp. 208 and 209), a circuit with two photoelements and two electrometers is given, whose author is Pohl.⁵ In fact, in principle this circuit is an exact reproduction of the second of the control methods described above. There too (p. 208) is described Dobson’s circuit,⁶ typical for the “method of flickering light” (Flimmermethode), which is a modification of Stoletov’s first method. This modification, as well as the general applicability of the flickering-light method, was made possible by the appearance of highly sensitive and low-inertia measuring instruments (string electrometers and galvanometers). Further on, in the same section of Simon and Zuurman’s book, a number of photometric instruments based on these circuits and produced by German firms are described.

Another circumstance that greatly interfered with the measurements was the phenomenon of photoelectric

¹ In the above-mentioned review by Hallwachs¹, he especially emphasizes that the phenomenon of photoelectric fatigue was discovered by Stoletov, and not by Kreusler,⁷ as is usually assumed.

A. G. Stoletov’s Works on the Photoelectric Effect

fatigue, which consisted in a lowering of the sensitivity of his photo-elements (or, as he called them, condensers) with time under the action of light. This was especially evident in the control photo-element, which was subjected to more intense illumination.

With the aid of the apparatus described, Stoletov carried out the majority of his investigations.

III. First of all, Stoletov established with complete precision that ultraviolet light exerts a discharging action only on a negatively charged electrode—a view which he expressed in his first publications ⁸˒⁹, which Wiedemann and Ebert ¹⁰ shared, and against which Hallwachs ¹¹ ¹) and Righi ¹² argued. When this last author became convinced of the erroneousness of his assertions ¹³, he tried to present the matter as though it was not he, but Stoletov, who had doubted the insensitivity of the positive electrode ¹⁴. On this occasion Stoletov wrote: “I have never yet encountered a more unceremonious way of shifting one’s sins onto another’s head.”

What questions, then, occupied Stoletov? First, he devoted much time to comparing the photoelectric properties of different metals and different states of metallic surfaces. With respect to metals, no substantial differences were found. Stoletov writes: “If the surface is smooth and well cleaned, every metal proves to be almost the same from this point of view.” On the other hand, the state of the surface proved to be very important: the better the cleaning and polishing, the higher the sensitivity. On the other hand, a freshly cleaned surface proves to become fatigued more strongly.

The question of the connection between the properties of the substance exhibiting the photoelectric effect, the character of the radiation, and the magnitude of the effect deeply interested Stoletov. In this direction he carried out a series of experiments, but it was impossible to draw final conclusions on their basis—for this an entirely different level of experimental technique was needed, which was achieved only 10–15 years later.

Stoletov expressed his view of the nature of the photoelectric effect, which he partly succeeded in confirming, in the following manner: “From the very beginning of my investigations I suspected that its sensitivity to actino-electric action stands in direct connection with the absorption of active rays by one or another plate...” and further: “Already the very unipolarity of the act—

¹) In Hallwachs’s early experiments this error was due to the fact that the observations were made with an electroscope, the charge of which leaked away simply because of imperfect insulation. Hallwachs ¹¹ wrote: “With a positive charge the falling of the leaves [of the electroscope.—N. Kh.] at first glance is absent, but upon careful investigation it becomes noticeable after a long period.” The method developed by Stoletov, owing to the presence of a battery, was free from this defect.

...of action shows that the electrodes... play an essential role in the phenomenon... Rays that illuminate the air layer, without striking the surface of the (negatively) charged body, produce no action (Hallwachs): the rays must fall upon it. Moreover, the rays must be absorbed by the negatively charged surface. Evidently, what is important here is absorption in the thinnest upper layer...”

In testing this view, Stoletov covers the cathode of his apparatus with various strongly absorbing dyes and finds that in many cases (fuchsin, eosin, fluorescein, methyl violet, methyl green, etc.) the sensitivity of such an electrode proves to be several times higher than that of any metal. In connection with these experiments, Hallwachs, in his review,^1 says that they did not directly confirm the views expressed by Stoletov, but emphasizes that they were the first to be directed toward establishing a connection between the absorption of light and the photoeffect, just as the idea of this connection was first expressed by Stoletov.

Stoletov returned to the same fundamental question once more, in connection with the fact he had discovered that a photocurrent can also be observed without the presence of batteries, if the contact potential difference has the proper direction. Here he says the following: “The system Zn, Ag, and air, under the condition that Ag is illuminated by active rays... is transformed into a real galvanic cell. Taking into account that in this case the rays must be absorbed by the silver, we may say (whatever the mechanism of the phenomenon may be) that the energy of the current in this air element arises at the expense of the energy of the illuminating rays.”

From this it is clear how close Stoletov was to understanding the true nature of the photoeffect as early as 1889. However, the formulation of this law was given only in 1905 by Einstein, who based himself on the electron theory (1895, Lorentz), the quantum theory (1900, Planck), and data on the limiting velocities of photoelectrons (1901, Lenard).

