A FLICKER PHOTOMETER ATTACHMENT FOR THE LUMMER-BRODHUN CONTRAST PHOTOMETER.*

BY

E. F. KINGSBURY,

Physical Laboratory, The United Gas Improvement Company, Philadelphia.
Member of the Institute.

EQUALITY-of-brightness photometers and flicker photometers have heretofore been built entirely separately from one another. In the literature of the measurement of light the design and use of the one has been a distinct problem from the design and use of the other, though the results the former gives are often used as a criterion for the practicability of the latter. Consequently, in constructing a flicker photometer, it has been built from the beginning just for the one purpose, notwithstanding the fact that about half of a flicker photometer is identical with an equality-ofbrightness photometer in its essential requirements. It should be possible, by interchanging the two dissimilar parts on the part common to both, to make either one easily converted into the other type of photometer. The result should be a photometer that is simple, compact, and capable of measuring either lights of one color or lights of widely different colors with the best precision of the equality-of-brightness method in the one case and of the flicker method in the other.

At the same time it is felt that such an instrument should embody the latest approved features that are coming to be recognized as desirable in a flicker photometer.

It is the purpose of this paper to describe such an instrument which the author has devised in the shape of a flicker attachment to replace the telescope of the ordinary Lummer-Brodhun contrast photometer when it is desired to measure differently colored lights. The high precision and the wide distribution of this type of equality-of-brightness photometer justify its selection for this purpose, although there are other good instruments to which this idea could be even more easily applied with a slight modification of details.

The various details concerning this attachment will be better appreciated if, before describing it in detail, we briefly review the essential requirements of a flicker photometer.

* Communicated by the Author.

VOL. CLXXX, No. 1076—16

215

1

Wilde says the desiderata for an ideal photometer are:

1. It must be sensitive and certain on lights of different color. 2. Rays must fall upon the illuminated surface perpendicularly to reduce angle errors.

3. Illuminated surfaces should be directly over pointer on

the bar.

4. When balance is nearly obtained, it should be easy to perceive which way the head must be moved to improve the balance, without the necessity of rocking it to and fro.

5. If complete reversibility is required, both illuminated surfaces should be identical and must be the same distance from the eye.

6. In a flicker photometer, in order to secure sensitiveness, the transition from one surface to the other must be sudden and there must be no dark line crossing the field of view, such as an unilluminated edge of a card.

These requirements are subject to considerable modification and addition. They are given, however, as illustrating a certain viewpoint in the construction and use of flicker photometers and to emphasize the employment of a proper photometric procedure.

The use of the substitution method of photometry is strongly recommended with the direct measurement of distances wherever necessary.

sary.

This makes the second, third, and fifth requirements unneces

The fourth requirement is misleading, because one cannot obtain a satisfactory balance with the flicker photometer by merely looking at it in one position. The proper method of obtaining a setting is rapidly to vary the light intensity back and forth over the no-flicker peak, at the same time gradually narrowing the range until the point of minimum or no-flicker is found. The essential requirements only will be outlined below. They will not be discussed in detail, as they have been fully covered elsewhere.2

The sensitiveness or precision requirements of a flicker photometer depend fundamentally upon one thing-the focusing of the observer's maximum attention upon the elimination of intensity flicker. The attention is secured in a rather negative way by eliminating all external factors that can divert it.

1London Illuminating Engineer, p. 825, i, 1908.

2 Phil. Mag., July-December, 1912, pp. 149, 352, 744, 845.

First, undue tiring of the eyes should be guarded against. This is secured mainly by making the field surrounding the flickering field light instead of leaving it dark. The intensity of this illumination should be equal to, or less than, the intensity of the comparison field. If brighter, it attracts the attention. The intensity and color of this field do not seem to influence the accuracy, at least within wide limits. This field should be illuminated uniformly and should be free from mechanical defects.

Secondly, there should be no mechanical flicker in the comparison field itself. This is secured by keeping the optical system clean and making its focus in space and not on a surface. The focus of the eye-piece lens, of course, should be on the plane of the opening through the external field just referred to.

Thirdly, in securing a balance, the color flicker (as distinguished from the intensity flicker) should be reduced to a minimum by a proper regulation of the speed of alternation of the colors.

The requirements that make for agreement with the equalityof-brightness method are in each method:

First, a restricted field; and

Secondly, a high illumination on this field.

Dr. H. E. Ives has found that on a field size of about 2° and at an illumination of about 25-metre candles the flicker method and the equality-of-brightness method agree in their results, provided the same conditions are complied with in each case. A restricted field and a high illumination are somewhat interchangeable, but the candle-power of many standard lamps and other light sources places a limit on the illumination, so that the field is restricted to make it equivalent to a higher illumination.

The requirements having to do with procedure have been mentioned; namely, that the substitution method be used with the direct measurement of distances. Where it cannot be done, care must be taken to reverse the comparison head and also the lamps under test, unless complete reversibility of each has been shown to yield identical results.

The method of securing a balance with the flicker photometer mentioned above should be emphasized here. The relative intensities of the two lights under test should be capable of being easily and quickly varied over a considerable range. One method of doing this is shown below.

In making a setting the light should be varied back and forth

over the region of minimum or no-flicker, and the limits gradually narrowed down to a balance. It is impossible to obtain good precision or accuracy by leaving the setting at any one point and trying to decide whether it is the correct point. There is no criterion for the judgment unless the intensity is changed.

The last general requirement is one of convenience. The parts

should be easily accessible for cleaning, repairing, and demonstration.

The attachment, with the motor and a neutral tint-absorbing screen, is shown in position on a Lummer-Brodhun head in Fig. 1. Fig. 2 shows an elevation of the same.

The attachment (A) screws into the head in place of the regular telescope. The only other change necessary in the Lummer-Brodhun is the removal of the contrast glasses, which

the heads are built to permit readily. A belt (H) runs from the rotator within the tube to the motor (M), which is fastened to a clamp (F) that fastens to the horizontal casting beneath the box. The belt (H) is made of spirally-wound steel wire. It acts

Elevation of Lummer-Brodhun with flicker attachment, motor, and neutral tint-absorbing screen, showing the method of clamping it to the head.

as a spring to take up any slack, and allows of the belt being easily slipped off the motor pulley. The part shown to the right is an addition, proposed by Dr. H. E. Ives, to facilitate reading, in the shape of a neutral tint-absorbing screen formed of parallel wires