MECHANISM OF FORMATION OF THE LATENT AND VISIBLE PHOTOGRAPHIC IMAGE

K. Chibisov, Moscow

Our ultimate aim was to exclude naked empiricism, to eliminate the idea that corrections during development are impossible, to ensure the successful production of a satisfactory negative, and in general to place photography among the exact sciences.

Hertter and Driffield (1898)

The photographic plate, as in the age of Daguerre, represents an achievement that may serve as an auxiliary tool in scientific research, but it itself cannot be the subject of scientific investigation.

I. Plotnikov (1923)

The great advances of recent years in the field of physicochemical disciplines could not fail to affect the development of photography; indeed, during this time considerable progress has been made in theoretical respects. This progress concerns chiefly the most important question in photography—the theory of photosensitivity and the nature of the latent image.

It is quite natural that the question of the nature of the latent image should have been linked with the physicochemical interpretation of the mechanism of the action of light on a photographic emulsion. This circumstance is one of the reasons for the successes achieved; but the chief reason that has contributed to the development and deepening of knowledge here is the application to the study of photographic processes of the achievements in the technique of physicochemical research.

A significant role in this direction was played by the application of the X-ray spectrographic method of investigation, on the basis of the theoretical data of modern atomistics, that is, the doctrine of the structure of atoms, of the nature of chemical affinity forces, and of the structure of crystalline bodies. The investigation of the nature of the latent image is a task considerably more difficult than the investigation of the formation of the visible image, since the former is known only as the capacity of the halide salts of the photographic emulsion, after the action of light, to be developed, but not as a ponderable or visible substance. All the existing hypotheses that attempted to explain the mechanism of the primary action of light, which creates the capacity for development, were either physical or chemical in character. Recent investigations, in connection with the general development of ideas about chemical reactions and, in particular, about photochemical processes, have successfully united these opposing points of view.

THEORY OF PHOTOSENSITIVITY AND THE NATURE OF THE LATENT IMAGE OF PHOTOGRAPHIC EMULSIONS

1. It has long been known that a photographic emulsion consists of fine particles of silver halide, the so-called grains, suspended in gelatin. The study of such emulsions has shown that in most cases the grains have an evidently crystalline structure. Recently, the question of the crystalline structure of emulsion grains—both from the standpoint of their crystallographic properties and the conditions of their growth, and from the standpoint of their internal structure—has been the subject of extensive investigation. It has been shown that the grains of silver halide in photographic emulsions are, without exception, crystals of the regular (cubic) system. Owing to the peculiar conditions of growth in the emulsion, very different external crystalline forms of grains are observed, as is seen from Fig. 1 (Table I). Most often the grains have the form of plates, which, owing to special phenomena during the drying of the emul-

MECHANISM OF FORMATION OF THE PHOTOGRAPHIC IMAGE

... axes are arranged with their larger surface parallel to the surface of the layer.

The size of emulsion grains ranges from tenths of a micron to 5 μ and even 10 μ in diameter; crystals in the form of plates usually have a thickness equal to \(1/10\)—\(1/15\) of their diameter.

The latest X-ray investigations have confirmed the data on the crystalline structure both of silver halide salts in pure form and of emulsion grains. These investigations have shown that silver chloride and silver bromide have the structure of a spatial crystal lattice of the NaCl type, in which the silver and halide ions are held by virtue of the equilibrium of electrostatic forces of attraction and repulsion and are arranged in the form of two interpenetrating face-centered cubic lattices in such an order that each point of the simple cubic lattice is occupied by a cation or an anion. Silver iodide exhibits dimorphism: it crystallizes in a hexagonal form with a spatial lattice of the ZnO and CdS type and in a cubic form—with a lattice of the diamond or ZnS type. In the case of the hexagonal form, stable at low temperatures, ions of opposite signs are arranged in the form of close-packed hexagonal lattices, which interpenetrate in exactly the same way, so that each ion is surrounded by four ions of the opposite sign. The cubic form is stable above 146°; this temperature is the transition point.

The method of X-ray analysis has shown, moreover, that even in those cases when emulsion grains do not have a clearly expressed crystalline form, as is observed in the case of such fine-grained emulsions as Lippmann emulsions, they nevertheless have an internal structure corresponding to a crystal.

Of particular interest for understanding the photochemical nature of emulsion grains is the study of mixed crystals of silver bromide and silver iodide, \(Ag(Br, I)\), since in most cases negative photographic...

silver emulsions contain silver iodide in amounts from 1 to 10%. Homogeneous crystals are formed under the condition that the content of silver iodide in the mixture does not exceed 32 mol %, as was shown by investigation of the electrical potentials of the mixtures and of the spectra of selective sensitivity of iodobromide emulsions, and was confirmed by X-ray analysis. Mixed crystals Ag (Br, I) form a lattice like that of pure silver bromide, the iodine atoms being distributed at lattice sites among the bromine atoms according to the law of chance; the dimensions of the elementary cube in the mixed crystal are somewhat increased in comparison with the dimensions for pure silver bromide, and the more so the greater the AgI content in the crystal. One of the characteristic features of mixed crystals Ag (Br, I) is the presence in them of an optical anomaly (double refraction), which is connected with internal stress in the crystal due to deformation of the space lattice.

1. The grains of silver halide constituting a single emulsion differ very greatly from one another in their dimensions. The distribution of grains by size is one of the important characteristics for explaining the photographic properties of silver-halide emulsions, since these properties must be regarded as the sum or integral of the probabilities of a very large number of elementary sensitivities of individual grains.

The dependence between light sensitivity and grain size is essentially statistical in the sense that individual large grains need not necessarily possess greater light sensitivity in comparison with grains of small size; but in the general count it is observed that, for a given emulsion, light sensitivity increases with increasing grain size. This dependence is complicated by the circumstance that, in the case of iodobromide-silver emulsions, grains of different sizes have a non-identical quantitative composition; namely, grains of larger size contain a greater quantity of silver iodide. It must be noted, however, that the dependence of light-

sensitivity on grain size is determined not only by a quantitative difference in the composition of grains of different sizes, since a similar dependence is also observed in the case of pure silver-bromide emulsions, but by other causes, more essential for light-sensitivity.

