temperature though exposed to a considerable range of external heat and cold. It is to such a power of accumulation of heat in our soil and lower atmosphere that we must impute the overwhelming contrast between our climate and that of the moon. With us, the solar heat that penetrates our vapour-laden and cloudy atmosphere is shut in by that same atmosphere, accumulates there for weeks and months together, and can only slowly escape. It is this great cumulative power which Mr. Lowell has not taken account of, while he certainly has not estimated the enormous loss of heat by free radiation, which entirely neutralises the effects of increase of sun-heat, however great, when these cumulative agencies are not present.[12] [Footnote 12: The effects of this 'cumulative' power of a dense atmosphere are further discussed and illustrated in the last chapter of this book, where I show that the universal fact of steadily diminishing temperatures at high altitudes is due solely to the diminution of this cumulative power of our atmosphere, and that from this cause alone the temperature of Mars must be that which would be found on a lofty plateau about 18,000 feet higher than the average of the peaks of the Andes!] _Temperature on Polar Regions of Mars._ There is also a further consideration which I think Mr. Lowell has altogether omitted to discuss. Whatever may be the _mean_ temperature of Mars, we must take account of the long nights in its polar and high-temperate latitudes, lasting nearly twice as long as ours, with the resulting lowering of temperature by radiation into a constantly clear sky. Even in Siberia, in Lat. 67-1/2°N. a cold of-88°F. has been attained; while over a large portion of N. Asia and America above 60° Lat. the _mean_ January temperature is from-30°F. to-60°F., and the whole subsoil is permanently frozen from a depth of 6 or 7 feet to several hundreds. But the winter temperatures, _over the same latitudes_ in Mars, must be very much lower; and it must require a proportionally larger amount of its feeble sun-heat to raise the surface even to the freezing-point, and an additional very large amount to melt any considerable depth of snow. But this identical area, from a little below 60° to the pole, is that occupied by the snow-caps of Mars, and over the whole of it the winter temperature must be far lower than the earth-minimum of-88°F. Then, as the Martian summer comes on, there is less than half the sun-heat available to raise this low temperature after a winter nearly double the length of ours. And when the summer does come with its scanty sun-heat, that heat is not accumulated as it is by our dense and moisture-laden atmosphere, the marvellous effects of which we have already shown. Yet with all these adverse conditions, each assisting the other to produce a climate approximating to that which the earth would have if it had no atmosphere (but retaining our superiority