THE PRESIDENTIAL ADDRESS 9 



the other alternative that we might, in order to work with the shortest 

 waves, dispense with lenses altogether : and in fact in using X-rays 

 this is done. We are then limited to controlling the course of the 

 rays by means of tubes or pinholes. This restriction is so serious 

 that it altogether defeats the possibility of constructing a useful 

 X-ray microscope analogous to the optical or the electron microscope. 

 In spite of this the use of X-rays is of fundamental value for dealing 

 with a particular class of objects, namely, crystals, which themselves 

 have a regular spacing, comparable in size with the length of the 

 waves. Just as the spacing of a ruled grating (say one 1/20, 000th 

 of an inch) can be compared with the wave-length of light by measur- 

 ing the angle of diffraction, so the spacing of atoms in a crystal can 

 be compared with the wave-length of X-rays. But here the indica- 

 tions are less direct than with the microscope, and depend on the 

 object having a periodic structure. So that the method hardly falls 

 within the scope of this address. How essential the difference is will 

 appear if we consider that the angle to be observed becomes greater 

 and not less the closer the spacing of the object under test. 



Colour vision is one of nature's most wonderful achievements, 

 though custom often prevents our perceiving the wonder of it. 

 We take it for granted that anyone should readily distinguish the 

 berries on a holly bush, and we are inclined to be derisive of a colour- 

 blind person who cannot do so. But so far anatomy has told us 

 little or nothing of how the marvel is achieved. Experiments on 

 colour vision show that three separate and fundamental colour 

 sensations exist. It is probable that the cones of the retina are 

 responsible for colour vision and the rods for dark adapted vision 

 which does not discriminate colour. But no division of the cones 

 into three separate kinds corresponding to the three colour sensa- 

 tions has ever been observed. Nor is any anatomical peculiarity 

 known which allows a colour-blind eye to be distinguished from a 

 normal one. 



Can artificial resources help to improve colour discrimination ? 

 In some interesting cases they can. Indeed, the whole subject of 

 spectroscopy may be thought of as coming under this head. We 

 can recognise the colour imparted by sodium to a flame without 

 artificial help. When potassium is present as well, the red colour 

 due to it can only be seen when we use a prism to separate the red 

 image of the flame from the yellow one. Such a method has its 

 limitations, because if the coloured images are more numerous they 

 overlap, and the desired separation is lost. To avoid this it is 

 necessary to make a sacrifice, and to limit the effective breadth of 

 the flame by a more or less narrow slit. And if the images are very 

 numerous the slit has to be so narrow that all indication of the 

 breadth of the source is lost. This, of course, is substantially the 



B 2 



