from the lower ones.  Again, if the night be a very calm one, there will be no supply of warmth from fresh currents of air falling down upon it.  Hence, in the treble event of a clear blue sky, a non-conducting soil, and a perfectly still night, we are liable to have great cold on the surface of the ground.  This is shared by a thin layer of air that immediately rests upon it; while at each successive inch in height, the air becomes more nearly of its proper temperature.  A vast number of experiments have been made by Mr. Glaisher on this subject (’Phil.  Trans.’ 1847), the upshot of which is that a thermometer laid on grass, under a blue sky on a calm night, marks on an average 8 degrees Fahr. colder than one 4 feet above it; 1 inch above grass, 5 1/2 degrees; 1 foot, 1 degrees; 4 feet, 1/2 degrees; on gravel and sand the differences are only about one-third as much.  Sheep have a practical knowledge of these differences.  Often, in an early walk on dewy mornings, I see all the sheep in Hyde Park bivouacked on the gravel walks of Rotten Row.  The above figures are the results of experiments made in England, where the air is always moist, and the formation of dew, while it testifies to the cold of the night, assists largely to moderate it.  In arid climates the chill would be far greater; such would also be the case at high elevations.  One of Mr. Glaisher’s experiments showed a difference of no less than 28 degrees between the cold on the ground and that at 8 feet high.  This might often be rivalled in an elevated desert, as in that of Mongolia.  Hence the value of the protection of a roof and of a raised sleeping-place, to a man sleeping under a blue sky in still weather, admits of easy interpretation.

Various Methods of Bivouacking.—­Unprotected.—­Mr. Shaw, the traveller in Thibet, says:—­“My companion and I walked on to keep ourselves warm, but halting at sunset, had to sit and freeze several hours before the things came up.  The best way of keeping warm on such an occasion, is to squat down, kneeling against a bank, resting your head on the bank, and nearly between your knees.  Then tuck your overcoat in, all round you, over head and all; and if you are lucky, and there is not too much wind, you will make a little atmosphere of your own inside the covering, which will be snug in comparison with the outside air.  Your feet suffer chiefly, but you learn to tie yourself into a kind of knot, bringing as many surfaces of your body together as possible.  I have passed whole nights in this kneeling position, and slept well; whereas I should not have got a wink had I been stretched at full length with such a scanty covering as a great-coat.”

Bushes.—­I have shown that the main object before sleeping out at night is to secure a long wind-tight wall, and that the next is to obtain a roof.  Both these objects may be attained by pleachingtwo or three small neighbouring bushes into one; or branches may be torn off elsewhere and interwoven between the bushes.  A few leafy boughs, cut and stuck into the ground, with their tops leaning over the bed, and secured in that position by other boughs, wattled-in horizontally, give great protection.  Long grass, etc., should be plucked and strewn against them to make them as wind-tight as possible.