1. Thanks to the work of a whole series of investigators, especially in the last decade, it has been fully proved that the process of development of grains subjected to brief illumination begins from individual points, the so-called centers of development. The picture of the initial stage of this process is analogous to what is observed in the visible photochemical decomposition of the grain; that is, the reduction of silver halide in the process of development, just as under the action of strong light, begins not uniformly throughout the whole grain, but only at certain discrete points, distributed according to the law of chance. The centers of development are by no means formed under the influence of the developer solution alone; they are present in the grain before its action, arising as a result of a photochemical reaction on pre-existing sensitizing nuclei, whose presence on the surface of the grain is confirmed by their destruction, and consequently by a general light-sensitivity, by certain oxidizing agents. By counting the centers of development and the developed grains it has been established, as a limiting condition, that the average number of centers is proportional to the surface of the grain; consequently, the probability that a grain will acquire the capacity for development also depends on its surface. One should not, however, attach too great an importance to this circumstance, since it proves necessary to assume the existence of a differential distribution of light-sensitivity over the surface of grains, which apparently is not in direct dependence on the size of the grains. The nature of this latter element of light-sensitivity, arising as a consequence of the presence on the surface of the grain of nuclei in the form of inclusions of a foreign substance, is revealed chiefly in the formation of the latent image. By applying an exceedingly original statistical

method, Svedberg proved that the centers of development are located on the surface of the grain.

The factual data presented, together with certain phenomena of chemical action on emulsion grains, also led to the assumption that the grains possess a certain kind of heterogeneity in their structure and that this heterogeneity is one of the principal causes of their great photosensitivity. The heterogeneity of the emulsion grain, as was indicated above, is explained by the fact that into its spatial crystalline lattice, composed of Ag+ and Br− ions, a foreign substance is included in negligible quantities. These inclusions play the role of sensitizing nuclei, around which both the action of light and the action of chemical agents are concentrated. The inclusion of a foreign substance in the spatial crystalline lattice of the grain must occur during the ripening of the emulsion, when its photographic properties are being formed, by means of the syncrystallization of silver halide with the substance constituting the nuclei. The nuclei altered by light then form centers of development, from which the process of reduction of the silver halide to metallic silver begins; consequently, they play the role of a seed for the beginning of chemical reduction, as is often observed also in the case of other physico-chemical phenomena. For chemical development, only those centers are significant which are situated on the surface of the grain or very close to it and, consequently, come into contact with the developing solution. In this connection, the usual photosensitivity of an emulsion must likewise be determined by nuclei situated on the surface of the grains.

1. It is well known what enormous importance gelatin has in photographic production with respect to the properties of the emulsion being prepared. Therefore there was every reason to see the cause of nucleus formation precisely in the gelatin. Gelatin, as a colloidal medium, plays in the process of preparing a photographic emulsion the role of two basic functions: 1) it regulates recrystallization, or the growth of individual grains, and 2) independently of this, it

Fig. 1. Microphotographs of emulsion grains
(linear magnification \(2500\times\))

Fig. 4. Microphotographs of emulsion grains subjected to photochemical decomposition (linear magnification \(3000\times\))

Fig. 5. Geometrically regular photochemical decomposition of an emulsion grain (photographing was carried out after every 15 min. of illumination (linear magnification \(2500\times\)).

Fig. 6. Individual moments in the development of normally illuminated emulsion grains (linear magnification \(3300\times\))

Fig. 7. Successive stages of development of normally illuminated emulsion grains (photographing was carried out after every 3 min. of development (linear magnification \(3300\times\))

Fig. 8. Formation of “protuberances” of metallic silver during the photochemical decomposition of an emulsion grain (1 — grain, normally illuminated and processed with developer; 2 — after additional strong illumination)

Fig. 9. Microcinematography of the process of development of emulsion grains (magnification from individual frames; linear magnification in the frame \(800\times\))

increases the light-sensitivity of the grains by causing the formation of nuclei on their surface. The question of the chemical nature of those traces of foreign substance which constitute the nuclei in emulsion grains is of enormous importance not only for the theory of light-sensitivity, but also for production. So long as the composition of the nuclei remained undetermined, all the results obtained and described above represented rather only a working hypothesis. The earliest assumption was that the grains contain nuclei of metallic silver, which is formed during the ripening of the photographic emulsion as a result of the reduction by gelatin of small amounts of silver halide. This possibility has recently found confirmation, and therefore this assumption is now no longer a mere hypothesis.

Veitert and Luhr, applying an electrometric, very precise quantitative method that makes it possible to take into account traces of metallic silver, determined the amount of this metal capable of passing into nitric-acid solution in an unexposed emulsion. The amount of metallic silver, determined in four different sorts of plates, ranges from \(1.82\) to \(2.68 \cdot 10^{-4}\) mg per \(1\ \mathrm{cm}^2\). Determination of the metallic silver at various stages of ripening of the photographic emulsion showed that its amount increases with the duration of the ripening process. These authors further showed that treatment of the emulsion with a solution of an oxidizing agent (a solution of potassium persulfate), which causes a decrease in light-sensitivity, also reduces the amount of free metallic silver in the emulsion.

However, a far more important role in increasing light-sensitivity is apparently played by minute impurities in the gelatin of certain organic sulfur compounds which, entering into chemical interaction with silver halide, form nuclei of silver sulfide. This question received a comparatively complete solution only in recent years, thanks to work carried out on a broad scale by Sheppard in the laboratories of the Eastman Kodak Company. To clarify the chem-

of the role of gelatin in the emulsion-ripening process, it was first of all necessary to develop a method by means of which this influence could be distinguished from its colloidal function in the sense of regulating the recrystallization of the emulsion grains. With the aid of the statistical method developed for taking account of the quantitative distribution of grains by size, it was shown that on different sorts of gelatin photographic emulsions can be prepared which, with a practically identical distribution of grains by size, will have very different light sensitivity; that is, gelatin may possess, photographically, various degrees of activity. As a result of many years of work it became clear that the sensitizing role in gelatin is played by allyl mustard oil (allyl isothiocyanate, $\mathrm{CH_2=CH.CH_2.NCS}$), present in an amount of 1 part by weight per $300\,000$—$1\,000\,000$ parts by weight of active gelatin.

1. Allyl isothiocyanate itself does not react with silver halide, since it is insoluble in water; but isothiocyanates readily react with ammonia and primary amines, forming thiocarbamides according to the following scheme:

\[ \mathrm{R.N=C=S + NH_3 \longrightarrow C=S} \begin{matrix} \nearrow \mathrm{NH.R}\\ \searrow \mathrm{NH_2} \end{matrix} \]

Similar conditions apparently also obtain in the preparation of a photographic emulsion; therefore the supposition naturally arose that the directly sensitizing action is exerted by allylthiocarbamide, i.e. allylthiourea or thiosinamine:

\[ \begin{matrix} & \mathrm{NH.C_3H_5}\\ & /\\ \mathrm{CS}&\\ & \backslash\\ & \mathrm{NH_2} \end{matrix} \]

The experiments carried out fully confirmed this supposition.

