Since it has been known that liquid oxygen is blue, many people have concluded that this explains the blue colour of the sky. But it has really nothing to do with the point at issue. The blue of the liquid oxygen becomes so excessively faint in the gas, further attenuated as it is by the colourless nitrogen, that it would have no perceptible colour in the whole thickness of our atmosphere. Again, if it had a perceptible blue tint we could not see it against the blackness of space behind it; but white objects seen through it, such as the moon and clouds, should all appear blue, which they do not do. The blue we see is from the whole sky, and is therefore reflected light; and as pure air is quite transparent, there must be solid or liquid particles so minute as to reflect blue light only. In the lower atmosphere the rain-producing particles are larger, and reflect all the rays, thus diluting the blue colour near the horizon, and, by refraction and reflection combined, producing the various beautiful hues of sunrise and sunset. This production of exquisite colours by the dust in the atmosphere, though adding greatly to the enjoyment of life, cannot be considered essential to it; but there is another circumstance connected with atmospheric dust which, though little appreciated, might have effects which can hardly be calculated. If there were no dust in the atmosphere, the sky would appear black even at noon, except in the actual direction of the sun; and the stars would be visible in the day as well as at night. This would follow because air does not reflect light, and is not visible. We should therefore receive no light from the sky itself as we do now, and the north side of every hill, house, and other solid objects, would be totally dark, unless there were any surfaces in that direction to reflect the light. The surface of the ground at a little distance would be in sunshine, and this would be the only source of light wherever direct sunlight was cut off. To get a good amount of pleasant light in houses it would be necessary to have them built on nearly level ground, or on ground rising to the north, and with walls of glass all round and down to the floor line, to receive as much as possible of the reflected light from the ground. What effect this kind of light would have on vegetation it is difficult to say, but trees and shrubs would probably grow laterally towards the south, east, and west, so as to get as much direct sunshine as possible. A more important result would be that, as sunshine would be perpetual during the day, so much evaporation would take place that the soil would become arid and almost bare in places that are now covered with vegetation, and plants like the cactuses of Arizona and the euphorbias of South Africa would occupy a large portion of the surface.