by Helmholtz and Abney. It can be used for many, if not most, of the purposes for which a color-mixer is needed. It consists of a diaphragm D1 the opening in which serves as an object to be imaged in the mixed colored light and as a guide for definiteness of focussing in the formation of that image; a double convex lens Lз to form the image; and a diffusing surface of a high coefficient of reflection, e.g., magnesium oxide or magnesium carbonate, to receive the image. The diaphragm D1 may be placed just back of the objective lens L2 of the spectroscope, or even just in front of it; the focussing lens. L3 just back of the objective slit OS1; and the screen S in the image plane of the diaphragm. When this is done an image of the diaphragm is formed in light uniformly mixed from the two, three or four slits as the case may be, the size of which will depend on the diopter value of the lens used and the distance of the diaphragm from that lens. As already indicated, the proportions of the components entering into the mixture may be regulated at will by means of the control plates, provided to vary the heights of each of the slits. Obviously these plates may be used also for the control of the intensity of the mixed beam of light, i.e., the components can be mixed at any level of intensity from full to extinction. To facilitate the use of the plates for this and other purposes in the control of intensity, each plate has been provided with a scale and micrometer screw. Also, if desired, other means may be used for the control of the intensity of the mixed beam of light, e.g., a sectored disc; two filters of exactly similar gradations of transmission, travelling in opposite directions across the beam of light; an iris diaphragm, with a slight modification in the optical system; the collimator slit; etc.

If it is desired to color-match or to spectrocolorimeter the mixed color obtained, a simple photometer head, such as the Ritchie wedge W, or a more complicated one, such as the Lummer-Brodhun cube supplied with an eye-piece, may be used to receive the lights to be compared; and a second or auxiliary spectroscope similar to the first, to furnish the standard color. In order that an image of the same size and shape may be obtained for the unmixed spectrum color, this

second spectroscope may be provided with an optical system, consisting of a diaphragm D2, a slit OS, and a lens L4, similar to that used for the color mixer, and with a control plate for the variation of the intensity of light. It may be supplied either with a single slit and control-plate or, if it is desired to add a variable achromatic component, the energy of which may be specified as well as the energy of the standard color, three slits may be provided, two upper slits spaced for a complementary pair of colors with control plates to vary the intensity, and a third lower slit to supply the standard color.

If it is desired to use the mixer as a colorimeter by the monochromatic, two-, three- or four-color methods, the opposite surface of the Ritchie wedge may be used to receive the unknown color. When so used, as already indicated, an achromatic component formed by the combination of complementary colors, the energy specification of which can be made, may be added when the monochromatic or two-color methods are used; and by reflecting white light into the beam from the surface of the prism, when the three- or the four-color methods are employed.

It is quite obvious that the optical system just described can not be used for focussing the light on the eye as a method of securing a color-field, even though the light emerging from the image formed by the mixing lens were collimated and focussed on the eye in accord with the usual method of so presenting the light. When this is done the image is formed within the refracting system of the eye, theoretically at the optic center, and the rays diverge and tend to separate out again before reaching the retina, presenting to the view of the observer a surface in the mixed color surrounded by fringes of the component colors. Plans are under way for utilizing slits of the type noted above in combination with the method of focusing on the eye, but they will not be discussed until they are satisfactorily completed.

While the instrument described has been devised to meet the needs of the work in our laboratory and to conform to its resources and equipment, there are some advantages of a more general purport which may be noted. (1) The mechanical

construction is simple and should fall within the capabilities. of any good instrument-maker of the college type. (2) With the exception of the special slits, the apparatus can be assembled from any good laboratory equipment and dismantled at will. Excessive resources and equipment are not thus tied up in a fixed piece of apparatus which can serve only a limited purpose. (3) The optical equipment is simple. Bi-refracting prisms, Nicol's prisms, etc. are not involved. Carbon bisulphide prisms are used in our spectroscopes. They are inexpensive and give little trouble if reasonable precautions are taken to keep the room at uniform temperature. The only expensive items are the triple achromatic collimating and objective lenses of the large size needed for a spectroscope designed to give a high intensity of light. We wish to note here also that the ribbon filament tungsten lamp of recent invention is a less troublesome high intensity source of light than the Nernst filament which we have previously used. (4) In the use of a bi-refracting prism, for example, for obtaining the colors to be combined and a Nicol's prism for varying intensity, a great deal of light is wasted by the low transmission of the prisms employed. In the use of multiple slits for obtaining the colors to be combined and a variable length of slit for the control of intensity, this wastage is not involved. (5) The needs for qualitative work in color-mixing are satisfied over a wide range of possibilities. And (6) colormixing may be made quantitative. This opens up a wide range of possibilities of investigation of this more complex of the visual responses and of its comparison with the simpler

responses.

Special advantages accrue from focusing the light on the eye which may also be indicated here. (1) The method is much more conservative of light. A very wide range of intensities may be covered which is very important in some of the problems we have in mind. (2) The work may be done under light- as well as dark-adaptation without the damaging admixture of light from the room. (3) The need to use the unsatisfactory artificial pupil for the control of the amount of light entering the eye as the level of intensity is changed is

eliminated. The beam of light is focussed into an image which is smaller than the pupil and of constant size. And (4) in connection with our rotary campimeter,' constructed for use with the spectroscope, color-mixing can be done with any part of the retina from fovea to far periphery over a wide range of intensity of light. When using the method of focusing on the eye it is possible also to do this work with the proper control of brightness of pre-exposure and surrounding field, without which the investigation of the color-functions in the periphery of the retina is practically valueless.

1 C. E. Ferree, Description of a rotary campimeter, Amer. J. Psychol., 1912, 23, 449-453; C. E. Ferree & Gertrude Rand, A spectroscopic apparatus for the investigation of the color sensitivity of the retina, central and peripheral. This JOURNAL, 1916, I, 276.

1

A PENDULUM-PHOTOCHRONOGRAPH 1

BY RAYMOND DODGE

Institute of Psychology, Yale University

Our pendulum photographic recorder is a combination of photochronograph and general myograph with the following characteristics: (1) It has the important advantage for psychological work of being practically noiseless; that is to say, it is noiseless until the interval to be recorded is well over when the pendulum scrapes along a piece of velvet as it is caught to prevent a return swing. (2) It is adapted to time measurements varying from .0001" to 1". (3) As many as fifty reaction records may be taken on a single strip of sensitized paper 2 inches by 7 inches in size. (4) A single count of the time-ordinates suffices for reading the entire series. If every tenth time-ordinate is emphasized, the process of reading the chronographic record is but little more exacting than reading elapsed time on a chronoscope. (5) In addition to these advantages, the pendulum recorder is economical in respect to time of operation, laboratory space, and record material.2

Our laboratory model was built up on a heavy pendulum which happened to be available. The pendulum was supported on steel knife-edges which were mounted on a heavy maple frame. The maximal amplitude of swing is an arc of about sixty degrees.

1 This instrument of precision for psychological laboratories was developed at the Wesleyan Laboratory in 1915 in connection with plans for a study of human variability, which were carried out with the support of the Ernest Kempton Adams Fellowship of Columbia University. It is being described separately to facilitate the publication of a dependent study.

2 A simple but serviceable model was constructed on the general principle of our laboratory instrument by a former assistant in the Wesleyan laboratory, Mr. H. A. Richmond. The expense of the material for this simplified model was under fifteen dollars including the cylindrical lens. The simplified instrument could be used in a semi-darkened room either on a table or on the wall. Unfortunately the war prevented him from publishing a description of his construction.