Frequently it happens that before the ascent of vapour has gone quite far enough to produce a cloud, other causes co-operate, and the cloud makes its appearance suddenly over considerable patches of sky. The most potent of these is a fall of the barometric pressure, which is brought about by some of the air far above the region of even the highest clouds flowing away to some other district. The air at all lower levels being thus relieved of the superincumbent pressure, immediately expands, and is thereby cooled throughout. Consequently, if at any level it was near its point of saturation, it will be carried beyond that point, and cloud will rapidly make its appearance over a large part of the sky, possibly at more than one level. Stratiform arrangements will be the rule; but if interconvection is going on at the time, its presence will be betrayed by a granular or cumuloid structure. Interconvection clouds should then be most frequent, and best formed when the air as a whole is still or moving slowly (so as not to create great eddies), when the temperature is rising rapidly, and when the barometer is making a sudden fall. All these conditions are met in thunder weather, and at the time when a summer anticyclone is giving way. It will be remembered that many of the most beautiful forms have been described as forming under one or the other of these very conditions. A second contributing cause, and one which tends to make the condensation in patches or long broad bands ranged roughly at right angles to the direction in which the air is moving, has been referred to earlier. It is the passage of the air over an undulating country; the up-and-down movements of the lower air being transmitted upwards to great altitudes, as ever broadening and flattening waves. If the upper air is flowing more rapidly than the lower, these broad waves may be far ahead of their real cause, which will, therefore, quite escape recognition, but the phenomenon is constantly to be detected in the arrangement of the lower clouds. Two instances in the writer’s experience will suffice. It was desired one morning to measure the altitude of some small clouds which were passing from the north-west at a height of probably between 2000 and 4000 metres, over a hill only about 150 metres higher than the valley in which the apparatus was fixed. In order to make the measurement, it was necessary for the cloud to cross the valley and appear in the same field of view as the sun, according to the method that will be described further on. But in order to cross the valley the air had to descend, and so, of course, had the cloud stratum, though to a less extent. But small as the descent was, it was enough to dry up the clouds entirely, and for more than a couple of hours the clouds came sailing over the hill, disappearing entirely, and then reforming so far beyond that no measurement was possible, since not one single fragment came near enough to the position of the