Mechanism of Formation of the Photographic Image

Thus, for the phenomenon of photographic sensitization, the group

\[ \begin{array}{c} \mathrm{N}\ldots\\[-2mm] \diagdown\\[-1mm] \mathrm{C}=\mathrm{S}\\[-1mm] \diagup\\[-2mm] \mathrm{N}\ldots \end{array} \]

is of fundamental importance, where the sulfur atom is attached by a double bond. Sulfur compounds in which sulfur occupies a different position, such as mercaptans, thioethers, and sulfur-containing heterocyclic compounds, do not possess a sensitizing action.

Despite their sensitizing capacity, thiosinamines do not, however, enter into the composition of the nuclei that constitute inclusions of silver sulfide formed as a result of the following chemical reactions:

1) allyl isothiocyanate is converted into allylthiocarbamide:

\[ \mathrm{C_3H_5\cdot NCS}+\mathrm{NH_3}\longrightarrow \begin{array}{c} \mathrm{NH\cdot C_3H_5}\\[-1mm] \diagdown\\[-1mm] \mathrm{C}=\mathrm{S}\\[-1mm] \diagup\\[-1mm] \mathrm{NH_2} \end{array} ; \]

2) allylthiocarbamide reacts with silver halide with the formation of an intermediate complex compound of the composition:

\[ \mathrm{AgX}\cdot \begin{array}{c} \mathrm{NH\cdot C_3H_5}\\[-1mm] \diagdown\\[-1mm] \mathrm{CS}\\[-1mm] \diagup\\[-1mm] \mathrm{NH_2} \end{array} , \]

where \(\mathrm{X}=\mathrm{Cl}, \mathrm{Br}, \mathrm{J}\);

3) the intermediate complex compound decomposes in an alkaline medium with the formation of \(\mathrm{Ag_2S}\).

The intermediate complex compound of the indicated composition was in fact isolated for silver chloride, bromide, and iodide, and it was also shown by a microphotographic method that the formation of silver sulfide nuclei takes place at discrete points of the grain, which is promoted by the formation of intermediate compounds.

The degree of photosensitivity depends on the amount of thiosinamine added to the emulsion as a chemi-

of the chemical sensitizer; namely, there exists an optimal concentration at which the greatest photosensitivity is obtained; further addition of thiosinamine, on the contrary, leads to a fall in photosensitivity and an increase in fog, which is connected, on the one hand, with an excessive increase in the size of the nuclei, in the presence of which the grains become capable of development without exposure to light; and, on the other hand, with an increase in the number of nuclei in individual grains, which, in turn, creates a certain kind of competition in the distribution of photochemical work during the formation of the latent image.

1. The essence of the problem of the formation of the latent image at the present time is the question of how sensitizing nuclei in grains facilitate the photochemical process.

If the existence of nuclei of metallic and sulfurous silver, or of both together, as Clark, Wightman, and Quirk assume, may be regarded as fully proven, their role in the formation of the latent image still remains not entirely clarified.

First of all it is necessary to point out that nuclei are significant only for ordinary light rays (visible and ultraviolet), but not for X-rays and a stream of α-particles, a fact which is highly characteristic for understanding the photochemical role of nuclei. X-rays in a single quantum and individual α-particles impart to the grain an amount of energy sufficient to liberate such a number of silver atoms that together they give a development center of the required dimensions; in such a case the sensitivity with respect to this energy will not depend on the presence of nuclei in the grain, which is confirmed experimentally in tests of the comparative use of low-sensitivity and high-sensitivity emulsions and of the influence of desensitization by oxidizers with respect to the action of α-particles and X-rays. The formation of a latent image under the influence of α-particles or X-rays apparently occurs as the result of the same basic—

of a photochemical reaction such as occurs under the action of ordinary light. A quantum of X-ray radiation must be absorbed by one bromine atom in the emulsion grain; this absorption is carried out by one of the inner electrons of the atom, which, as a result, is ejected beyond the limits of the atom with a velocity determined by the amount of energy of the quantum. The liberated electron, upon collision with another atom, causes the formation of a new quantum of X-ray radiation with a somewhat diminished store of energy. On the other hand, the replacement of the ejected electron by another from the nearest outer electron shell gives rise to the phenomenon of X-ray fluorescence. Such multiple degradation of energy, proceeding in accordance with the principle of conservation of energy and momentum, ultimately leads to the formation of several hundred quanta of ultraviolet radiation in the region of ordinary absorption of silver bromide, which apparently also form the same number of silver atoms.

1. For explaining the mechanism of formation of the latent image and the role of sensitizing nuclei, it is extremely important to consider photochemical decomposition from the standpoint of the photoelectric effect. On the basis of ideas about the structure of silver halide salts as heteropolar lattices, first Fajans, and then independently of him Sheppard and Trivelli in 1921, proposed that the elementary photochemical process consists in the liberation, under the influence of light, of a valence electron of the bromine ion, which, passing to the silver ion, converts it into a neutral atom; thus, the elementary process is represented in the following form:

\[ \mathrm{Br}^{-}+h\nu=\mathrm{Br}+\Theta; \]

\[ \mathrm{Ag}^{+}+\Theta=\mathrm{Ag}; \]

\[ 2\,\mathrm{Br}=\mathrm{Br}_{2}. \]

Consequently, in order to prove the validity of this conception, it is necessary first of all to establish the parall-

between the primary photochemical process and the photoelectric properties of silver halide salts.

It has long been known that silver halide salts exhibit both an external photoelectric effect, consisting in the complete liberation of electrons, and photoconductivity, or an internal photoeffect, owing to an increase in the mobility of electrons within the spatial crystal lattice under the action of light. Since silver halide salts, on the one hand, do not show an external photoelectric effect under the action of light with wavelength greater than \(\lambda = 300\,m\mu\), and, on the other hand, possess photochemical activity in the rays of the visible part of the spectrum, the external photoeffect therefore cannot be the cause of the formation of the latent image; that is, complete liberation of electrons is not necessary for the primary photochemical process.

The phenomenon of photoconductivity of silver halide salts has been the subject of investigation by several authors, among whom especially valuable results, in the sense of explaining the mechanism of the chemical action of light in the formation of the latent image, were obtained by Koblenz and Toy. From a comparison of the spectral distribution of the photoconductivity of silver halide salts with the photoconductivity of halide salts of other metals (thallium, lead, mercury), obtained in Koblenz’s investigation, it is seen that in various heteropolar salts the light-sensitive ion which liberates the electron is the anion, i.e. the halogen ions. The photoconductivity, measured by the named author for thallium iodide in the temperature interval from \(23^\circ\mathrm{C}\) to \(-60^\circ\mathrm{C}\), proves to be practically independent of temperature, as is also the light-sensitivity in the sense of the formation of the latent image. And, finally, Toy in his investigation showed that the light-sensitivity, in the sense of an increase in photoconductivity in a very thin surface layer (of thickness \(1\)—\(3\,\mu\)) of a silver bromide crystal, increases rapidly with decreasing wavelength down to \(\lambda = 360\,m\mu\). In such thin layers the spectral distribution of the relative photographic effect, of the relative magni-

...the magnitudes of photoconductivity and light absorption, for equal incident energy, turn out to be almost identical. All these facts indicate a high probability of the supposition that the mechanism of the primary stage of the photographic process is analogous to the mechanism by which photoconductivity arises.

