It has already been stated that a layer of water less than the twentieth of an inch in thickness suffices to stop and destroy all waves of radiant heat emanating from an obscure source. The longer waves of the obscure heat cannot get through water, and I find that all transparent compounds which contain _hydrogen_ are peculiarly hostile to the longer undulations. It is, I think, the presence of this element in the humours of the eye which prevents the extra red rays of the solar spectrum from reaching the retina. It is interesting to observe that while bisulphide of carbon, chloride of phosphorus, and other liquids which contain no hydrogen, permit a large portion of the rays emanating from an iron or copper ball, at a heat below redness, to pass through them with facility, the same thickness of substances equally transparent, but which contain hydrogen, such as ether, alcohol, water, or the vitreous humour of the eye of an ox, completely intercepts these obscure rays. The same is true of solid bodies; a very slight thickness of those which contain hydrogen offers an impassable barrier to all rays emanating from a non-luminous source.[A] But the heat thus intercepted is by no means lost; its _radiant form_ merely is destroyed. Its waves are shivered upon the particles of the body, but they impart warmth to it, while the heat which retains its radiant form contributes in no way to the warmth of the body through which it passes. [Sidenote: FINAL COLOUR OF ICE AND WATER BLUE.] Water then absorbs all the extra red rays of the sun, and if the layer be thick enough it invades the red rays themselves. Thus the greater the distance the solar beams travel through pure water the more are they deprived of those components which lie at the red end of the spectrum. The consequence is, that the light finally transmitted by the water, and which gives to it its colour, is _blue_. [Sidenote: EXPERIMENT.] I find the following mode of examining the colour of water both satisfactory and convenient:--A tin tube, fifteen feet long and three inches in diameter, has its two ends stopped securely by pieces of colourless plate glass. It is placed in a horizontal position, and pure water is poured into it through a small lateral pipe, until the liquid reaches half way up the glasses at the ends; the tube then holds a semi-cylinder of water and a semi-cylinder of air. A white plate, or a sheet of white paper, well illuminated, is then placed at a little distance from one end of the tube, and is looked at through the tube. Two semicircular spaces are then seen, one by the light which has passed through the air, the other by the light which has passed through the water; and their proximity furnishes a means of comparison, which is absolutely necessary in experiments of this kind. It is always found that, while the former semicircle remains white, the latter one is vividly coloured.[B]