In relation to this important fact it will be convenient to refer to a familiar but very beautiful experiment--the formation of a spectrum. An electric lamp is enclosed in an iron lantern, having in its front an upright slit; from this slit there issues a narrow beam of white light, which is made up of rays of many different wave-lengths, all mixed up together. By causing the light to pass through a prism the mixed rays are sorted out side by side according to their several wave-lengths, forming a broad, many-hued band or "spectrum" upon a white screen placed to receive it. (See Fig. 1.) To the visible rays of the longest wave-length is due the red colour on the extreme left. Waves of somewhat shorter length produce the adjoining stripe of orange, and the succeeding colours--yellow, green, and blue--correspond respectively to waves of shorter and shorter lengths. Lastly there comes a patch of violet due to those of the visible rays whose wave-length is the shortest of all. The wave-length of the light at the extreme edge of the red is about 1/34000 inch, and as we pass along the spectrum the wave-length gradually diminishes, until at the extreme outer edge of the violet it is about 1/64000 inch, or not much more than half that at the other end. [Illustration: _Fig. 1.--Image of Slit and of Spectrum._] The two ends of the spectrum gradually fade away into darkness, and the point that I wish to insist upon and make perfectly clear is this:--The position of the boundaries terminating the visible spectrum does not depend upon anything whatever in the nature of light regarded as a physical phenomenon. Ether waves which are much longer and much shorter than those which illuminate the spectrum certainly exist, and evidence of their existence is easily obtainable. But we cannot see them; they fall upon our eyes without exciting the faintest sensation of light. The visible spectrum is limited solely by the physiological constitution of our organs of vision, and the fact that it begins and ends where it does is, from a physical point of view, a mere accident. The spectrum actually projected upon the screen is in truth much longer than that portion of it which any one can see: it extends for a considerable distance beyond the violet at the one end and beyond the red at the other, these invisible portions being known as the ultra-violet and infra-red regions. People's eyes differ in regard to range of sensibility just as their ears do. I believe the sensibility of my own eyes to be normal, but if I were to indicate the two points where the spectrum appears to me to begin and to end, a great many persons would certainly be inclined to disagree with me and place the boundaries somewhere else. Some, indeed, could see nothing whatever in what appears to most of us to be a brilliant portion of the red.