1. For a further clarification of the connection between the photochemical process and the photoelectric effect, Sheppard and Vanselow recently undertook a detailed investigation of the Becquerel effect, which consists in the appearance of photopotentials on electrodes made of metallic silver and a silver halide immersed in an electrolyte; in the experiments of the authors mentioned, the photoelement was filled with a solution of KBr saturated with AgBr and was placed in a thermostat. If both electrodes are darkened, the potential difference is practically zero; when one of the electrodes is illuminated, a potential difference appears immediately, and in the first second of illumination a considerable negative difference is observed, which, with further illumination, passes into a positive one. The peculiar form of the curve showing the growth of the potential difference with time, the absence of inertia in the effect, the influence of the thickness of the AgBr layer, the influence of free bromine and of bromine acceptors introduced into the electrolyte on the course of the curve—all together indicate that the observed effect may be regarded as the result of the simultaneous liberation of electrons and formation of bromine atoms. The existence of a maximum negative effect and its subsequent transition into a positive one can be explained as follows: upon the liberation of electrons from the surface layer of an Ag⁺ Br⁻ crystal, a positive space charge is formed in it due to the remaining Ag⁺ ions, which causes a reverse attraction of the electrons; therefore, after the initial influx of electrons to the silver electrode, their reverse motion gradually sets in, as a result of which the negative potential difference at first decreases rapidly and then ever more slowly, as a certain mobile equilibrium is approached. In this process there occurs not only neutralization of the silver ions, but also photo-

chemically liberated bromine atoms recombine with silver atoms, which lowers the positive potential difference; the bromine ions thereby formed will again undergo secondary photolysis; this also explains the presence of a quasi-stationary potential, representing the resultant of its positive and negative values. The results presented give grounds for concluding that the external photoelectric effect, photoconductivity (the internal photoeffect), and the photochemical process in silver bromide begin under the action of light simultaneously, i.e., as the result of one and the same primary disturbance.

1. To explain the elementary photochemical process, besides the assumption presented above, another point of view may also be advanced, namely: there is reason to suppose that the elementary photochemical process proceeds according to the following scheme:

\[ \mathrm{Ag}^{+}\mathrm{Br}^{-}+h\nu \rightleftarrows (\mathrm{AgBr}) \longrightarrow (\mathrm{Ag})+(\mathrm{Br}). \]

This scheme is based on the latest investigations of the mechanism of elementary photochemical processes, belonging chiefly to Franck and his co-workers. According to these investigations, optical dissociation of molecules absorbing light is possible. This means that a molecule, as a result of absorbing a light quantum, can decompose without a subsequent collision, which at one time was considered a necessary condition for decomposition. The nature of the decomposition products is determined, among other things, by the nature of the chemical bond. If the molecule is homopolar, then, as Franck showed, it decomposes into atoms, one of which is in the normal state and the other in an excited state. Conversely, a heteropolar (or, more precisely, ionic) molecule always decomposes into normal, unexcited, and uncharged atoms. The mechanism of this decomposition in the case of atomic (homopolar) molecules, according to Franck, consists in the fact that, as a result of absorption, an electron makes a quantum jump, whereby an excited molecule is obtained; if, moreover, under

Under the influence of this excitation, the bonds in the molecule weaken, and the intrinsic vibrations of the nuclei, arising as a result of the displacement of the equilibrium position in the excited molecule, can break the bonds; the molecule dissociates, and in one of the resulting atoms the electron will be in some excited orbit. Thus dissociation will lead to decomposition into normal and excited atoms. In the case of an ionic (heteropolar) molecule, the quantum jump consists in the transition of an electron from the anion to the cation. In this case the bonds, naturally, are sharply weakened, and the molecule can easily be broken by its own vibrations, while as a result of the decomposition normal, i.e. unexcited, atoms are obtained, not ions.

The investigation of the spectra of silver halide salts in the vapor state, carried out by Franck and Kuhn, showed that in these spectra there are continuous regions of absorption, which indicate the presence of processes of optical dissociation. However, this decomposition, according to Franck and Kuhn, corresponds to the formation of a halogen atom in an excited state, i.e. it testifies to the fact that the molecules of the silver halide salts in the gaseous state are atomic molecules, not ionic ones. This circumstance, as the authors mentioned suppose, may be reconciled with the idea that in the solid state the silver halide salts form an ionic lattice: upon transition to the gaseous state the molecule is freed from the influence of the surrounding molecules, which entails such a redistribution of charges in which the character of the bonds changes. If, however, the lattice is ionic in character, then in optical dissociation there should occur a transition of an electron from the anion to the cation, i.e. the liberation of atomic silver should occur according to the scheme indicated above.

In the course of an elementary photochemical process, according to the scheme set forth in §§ 7 and 8, that is, in the presence of an internal photoelectric effect, it would seem that only optical sensitization is possible, consisting in a shift of the wavelength capable of liberating an electron,

and is, in essence, photochemical sensitization. The process expressed by the second scheme is more capable, in what is understood to be photographic sensitization, in the sense of reducing the energy, at the same wavelength, that makes the grain capable of development. Sheppard, who put forward the considerations cited, assumes that photographic sensitization is possible only by concentrating the photoproduct around the nuclei included in the grains, since he believes that, in the absence of explicit optical sensitization, an absolute increase in the amount of photochemically formed product per unit of absorbed energy, in the case of an endothermic reaction, is impossible. The rapid concentration of the reaction product, that is, of silver atoms, around the nucleus eliminates the possibility of reverse and secondary reactions, since, in the absence of sensitizing nuclei, under the action of ordinary light individual atoms are formed which have very little chance of prolonged existence: they can easily recombine, and also, as Clerk suggested, be peptized by the developing solution. Only in the case of X-ray radiation, in the presence of quanta of great power, do many hundreds of silver atoms formed as a result of the action of a single quantum have time to condense and form their own lattice in the form of a nucleus of sufficient size, which then becomes the center of development.

