_April 4, 5._ Frost, wind north-east, the wind changed in the middle of the day to the north-west without rain, and has done so for three or four days, becoming again north-east at night. For the sun now giving greater degrees of heat, the air ascends as the sun passes the zenith, and is supplied below by the air on the western side as well as on the eastern side of the zenith during the hot part of the day; whence for a few hours, on the approach of the hot part of the day, the air acquires a westerly direction in this longitude. If the north-west wind had been caused by a retrograde motion of some southern air, which had passed over us, it would have been attended with rain or clouds. _April 10._ It rained all day yesterday, the wind north-west, this morning there was a sharp frost. The evaporation of the moisture, (which fell yesterday) occasioned by the continuance of the wind, produced so much cold as to freeze the dew. _May 12._ Frequent showers with a current of colder wind preceding every shower. The sinking of the rain or cloud pressed away the air from beneath it in its descent, which having been for a time shaded from the sun by the floating cloud, became cooled in some degree. _June 20._ The barometer sunk, the wind became south-west, and the whole heaven was instantly covered with clouds. A part of the incumbent atmosphere having vanished, as appeared by the sinking of the barometer, the remainder became expanded by its elasticity, and thence attracted some of the matter of heat from the vapour intermixed with it, and thus in a few minutes a total devaporation took place, as in exhausting the receiver of an air-pump. See note XXV. At the place where the air is destroyed, currents both from the north and south flow in to supply the deficiency, (for it has been shewn that there are no other proper winds but these two) and the mixture of these winds produces so sudden condensation of the moisture, both by the coldness of the northern air and the expansion of both of them, that lightning is given out, and an incipient tornado takes place; whence thunder is said frequently to approach against the wind. _August 28, 1732._ Barometer was at 31, and _Dec. 30_, in the same year, it was at 28 2-tenths. Medical Essays, Edinburgh, Vol. II. p. 7. It appears from these journals that the mercury at Edinburgh varies sometimes nearly three inches, or one tenth of the whole atmosphere. From the journals kept by the Royal Society at London it appears seldom to vary more than two inches, or one-fifteenth of the whole atmosphere. The quantity of the variation is said still to decrease nearer the line, and to increase in the more northern latitudes; which much confirms the idea that there exists at certain times a great destruction or production of air within the polar circle.