A Standardized Colour-vision Testing Lantern (ii) Transport Type.

THE purpbse of this note is to describe a modification of the standardized colour vision testing lantern designed by the present writer (Martin, 1939), which should render it specially suitable for use by. transport undertakings. Experience has shown, however, that the mariner's test was a very satisfactory one in itself. Therefore the modifications are supplementary, and the marine form of the test can be carried out with the new form of the lantern if desired, as well as the " transport " tests. 1 he additions consist of an orange-yellow filter in addition to the original red, green, and white of the first form; so that pairs of "red with orange " and " orange with green " can be shown. Rail and road signals now employ orange signal lights extensively, and. on many grounds familiar to those dealing with the personnel to be tested it is desirable to make the test simulate practical conditions. For practical trials, a selection of orange filter glasses was obtained from Messrs. Chance Bros. and Co. Ltd. and combined with suitable neutral--filters so as to bring the intensity of the transmitted light to equality with that of the standard lights of the earlier'lanternr. Trial with the smallest apertures and standard conditions then showed that normal persons with good colour -vision sometimes made mistakes; for example a combination of "green-orange," i.e. " green with orange " might be called " green-red," or " orange-red " be called " white-red." It is therefore necessary when using the smallest apertures (0-02 in.) to restrict the lights shown to red, green, and white as in the well-established mariners' test; and this is even more important when the modifying glasses are used with the small apertures. / Trial showed, however, that when the apertures were enla'rged to 0 05 ins., the distinctions were very easy to-normal vision, and this suggested the use of 0.04. in. as a standard size for the pairs" 6f apertures with which the supplementary part of the test can be conducted. Experiments were made with a number of orange Flasses and it appeared that, out of those available, a filter stated to have a hue wave-length of 5970A * was the most readily distinguished from