1. The mechanism of the concentration and growth of the nucleus to the dimensions of a development center may be represented in the following way. The inclusion, in the spatial crystal lattice, of a sensitizing nucleus consisting of a foreign substance causes a certain orientation and deformation of the atoms or ions at the interface between two phases; consequently, they are in a certain way disoriented with respect to their own lattice; in the halide ion there appears, apparently, a considerable asymmetry and a strong electric moment, since the excess electron is attracted by the silver ion in one direction, while the remaining part of the ion is attracted in another direction by the lattice of the sensitizing nucleus. The deforming action—

formation of the nucleus must spread, gradually diminishing, over some small distance, and may be represented, according to Sheppard, in the form of the following scheme (Fig. 2), showing a transverse section of the crystal at the location of the sensitizing nucleus; in this scheme the continuous lines denote equipotential regions of lattice disorientation; the dashed lines denote the directions of the potential gradient.

Fig. 2. Distribution of stress around the sensitizing nucleus.

Sheppard further assumes that the absorbed energy in the crystal tends to move along the potential gradients toward the surface of separation between the silver halide lattice and the sensitizing nucleus and to produce there a quantum transformation; if the absorption of light is associated with the transformation of the bromine ion into an excited state, then, for Sheppard’s hypothesis, it is immaterial whether the liberation of an electron and its transition to a silver ion occurs at once, or whether excitation under the influence of the absorbed energy is transmitted in the form of a displacement current from ion to ion until it reaches the boundary of separation, where the silver atom is liberated. Sheppard’s view is in agreement with certain facts of heterogeneous catalysis.

1. If we assume that the sensitizing nucleus consists of metallic silver in contact with silver sulfide, then, evidently, the following heterogeneous system must be regarded as photosensitive:

\[ \mathrm{Ag}—\mathrm{AgBr}—\mathrm{Ag}_2\mathrm{S}. \]

In the occurrence of the internal photoelectric effect, metallic silver may play the role of the cathode, silver sulfide that of the anode, and silver bromide will be the solid electrolyte. Thus, according to this hypothesis, proposed by Trivelli, the sensitizing nucleus will constitute a galvanic couple which, under the action of light, when the silver halide possesses photoconductivity, will produce a true photoelectrolysis of the halide salt. In the indicated system the elementary photocurrent will pass from the anode—\(\mathrm{Ag_2S}\)—to the cathode—\(\mathrm{Ag}\), without

Fig. 3. Mechanism of the formation of the latent image according to Trivelli’s hypothesis.

leaving the limits of the grain, as a result of which the silver cations will move toward the cathode, gradually enveloping the inclusion of silver sulfide, as shown in the accompanying scheme (Fig. 3), and increasing the sensitizing nucleus to the dimensions of a development center. It is possible that, in parallel with the photoelectrolysis of the halide salt, decomposition also takes place of the silver sulfide entering into the composition of the sensitizing nucleus; this latter supposition agrees with the results of the investigations of Tubandt and Tedicke, who showed that silver sulfide behaves as a conductor of the second kind, that is, represents a solid electrolyte in which the cations are mobile. Therefore the photoelectric current passing through the silver sulfide must cause dissociation of the latter and move the silver ions toward the cathode, while the sulfur ions, losing their charge, remain in their places in the spatial lattice. This supposition is confirmed by the circumst—

that desensitization by the action of an oxidizing agent (No. 20 solution of chromic acid) is more effective for an emulsion subjected to preliminary illumination, as compared with the same emulsion without preliminary illumination. In the latter case, after the action of the oxidizing agent, which first of all dissolves the metallic silver, the sensitizing nuclei of the remaining silver sulfide will change their initial dimensions only insignificantly, and therefore the light sensitivity of the emulsion will be greater than after the action of the oxidizing agent in the case of preliminary illumination. The joint formation of metallic silver in contact with sulfide can be represented in such a way that the reduction of the silver halide salt during ripening proceeds more readily at the sites of disturbances formed in the spatial crystal lattice of the grain, owing to the initial formation of inclusions of Ag₂S. Trivelli’s hypothesis, which undoubtedly deserves attention, does not, however, explain the sensitizing action of nuclei of metallic or silver sulfide alone; it rather concerns the following stage in the mechanism of latent-image formation in the case of the presence in the grain of sensitizing nuclei of silver sulfide after a certain amount of metallic silver has been photochemically separated out. Therefore it does not exclude Sheppard’s basic hypothesis, but supplements it, since the most important significance of the sensitizing nuclei apparently consists in producing deformation in the lattice and in orienting the elementary photochemical process in the grain. It is very probable that the effect of increasing light sensitivity by means of preliminary brief illumination can also be explained by the formation of elementary galvanic couples, as a result of the photochemical separation of metallic silver in contact with the initial nuclei of silver sulfide. Therefore one may expect that the effectiveness of preliminary illumination will be considerably greater in emulsions with a large number of grains sensitized by nuclei of silver sulfide.

1. The presence of ionic deformation near the surface of the disـ

for bromide silver and a sensitizing nucleus theoretically should cause a shift of the sensitivity threshold toward the region of the spectrum with longer wavelengths; this circumstance constitutes a certain difficulty with respect to the hypothesis of the orientation of the photochemical process and the condensation of metallic silver near the sensitizing nuclei. However, Sheppard believes that the principal feature of sensitization by the discrete inclusion of a foreign substance in the spatial lattice is that this sensitization may have the character not only of photographic sensitization, that is, may cause a more rapid compaction of the photochemically formed silver atoms and thereby prevent these atoms from recombining back with the halogen, but also of optical sensitization, shifting the threshold or changing the distribution of spectral sensitivity—everything depends only on the degree of dispersion of the sensitizing nuclei. If the emulsion grain contains a large number of highly dispersed nuclei of metallic or silver sulfide distributed over the surface, then, owing to the relatively increased number of strongly deformed ions, the effect of optical (photochemical) sensitization will be correspondingly greater, while the effect of concentration will diminish as a result of a kind of competition among a large number of sensitizing nuclei. In the limiting case, when there is nearly atomic or molecular dispersion of the nuclei, the phenomenon of a photoelectric effect will be observed in the region of the frequencies of the intrinsic absorption of the foreign substance distributed over the surface. Conversely, with a small number of large nuclei, the effect of concentration of the photochemically formed silver atoms and the growth of the nuclei to the size of development centers will be dominant, since around such nuclei a large region of ionic deformation will be observed, and the reduced competition will cause more rapid growth of the nuclei.

1. The explanation given of the role of sensitizing nuclei is very important, since it reconciles, to

of a certain degree, two opposite points of view on the question of whether the presence of nuclei in emulsion grains permits an absolute increase in the amount of photochemically decomposed silver halide by a quantum of absorbed energy, or not.

