FIRST PRINCIPLES 35 



accuracy of its colour-vision, surpassed by that of any other 

 animal. Yet the eye, considered as an optical apparatus, is 

 extremely faulty; the several refractive media are not cor- 

 rectly centred, and are guilty of spherical and chromatic 

 aberrations, besides a variety of minor faults. The layer in 

 which waves of light are converted into nervous impulses 

 (the rods and cones) is on the back of the retina, so that the 

 picture is more or less obscured, like a photograph taken 

 with the back of the sensitised paper in contact with the 

 negative. But the picture which the Mind sees does not 

 present the imperfections of the image on the retina. By 

 force of training, the Mind has come to ignore the faults of 

 the retinal image. It does not take cognizance of the ' ' blind 

 spot " or of the yellow colour and double refraction of the 

 "yellow spot," the only part of the retina which is suffi- 

 ciently sensitive for "direct vision." Nor, indeed, can the 

 retina be said to be very sensitive, since objects which 

 subtend an angle with the eye of less than 60" do not fall on 

 separate "sensational units" of its surface. They fail to 

 give rise to separate sensations, and are therefore seen not 

 as two objects but as one. Since the retina is a mosaic of 

 sensational units, every apparently continuous line is really 

 seen as a succession of points. Again, it is far from being 

 capable of responding to all vibrations of light. There are 

 vibrations slower and longer than the red and more rapid 

 and shorter than the violet to which it is insensitive. And 

 within its range, who shall say that it gives us correct 

 information as to the relative wave-lengths of rays of light 

 the quality of the different rays which we distinguish as 

 colour? The rays which produce the visible spectrum 

 present, except for certain gaps due to the absorption of 

 Frauenhofer's lines by the sun's atmosphere, every possible 

 rate of vibration from 381 billions to the second to 764 

 billions ; but the eye can distinguish them only as they coin- 

 cide with or approximate to three mean rates. It groups 

 them as red, green, violet, or combinations of these three 