In connection with investigations of the influence of the surface, mention should be made of Stoletov’s verification of the experiments of a certain Goor,^15 who had attempted to prove that the photoeffect should be attributed exclusively to adsorbed layers of gas. Goor, in particular, believed that he removed the adsorbed gases by heating the plates to... 55°C. Having spent much time on the verification, Stoletov described this work as “frivolous in every respect.” Stoletov himself found that heating gives a certain increase in sensitivity.

Finally, intending to investigate the photoeffect for different spectral compositions of the radiation, Stoletov carries out experiments with the introduction into the carbon arc of various metals, especially aluminum, on which Stoletov placed particular hopes, since this metal has a very extended ultraviolet spectrum. It turned out that the use of Al increa-

increases the effect in some cases up to 19 times. Stoletov indicated the most convenient method for bringing out Al—in the form of a wick for the positive carbon. This method was subsequently used by Righi, Hallwachs, and other authors. In Stoletov’s work these experiments received no further development, since the extreme inconstancy of the burning of such arcs with metallic vapors made it impossible to carry out quantitative investigations.

In exactly the same way, Stoletov attempted to apply another source of radiation available to him—the sun—but he discovered no effect of sunlight.

Usually Stoletov’s name is associated only with those works which he brought fully to completion, especially with the establishment of the existence of a saturation current. From the preceding it is clear that this is incorrect, since his works outlined a large number of directions of research which were not continued only because of the reasons indicated earlier, and which were developed much later.

The fundamental results on photoelectricity found by Stoletov consist in the establishment of a direct proportionality between the quantity of light energy and the strength of the photocurrent, as well as in the establishment (with the accuracy accessible to him) of the inertia-free character of the photoelectric effect.

The experiments that led to the establishment of these laws attract attention by the care with which they were carried out, of which some idea is given by the development of the control methodology described above. Another characteristic feature of them is the striving to verify results obtained by one method through the use of other procedures, and an extremely attentive attitude toward possible sources of error. All this characterizes Stoletov as an experimenter not only of great initiative, but also of enormous thoroughness in his work—a quality just as important as the first. Both of these, as well as a third—the striving to generalize results and to create an integral picture of the phenomenon, the “gift of foresight,” at the basis of which lies precisely this ability to see a phenomenon from many sides, as well as a strict attitude toward frivolity and bad faith—all this relates Stoletov to another greatest investigator, E. Rutherford.

In establishing the linearity of the dependence between the photocurrent and the quantity of light energy, Stoletov used three methods: a) the decrease of illumination with distance (main and control photocells); b) the change of the photocurrent as a function of the magnitude of the illuminated surface; and c) the rotating-disk method. The experiments carried out by this last method are especially interesting, since, first, they led to the posing of the question of the inertia of the photoelectric effect and to the establishment of the absence of such inertia (with an accuracy up to 0.001 sec.), and, second, because the corresponding

...ing setup of Stoletov is the prototype of all modern setups for measuring the inertia of photocells with a mechanical modulator of the light flux (perforated disks).

Having established direct proportionality between the light flux and the strength of the photocurrent, Stoletov was able to proceed to the study of the influence of the potential difference and the distance between the electrodes on the magnitude—as he considered it—of the photoeffect. Here he still made no distinction between the primary phenomenon, i.e. the photoeffect itself, and the secondary motion of electricity through the gas. But the law of the saturation current discovered as a result of this investigation gave him grounds to make such a separation in his last work, where the photoeffect was observed in a medium of rarefied gas. This, together with his views on the nature of the photoeffect quoted above, compelled him to consider that the photoeffect can be observed in its purest form in the highest vacuum.

In concluding the survey of Stoletov’s photoelectric investigations, it is necessary to dwell on two further facts noted by him. The first of them was discovered while investigating the dependence of the strength of the photocurrent on the quantity of light. Measuring the photocurrent under oblique illumination of the photocathode, the light beam being made parallel by means of a quartz lens and always falling completely within the cathode, he found, for a metallic surface blackened in order to reduce reflection, a noticeable (by 5–10%) increase of the current. Stoletov says that this result remained incomprehensible to him. It could be explained only much later, after the discovery by Pohl and Pringsheim of the vectorial selectivity of the photoeffect.

The second of the facts mentioned once again confirmed Stoletov’s ideas about the existence of an energy exchange between the radiation and the electrode emitting charges. In establishing the law of the saturation current, Stoletov, alongside the usual method—the observation of a constant current under the action of illumination—also made observations with the battery disconnected1. In this case the photocell was shunted by a capacitor of large capacitance (0.05 μF), on which there was a charge from the battery, gradually leaking away through the photocell. Knowing the capacitance and the strength of the discharge current, it was not difficult to calculate the discharge time, which agreed well with the time found experimentally. And it turned out that if the interval of time was greater than that needed for complete discharge, then the capacitor not only discharged but also recharged. A detailed investigation of this phenomenon would probably have led Stoletov to a number of further discoveries and conclusions. But—

...no such investigation was carried out by him. He merely expressed his conviction that charging with positive electricity is nothing other than discharging with negative electricity, which is in complete agreement with reality.