Toy inclines toward an affirmative answer, since he considers that the sensitizing nuclei themselves are photosensitive. Sheppard, on the contrary, on the basis of the laws of quantum transformations of energy, adheres to the opposite view and ascribes to the nuclei only a directing significance in the elementary photochemical process and the role of coagulation centers. The hypothesis concerning the dependence of the behavior of sensitizing nuclei on the degree of their dispersity, reconciling both points of view, seems to us very probable and may be confirmed to a certain extent by experimental data. Frankenburger, in his study of the spectral sensitivity of silver bromide and of the influence upon it of adsorbed substances, showed that silver bromide in the presence of excess AgNO₃ exhibits a considerable displacement of the sensitivity threshold toward the red rays (toward \(\lambda \geq 620\,m\mu\), whereas in the case of pure AgBr the sensitivity threshold lies between \(\lambda = 435\,m\mu\) and \(\lambda = 410\,m\mu\)); the author of the study explains this by the fact that adsorbed silver ions on the surface of the AgBr lattice are, from electrostatic considerations, more active acceptors of electrons than the surface silver ions properly belonging to silver bromide. The adsorption of OH⁻ ions also affects the displacement of the sensitivity threshold, but in this case a different photosensitive system is obtained, since the electron is transferred to the silver ion, apparently, not from the bromine ion, but from another anion (OH or O⁻⁻).

1. In the study on the standard source of light for sensitometry, reported by the Commission of the American Optical Society at the VI International Photographic Congress (Paris, 1925), there are very interesting data which may also serve as confirmation of the hypothesis considered. As is known, the magnitude of the photosensi-

...sensitivity, determined by the appropriate method, depends on the light source used, that is, on the spectral composition of the flux of radiant energy emitted by it, which, in turn, is determined by the color temperature of the source. If sources with high and low color temperature are used, then especially large discrepancies are obtained with emulsions containing optical sensitizers, that is, photochemically active dyes adsorbed by the emulsion grains; the introduction of such dyes into a photographic emulsion causes a considerable change in the spectral sensitivity, since the presence of adsorbed molecules on the surface of the grain gives rise to a photoelectric effect in the spectral region corresponding to the dye’s own absorption band. But it turns out that ordinary emulsions as well, having natural color sensitivity to blue, dark-blue, and violet rays, and also to ultraviolet rays, which, however, are absorbed when working with a glass optical system, exhibit a different distribution of spectral sensitivity in this region, even if they contain only silver bromide alone (without an admixture of iodide; different amounts of the latter could cause a nonuniform distribution of spectral sensitivity), as is evident from the following data:

In these experiments with emulsions of identical composition, containing only silver bromide, a very probable cause of the fluctuations of the coefficient \(K\) may be the unequal degree of dispersity of the sensitizing nuclei, which is a consequence of different emulsification conditions, since the influence of quantitative composition is excluded here.

1. Similar conclusions may be reached as a result of studying the chemical sensitization of a finished emulsion by immersing it in ammoniacal solutions of various silver salts. A systematic investigation in this direction is being conducted by the author of the present communication together with Mikhailova; it is true that we have not yet obtained convincing evidence of a shift of the sensitivity threshold toward rays of longer wavelength, owing to the insufficient accuracy of existing methods for the quantitative determination of spectral sensitivity and to the complex adsorption phenomenon observed here, but the results obtained indicate a change in the spectral distribution of sensitivity in the region of the emulsion’s natural sensitivity under the influence of various silver salts adsorbed on the surface of the grains. The method of investigation consisted in sensitizing plates by immersion in very dilute ammoniacal solutions of various silver compounds and in determining the relative photosensitivity, once from a light source with a color temperature reduced to \(5000^\circ\ \mathrm{K}\) (with the aid of a light filter), and a second time from the same source after changing the composition of its radiant-energy flux by placing a light filter that transmits rays in the region of the emulsion’s natural sensitivity; the nonconstancy of the ratio of the photosensitivity values should indicate a nonuniform distribution of spectral sensitivity, which in some cases may be explained by the formation of an electrostatic double layer on the surface of the grains, and in others, possibly, by ionic adsorption or substitution. The data obtained by us are given below:

In the table the results are arranged in order of increasing fog upon sensitization; it is interesting to note a certain parallelism in the change of the fogging capacity and the relative magnitude of the light sensitivity, which was to be expected, since it has been established experimentally that emulsion grains having a greater tendency to develop without the action of light upon them are also the most light-sensitive. Here it is also important to note the appreciable influence of chemical sensitization on the value \(\gamma\) (the gradient \(\dfrac{dD}{d\lg E}\), where \(D\) is density, \(E\) is the quantity of light), which must likewise be explained by a change in spectral sensitivity, since \(\gamma=f(\lambda)\).

1. Recently Hickman has put forward a supposition and made an attempt at experimental demonstra-

...of the chemical role of the nuclei of AgS₂. This hypothesis, while not contradicting thermodynamic considerations, consists in the fact that the sensitizing nuclei can chemically bind the liberated bromine atoms and, as a result of this reaction, release metallic silver, thereby increasing its quantity for the formation of the development center. However, Sheppard showed that the use in a photographic emulsion of thioanilide, which is a bromine acceptor, does not cause an increase in light sensitivity. Consequently, the chemical binding of bromine is not unconditionally necessary for the effect of photographic sensitization. But the hypothesis advanced compels one to give due attention to the question of the removal of bromine atoms and of the influence of reverse dark reactions.

To complete the consideration of the nature of the latent image, it is of some interest to touch upon the question of the magnitude of the altered sensitizing nucleus which, in the process of development, plays the role of the center from which the reduction of silver halide begins. Direct determinations of the sizes of development centers are not available at present; therefore only the order of this magnitude can be given, using analogies. It has been established that, for the onset of the reduction of gold from the corresponding reaction mixture, the introduction of a germ with a mass of not less than \(10^{-16}\) mg is necessary, containing about 300 atoms of gold; it is possible that the development centers must also have approximately the same order of magnitude.

Photochemical Properties and Development of Individual Emulsion Grains

1. In the volume corresponding to each \(cm^2\) of the surface of a light-sensitive layer, having a thickness of about \(0.02\) mm, there are from 50 to 300 million emulsion grains of various sizes, arranged in the form of a large number of stratifications (in cross section the emulsion layer may have up to 100 elementary layers of grains). From the properties of these grains

as independent units participating in the formation of the image, the properties of the photographic emulsion as a whole are composed. Therefore, for elucidating the mechanism of the action of light in photography and the physicochemical picture of the development process, the microphotographic study of the behavior of individual emulsion grains is of extremely great importance. Despite the obvious importance of this question, it has become a subject of investigation only in recent years.

