In the case of keeping for a very long period between exposure and development, especially in a warm and moist climate, other phenomena become manifest : progressive fog and, usually, retrogression of the latent image, this retro gression being sometimes fairly rapid with films owing to the chemical action of the products of spontaneous decomposition of the support.
201. The Law of Density. In order to set forth the accuracy in the rendering of different luminosities in the photographic process, we will consider, not a landscape or any other usual subject, where the luminosities are distributed at hazard, but a scale in which the same range of luminosities is arranged in increasing order, and preferably a translucent scale which can be printed by contact on the sensitive film to be tested in such a way as not to introduce any causes of error due to the lens. Let us choose, for example, a neutral grey prismatic screen (wedge), forming a graduated scale of tones such as that represented in section (but with the height very exaggerated) by Fig. 139A.
If we consider several points abcd equidistant from one another along the straight edge of the prism, we see that the thicknesses of absorbing matter at these points form an arithmetical progression; 3 the quantities of light transmitted at these different points form a gecinetrical progression'.
Take the case of several identical absorbing layers arranged so as to form a scale of graded tints. Suppose that each layer transmits half of the light which reaches it. The quantities of light transmitted by the successive steps will be as shown in the table, if we represent by unity the value of the originally incident light.
This fact is generally expressed by saying that the quantity of absorbing material (the thickness of the wedge in the example we are considering ; the mass of reduced silver per unit area, in the case of a photographic negative) is proportional to the logarithm of the opacity. The name optical density is given to the logarithm (to the base io) of the 202. If we could obtain a negative in which the silver was distributed as shown, on a con venient scale, in Fig. 1393, then the " wedge " and the negative, if superposed, would make a series of uniform densities, this negative allowing us to obtain, by a second print made under the same conditions, a perfect facsimile of the " wedge."
Actually, however, it is known that whatever kind of sensitive film be used for one or other of the prints, the ideal result is never attained. Under the most favourable conditions the dis tribution of reduced silver in the negative may be like that shown in Fig. 139C. Another copy made under identical conditions will not give a facsimile of the original screen but the irregular scale of tones indicated in Fig. 139D.
The curve in Fig. 139C is called the charac teristic curve of the 203. Numerical Expression of the Sensitivity of Photographic Emulsion. After attempts had been made to reduce the photographer's testing work by dividing the various emulsions into a number of arbitrary classes (one of the most fantastic of these empirical systems was that due to Warnerke), endeavours were made to denote each emulsion by a numerical value expressing some characteristic of this emulsion.. The different Austro-German sensitometric sys tems (Scheiner, Eder, etc.) are based on the measurement of the least quantity of light (corresponding with the threshold of sensitivity), which, after development under constant condi tions (which unfortunately is not always equiva lent to development carried to the same degree), produces on the emulsion an image which can just be detected in comparison with the region which has been completely protected from the light. The systems of English origin, which are all derived from the sensitometric method of Hurter and attempt to define the sensitivity of the emulsion by the position of the point A (Fig. 140), at least when this point is situated on the log E axis. These different methods of numerical expression, characterizing as they do different properties of the emulsion, have no common connection between them, and it is impossible to obtain precise equivalents of one system in another. The most that can be done is to indicate the approximate relation between the different methods of expression.