double; for the positive secondary image of the adjacent light areas will appear within it. To this end its apparent width to the naked eye must be some 8′ or 10′ (if eyes are alike in this dimension). In a telescope magnifying, say, 400 diameters, this width need be only a little over 1″. If the planet is 16″ in diameter (a rough average of its favorable position in recent years) this will amount to closely 10° on its surface. Now the double canals of Schiaparelli, in 1881–2, and of Perrotin and Thallon, in 1886, are frankly of this width and, I believe, are due to this cause. In any case the test to be applied is evidently the relation between the apparent width of a double and the radius of the halo illusion. The prevention of error in the future will evidently be the application of different powers to each canal, particularly a low power which will make its width appear less than 6′. This must be done with care for low powers increase the number of rays.
The halo illusion is also responsible for marginal canals. When a dark area becomes 6′ or 8′ wide, it appears double, having a light interior and dark edges. With any increase of width the dark edges, giving the effect of the marginal canals, remain. Hence along the edge of any dark area there appears a fictitious canal. Professor E. W. Maunder observed this in his excellent artificial planet study of a few years ago. I believe that high powers by reducing contrast will help to eliminate this error.
The mention of the chromatic rings draws attention to chromatic aberration in the eye and in the telescope. This effect in the telescope is so great that colors in a refracting telescope are not in the least trustworthy. The blue-green tint attributed to the dark areas on Mars is a product of the telescope. Its existence on our neighbor can only be verified by the use of a reflector.
Thus in conclusion we see that there are fundamental defects in the human eye producing faint canal illusions, that these have worked serious injury to our observations in the past and that in the future they may be avoided chiefly by the simple expedients of varying the position of the head and using a wide range of magnifying power.