Many different types of flicker photometer have been developed. Of the more accurate, the one shown in fig. io is, perhaps, the simplest in principle, and will serve to illustrate the method. S is a white surface illuminated, by way of the total reflection prism P, by the light from one of the sources to be compared. W is a sector disc, which can be rotated at any desired speed about a horizontal axis. Its surface is white and is illuminated by the other source of light so that, when viewed by the eye at E, the field of view seen through the small aperture A is alternately oc cupied by W and by S. A is of such a size as to subtend an angle of 2° at E. It is cut with a sharp (back-bevelled) edge in a con cave surface F, which is evenly illuminated by the small lamp L. L is adjusted to give F a brightness of about 8 candles per square metre, and the distances of the sources to be compared are then arranged so that the brightness of the field at A has approximately the same value. The test lamp and photometer are then fixed and measurements are made by moving the comparison lamp, the speed of rotation of W being reduced until flicker can be made almost to disappear. Settings of the comparison lamp are then made to the point of minimum flicker. It will be noticed that the criterion of equality used in this instrument is not dis appearance of flicker but minimum flicker, as it has been found that in practice this gives a rather more sensitive test of the balance point.
The instrument used for determining both the above sets of data is called a spectrophotometer, since it is essentially a device for enabling the intensities of two lights to be compared at any given part of the spectrum instead of as a whole. Every spectro photometer, therefore, consists of two parts, (i.) a spectrometer (see SPECTROSCOPY) for enabling any given wave-length interval of the lights to be isolated, and (ii.) some device for causing the portions of the lights so selected to illuminate the two parts of a comparison field, and a means for altering the brightness of one part of this field.
The principle will be made clearer by means of an example. Fig. II shows diagrammatically the Lummer-Brodhun instrument which consists of an ordinary spectrometer with the addition of a second collimator C2 and a Lummer-Brodhun cube. The latter is constructed similarly to the cube used in the ordinary photometer head (see p. 841) but the form of the field is that shown in fig.
II. The two sources to be compared respectively illuminate the slits of and C2 which may be covered with diffusing glass. For the adjustment to a photometric balance various devices may be used, e.g., a variable slit in one collimator, alteration of the distance of one source from its slit, a sector disc or neutral wedge in one beam, or a pair of nicol prisms (one capable of rotation through a measureable angle) mounted in one of the collimators. The last mentioned device is used in the Brace-Lemon instrument which is a modification of the Lummer-Brodhun spectrophotometer, the dispersing prism being used also to form the comparison field. This prism is shown in fig. II. As in ordinary photometry the substi tution method is generally employed. The source used as a standard of spectral distribution is frequently a tungsten filament vacuum lamp, since the spectral distribution of this type of lamp is very closely determined by its efficiency. Alternatively any suitable lamp may be used, the spectral distribution of which has been determined at a standardiz ing laboratory.
One of the most important uses of spectrophotometry is to determine at each part of the spectrum the transmission factor of a coloured medium, a chemical solution, etc. This may be done quite readily by first obtaining a balance for two sources of light at any one wave-length, and then inserting the coloured medium between one of these sources and the photometer. The amount by which the intensity of the other beam has to be altered to restore the balance enables the transmission factor of the medium at the particular part of the spec trum isolated to be reduced.
Physic al Photometers.— Many attempts have been made at various times to develop some physical instrument which can be used instead of the eye for making photometric measurements. The instruments proposed may be divided, roughly, into three classes. The first of these includes the vast number of chemical photometers in which the illumination is measured by the rate of change it produces in the constitution of some chemical mix ture. This form of photometer survives in the photographic method used in some branches of spectrophotometry and in stellar photometry (see below). The chief disadvantage of the method is, as in all physical photometry, that the response of the sensitive substance to equal rates of energy reception in the form of light of different wave-lengths is quite different from that of the eye. other words, the luminosity curve of the physical photometer is not approximately the same as the curve of fig. 1. This, i clearly, is of no importance in spectrophotometry where the com parison is made separately at each part of the spectrum. The principal difficulty associated with photographic spectrophotom etry is the lack of proportionality between exposure (illumination X time) and blackening. Schwartzchild's law that, for equal photographic densities, EtP is a constant (p has a value between 0.75 and 1 depending on the plate) holds for short ranges of E and t.