If an emulsion grain is subjected to strong illumination, it begins to darken as a result of the decomposition of silver halide and the separation of metallic silver. The actual formation of metallic silver was proved by the X-ray spectrographic method of Debye–Scherrer, thereby casting great doubt on the existence of subhalide compounds; they must rather be regarded as solid solutions or as adsorption compounds of silver halide with metallic silver; the colored shades in these cases are explained by the varying degree of dispersion of colloidal metallic silver.

1. The photochemical decomposition of silver halide proceeds in the crystal not uniformly over the entire illuminated surface; it begins at individual points located in the grain for the most part according to the law of chance. We investigated the decomposition of a large number of individual grains (from 20 different sorts of plates and films), and no signs whatever were found of any geometrical regularity in the localization of metallic silver, as is clearly seen from the accompanying typical microphotographs of this process (Table I, Fig. 4). However, in some comparatively rare cases, a geometrical regularity of the photochemical decomposition was observed.

The phenomenon of geometrical regularity in the decomposition of silver-halide crystals was first discovered and described by Trivelli and Sheppard; such crystals were obtained under special conditions from an ammoniacal solution, in the absence of gelatin. This latter circumstance is apparently the reason for the manifestation of geometrical

regularities: the absence of gelatin created conditions for the proper growth of the crystal. Further investigations by the authors cited led to the discovery of isolated cases of semi-regular photochemical decomposition of emulsion grains. We likewise found a number of emulsions in which grains occurred in large numbers that exhibited geometrical regularity in the localization of the metallic silver formed photochemically. Such grains were usually small in size (of the order of 1 μ), which apparently is one of the reasons for the preserved geometrical regularity (Table II, Fig. 5). It is interesting to note that, with short illumination and subsequent short development of such grains, no geometrical regularity is observed in the deposits of metallic silver, since development begins from centers located on the surface of the grain and subsequently proceeds differently from the visible photochemical decomposition.

In the photochemical reduction of silver halide, in contrast to development, the entire crystal as a whole is of much greater importance, i.e., the internal interrelations of the silver and halide ions arranged according to a definite law in the spatial crystal lattice, and not merely the outer surface of the crystal. This feature is demonstrated by the desensitizing action of oxidizing agents: treatment with a solution of an oxidizer almost completely destroys the light sensitivity of an emulsion, measured in the usual way after development, but this chemical action has no effect at all on the rate of visible photochemical decomposition. This difference is explained by the fact that the action of oxidizers consists in the destruction of surface sensitizing nuclei, which exclusively determine the high light sensitivity of photographic emulsions, but are not at all necessary for the visible photochemical decomposition of the grain.

1. The process of development of a briefly illuminated grain, which from the chemical point of view represents the reduction of silver halide to metallic silver, begins at separate points, centers of development, and the picture

at its initial stage fully coincides with what is observed in visible photochemical decomposition; subsequently, however, a considerable peculiarity of the process of development becomes apparent. Soon after the beginning of development the grain begins to send out sprouts of metallic silver and, by the end of development, completely loses its original crystal form (Table II, Fig. 6, and Table III, Fig. 7).

The diminution and irregular outlines which the emulsion grain assumes during development may indicate the gradual dissolution of the silver halide. Consequently, the mechanism of development of individual grains may then be represented as follows: owing to the presence in the developer of substances dissolving the silver halide, part of it, in a very thin layer around the grain, passes into solution, and here the silver ions are reduced to metallic silver in the colloidal state; as it forms, it is deposited around the centers of development, which play the role of a seed for the beginning of the coagulation of metallic silver; as a consequence of coagulation, the state of equilibrium expressed by the scheme is disturbed:

\[ \mathrm{AgBr}\ (\text{solid}) \rightleftarrows \mathrm{AgBr}\ (\text{dissolved}) \rightleftarrows \mathrm{Ag}^{+} + \mathrm{Br}^{-} + \mathrm{R} \rightleftarrows \]

\[ \rightleftarrows \mathrm{Br}^{-} + \mathrm{R}^{-} + \mathrm{Ag}\ (\text{colloidal solution}) \rightleftarrows \mathrm{Ag}\ (\text{solid}); \]

therefore the process proceeds in one direction and ends when the grain has been completely developed. If the grain has not been altered by light and there are no centers of development on its surface, then the gelatin protects the silver particles from coagulation, and the grain remains undeveloped owing to the rapid establishment of equilibrium; it must be noted, however, that the protective role of gelatin is limited to a certain time, and with very prolonged development there will occur a gradual coagulation of the colloidal silver particles up to the size of a center of development; then the unexposed grain will also prove to be developed; grains developing in this way also form a veil. Consequently, the selective character of reduction by the developing solution of silver halide in the places of formation-

the latent image that has appeared has to be explained by the presence of gelatin as a protective colloid, since in its absence unexposed silver halide layers are also reduced. On the other hand, the developer must have such a reduction potential that, within the intervals of time used in practice for developing the latent image, it does not go beyond the limits of the protective action of gelatin, i.e., does not cause the formation of an excessively large fog.

1. In 1920, Sheppard advanced another hypothesis, based on the phenomenon of adsorption of the reducer of the developing solution by the silver halide of the emulsion grains, which may be expressed by the following scheme:

\[ \frac{R^-}{\mathrm{Ag}+\mathrm{Br}}\quad \frac{R}{\mathrm{Ag}+\mathrm{Br}}\quad \frac{R^+}{\mathrm{Ag}+\mathrm{Br}}, \]

in which a part of the surface of the emulsion grain is represented; as a result of the subsequent rearrangement of the complex \((\mathrm{Ag}R)^-\), metallic silver is formed and the emulsion grain is gradually destroyed from the surface.

The role of nuclei of metallic or silver sulfide of definite sizes may be explained, from the point of view of this hypothesis, by an increase in the reactivity of silver ions at the points of contact of the nucleus with the silver halide surrounding it, owing to deformation of the electron orbits. This hypothesis satisfactorily explains the phenomenon of pseudomorphic replacement of crystals of silver bromide by metallic silver, which, according to X-ray analysis data, has likewise a crystalline structure, in the case of the use of developing solutions with a low reduction potential (for example, hydroquinone). A certain difficulty for this hypothesis is posed by the explanation of the peculiar changes in the shape of the emulsion grains, which seem to indicate the gradual dissolution of the grain and are accompanied by “explosions” and the formation of “protuberances” of metallic silver. The author has observed, however, a phenomenon which may also be regarded as confirming Sheppard’s hypothesis:

this phenomenon consists in the catalytic acceleration of the photochemical reduction of an emulsion grain that has previously been briefly exposed and treated with a developing solution; in this case, after the indicated preliminary treatment and drying of the layer, the microscopic picture under illumination with white light coincides extremely closely with what is observed with emulsion grains when the developing solution acts on them under illumination with non-actinic light, that is, “explosions” in the grain and the formation of “protuberances” are also observed (Table III, Fig. 8).

