380 Transactions of the Canadian Institute. [Vol. VII. 



mounted on another piece of glass forming the artificial spectrum. The 

 amount of the reproduction colours necessary to match each of these 

 glasses in luminosity and hue is accurately measured, giving the 

 analogue of Maxwell's colour-mixture curves. The photometric 

 measurement of the densities of the resulting negatives is avoided by 

 reducing the luminosities of these glasses sufficiently, by means of 

 rotating sectors or patches of developed grey, to render the quantity of 

 red light, say, coming through each glass equal. A negative of such an 

 apparatus, taken through a correctly adjusted red filter or screen, should 

 give a series of patches of equal density ; and, since the equality of the 

 densities of adjacent patches can be accurately estimated by the eye, no 

 photometric measurements are necessary, and the adjustment of filters 

 is much facilitated. A similar apparatus is, of course, required for the 

 green, and for the blue-violet filters. 



This device, due to Sir Wm. Abney, is usually called an Abney 

 sensitometer, and is of very great service in three-colour work ; a 

 simpler form of the same instrument is also very largely employed in 

 the adjustment of compensating filters for orthochromatic plates. The 

 set of filters employed in making the examples of three-colour work to 

 be presently shown were adjusted by an Abney sensitometer, and were 

 supplied by Mr. Sanger Shepherd, of London, who was awarded the 

 medal of the Royal Photographic Society in 1899 for his three-colour 

 filters. They were adjusted for use with the Cadett Spectrum plate, and 

 will evidently, when the varying colour sensitiveness of different plates 

 is remembered, only give correct results with this plate. The Cadett 

 Spectrum and the Lumiere Panchromatic are the most suitable plates 

 for three-colour work. There are other panchromatic plates manufac- 

 tured, but these have hardly, as yet, entered into serious competition 

 with the above named brands. The Cadett plate is not so sensitive to 

 red as the Lumiere and requires a longer exposure through the red 

 filter, but has the decided advantage of giving a much longer scale of 

 correct gradation, and, in subjects with much contrast, will be more 

 likely to produce correct results. It is possible to use a different brand 

 of plates with each filter, but not advisable where good work is desired 

 as it will be found almost impossible, on account of the different 

 qualities of the plates, to secure a harmonious set of negatives. In 

 three-colour photo-mechanical process work, this last procedure is 

 frequently followed but the negatives and positives require and receive 

 considerable retouching and etching. 



On comparing the absorptions of these taking screens with those 

 of the reproduction glasses [s/iown], it is at once seen that the former 