In conclusion, it seems interesting to us to present a résumé of the results of Stoletov’s photoelectric investigations, drawn up by him himself. It concerns all the works except the last, where experiments carried out in various gases and vapors at various pressures were reported. Here it is:

1. The rays of the voltaic arc, falling upon the surface of a negatively charged body, carry away charge from it. Depending on whether the charge is replenished, and how rapidly, this removal of charge may or may not be accompanied by a noticeable fall in potential.

2. This action of the rays is strictly unipolar: positive charge is not carried away by the rays.

3. In all probability, the apparent charging of neutral bodies by the rays is explained by the same cause.

4. The discharging action is possessed—if not exclusively, then with an enormous predominance over the rest—by the rays of the highest refrangibility, lacking in the solar spectrum \((\lambda 295 \cdot 10^{-6}\ \mathrm{mm})\). The richer the spectrum is in such rays, the stronger the action.

5. For discharge by rays it is necessary that the rays be absorbed by the surface of the body. The greater the absorption of the active rays, the more sensitive the surface is to their discharging action.

6. All metals possess such sensitivity, without significant differences, but it is especially high in certain coloring substances (aniline dyes). Water, which readily transmits the active rays, is devoid of sensitivity.

7. The discharging action of the rays is detected even under very brief illumination; moreover, no noticeable time elapses between the moment of illumination and the moment of the corresponding discharge.

8. The discharging action, ceteris paribus, is proportional to the energy of the active rays falling upon the surface being discharged.

9. The action is detected even at negligible negative charge densities; its magnitude depends on this density; as the density increases up to a certain limit it grows faster than the density, and then more and more slowly.

10. Two plates of metals heterogeneous in the Volta series, placed in air, represent a kind of galvanic element as soon as the electronegative plate is illuminated by active rays.

11. Whatever the mechanism of the actino-electric discharge may be, we are entitled to regard it as a certain current of elect—

of the substance, while the air (either by itself or owing to the presence in it of foreign particles) plays the role of a poor conductor. The apparent resistance to this current does not obey Ohm’s law, but under definite conditions has a definite value.

1. The actino-electric action increases with increasing temperature.

IV. The results of Stoletov’s investigations were published by him in six articles, namely:

1. Compt. Rend. CVI, 1149, 1888.
2. ” ” CVI, 1593, 1888.
3. ” ” CVII, 91, 1888.
4. ” ” CVIII, 1241, 1889.
5. ZhRFKhO, 21, 159, 1889.
6. Journal de Physique, 9, 468, 1890.

In addition, in 1889 he wrote the first survey in the literature in this field on the photoelectric effect, published by the Physico-Chemical Society at Petersburg University in that same year, 1889. This survey is based mainly on his own works; the results of other authors are cited chiefly for comparison.

LITERATURE

1. W. Hallwachs, Handb. de Radiologie, B. 11, 1916.
2. A. Stoletow, Journ. de Phys., 9, 468, 1890.
3. A. Righi, Rend. Linc. (4), 6, 151, 1890.
4. G. Simon and R. Zurman, Photoelements and Their Applications, ONTI—GTTI, 1936.
5. R. Pohl, Gött. Nachr. Math.—Phys. Kl., 185, 1926.
6. G. M. Dobson, Proc. Roy. Soc. (A), 104, 248, 1923.
7. Kteusler, Ann. Physik, 6, 398, 1901.
8. A. Stoletow, C. R., CVI, 1149, 1888.
9. A. Stoletow, C. R., CVI, 1593, 1888.
10. E. Wiedemann u. H. Ebert, Wied. Ann., 33, 248, 1888.
11. W. Haliwachs, Wied. Ann., 33, 304, 1888.
12. A. Righi, C. R., CVI, 1349, 1888.
13. A. Righi, C. R., CVII, 559, 1889.
14. A. Righi, N. Cimento, XXV, 15, 1889.
15. Hoor, Sitzungsber. Wiener Akad., XCVII Abt. IIa, p. 719, 1888.
16. O. D. Khvolson, Course of Physics, vol. V, p. 739, GIZ, 1923.
17. O. D. Khvolson, Course of Physics, vol. V, p. 711, GIZ, 1923.

1. One of the by-products of these experiments was the determination of the contact potential differences of various metallic pairs. This new method, the novelty of which lay in the use of a galvanometer, gave good agreement with the data of ordinary determinations. ↩