It may therefore be supposed that the phenomenon of catalytic acceleration is connected here with adsorption on the surface of the grain of molecules of the reducing agent, all the more so since the process of rapid photochemical decomposition gradually dies away. Thus, in the process of development one may assume two stages: 1) complex formation between silver halide and the molecules of the organic reducing agent, and 2) an internal rearrangement of the complex molecule, accompanied by the liberation of metallic silver, the oxidation products of the organic substance, and soluble bromide. As for the question of whether the process of internal rearrangement takes place in situ, in the adsorption layer of the grain, or whether the complex molecules pass into solution in the thin layer around the grain, here, it seems to us, it is impossible to draw a sharp boundary between the two cases.

1. An even more perfect method for studying the mechanism of development is cinematographic filming.

To carry out this method it is necessary, simultaneously with filming, to follow the process of development visually. For this purpose a special eyepiece may be used, which Tetel and Trivelli used in their work. In the absence of such an eyepiece, the author, who studied the development process together with Goppe by this method, had to proceed somewhat differently: first we followed the course of the process over time, having established in a preliminary experiment the moment of the beginning of development and its duration; subsequently we projected

MECHANISM OF FORMATION OF THE PHOTOGRAPHIC IMAGE

the image cast by the microscope, in such a way that part of it fell into the aperture of the motion-picture camera, while the other part fell upon a screen placed around the aperture; from the image on the screen it is very convenient to follow the process of development of the grains. Photography was carried out on orthochromatic film with the use of a green monochromatic light filter, and on panchromatic film with the use of a red light filter. For the work, objectives with water and oil immersion were used; in the first case a drop of water was placed on the slide bearing the applied smear of emulsion, and after focusing, into the drop in which the front lens of the objective was immersed, a drop of developing solution was introduced from a pipette. After some time, when the solution, owing to diffusion, reached the layer, filming began. In the second case the smear of emulsion was applied to a cover glass, which was placed on strips of glass cemented with Canada balsam to the slide; thus, between the cover glass and the layer of emulsion applied to it and the slide, there was a space approximately \(1/2\ \mathrm{mm}\) thick, which served as a kind of cuvette filled with the developing solution. Oil immersion was effected between the slide and the condenser and between the front lens of the objective and the cover glass.

The use of green illumination, of low actinic power with respect to the emulsion under study, and of red, non-actinic illumination, had the advantage that in photographing, the actual process of development was studied. The American investigators carried out photography in blue rays, and therefore two processes proceeded simultaneously in their work—the photochemical decomposition and the development of the grains, which were superposed one upon the other, as was shown by us in experiments on the catalytic acceleration of the process of photochemical decomposition of a grain. In Fig. 9 (Plate IV) an enlarged print from individual frames of the motion-picture film is presented.

A particularly striking picture is observed when projecting onto a screen: here, at an enormous magnification, the entire

the picture of development comes to life: during development the grains, it turns out, do not remain at rest; at times they are shaken; from individual areas of the grains “protuberances” of metallic silver are thrown out; the very rapid formation of metallic silver in the presence of great resistance, due to the elasticity of the gelatin, apparently serves as the cause of the shaking of the grains.

At present we are conducting further investigations in this direction, with the aim of discovering the individual features of the action of various developing substances and of the conditions of development in connection with the question of obtaining fine-grained deposits of metallic silver.

In conclusion to the present essay it is necessary to note that if in recent years major achievements have indeed been made, somewhat clarifying the mysterious mechanism of the action of light in obtaining a photographic image, photography owes the latter exclusively to the progress of scientific knowledge in the field of physicochemical disciplines. Thanks to the successful application of this knowledge and of refined methods of scientific research to questions of photography, it has been possible to obtain results which, in their grandeur in the microcosm, may seem so extraordinary in comparison with the rather simple templates used by the photographer to obtain an image.

But what has been obtained is insignificant in comparison with those questions which, as we move forward, arise in still more enigmatic obscurity; and there is no end to the path that opens as an immense horizon before the eyes of the investigator.

Literature

1. Eder J. M. Ausf. Handb. Phot. II. 1: Lüppo-Cramer. D. Grundlag. d. photogr. Negativverf. (1927).

2. Trielli A. P. H. & Sheppard S. E. The Silver Bromide Grain of Photogr. Emulsions. N. Y. (1921).

3. Bullock. Chemic. Reactions of the Photogr. Latent Image. N. Y. (1927).

4. Sheppard S. E. & Mees C. E. K. Investigat. on the Theory of the Photogr. Processes. London (1907).

5. Nietz A. H. Theory of Development N. Y. (1922).

6. Ross F. E. The Physics of the Developed Photogr. Image. N. Y. (1924).

7. Lüppo-Cramer. Kolloidchemie u. Photographie, (1921).

8. Sheppard S. E. The Theory of Photogr. Processes and Methods. Photogr. as a Scientific Implement. London, p. 103–209 (1923).

9. Sheppard S. E. The Formation of the Photogr. Latent Image. Phot. Journ. 68, 397 (1928).

10. Sheppard S. E. & Vanselow. W. The Lattice Energies and Photochemic. Decomposit. of the Silver Halides. Journ. Phys. Chem. 33, 250 (1929).

11. Vanselow W. and Sheppard S. E. Photovoltaic Cells with Silver-silver bromide Electrodes. Journ. Phys. Chem. 33, 331 (1929).

12. Rawling S. O. Recent Advances in our Knowledge of the Latent Photogr. Image. Phot. Journ. 69, 471 (1929).

13. Toy F. C. D. Mechanismus d. Entstehung d. latent. Bildes. Ztschr. f. Wissensch. Photogr. etc. 27, 85 (1929).

14. Trivelli A. P. H. Essai d’hypothèse sur l’image latent I, II. Science & Industr. photogr. 8^m, 8, 14, (1928).

15. Slater-Price, T. La sensibilité photographique. Sc. Industr. phot. 9^m, 39, (1929).

16. Bancroft, W. D. Les théories de la photographie. Sc. & Ind. photogr. 5^m, 81 (1925).

17. Chibisov K. Physico-chemical interpretation of the photosensitivity of photographic emulsions. Journal of Applied Chemistry. 1, 3, (1928).

18. Shpolsky E. Mechanism of elementary photochemical processes. Advances in the Physical Sciences 7, issue 6 (1927).