congealed water, it may fairly be classed as a cloud of Inclination, its form being determined by the descent of congealed matter through air-layers of varying velocities and directions. Observation of its movement and filature (or direction of slant) would greatly add to our knowledge of currents of air at a great elevation. But perhaps the most remarkable fact in connection with this extraordinary cloud is that it is gradually decreasing year by year, and will, in all probability, soon become extinct. Cirrus 92. Cirrus may be described as a vague, thin, very lofty cloud, resembling in appearance tufts or bunches of curled hair, not disposed in bands, either perfectly white or, when seen through haze, of a creamy or pale orange tint. Pure Cirrus is generally easily distinguishable in the daytime, as it is not disposed in rather slanting definite lines like Cirro-filum, does not extend into a broad horizontal sheet like Cirrovelum, and is very dissimilar to the rather rounded spots of Cirro-macula. Representations of this cloud are given in Plate V. and Illustration VI. Its altitude is very variable, as seen from Table III. (§ 25), but its average altitude is about 25,000 feet. 93. Cirrus is common in intra-tropical regions, but is not very abundant over the tropical calms. It is rarely seen over the interiors of the large continents, especially at the greatest distance from the leeward oceanic margins of the continents. It is not at all uncommon in the summer seasons of the higher latitudes, and in extra-tropical regions it is, even when thick and abundant, a cloud of the summer "anticyclone." 94. Pure Cirrus is the simplest type of Inclination cloud, because its formation is not complicated by other processes, mentioned in § $9. It nearly always has a slight curve, but the greater this curve the more nearly does the cloud approach the form of the variety Crostilom. La fact, it may sometimes be difficult to say whether the loud is Cs or Crime, but this difficulty is merely one of nomenclature, as the knowledge we gain from the observation of the amount of twist is independent of the name which we give to the cloud. At the same time, we must always remember, in estimating in our own minds the amount of twist, the effect that perspective will have. Thus if the cloud is moving towards or away from us, we shall not see so much of the curve as we should if it were moving across us. It will be seen that the sharper this curve the greater is the difference of velocities between the currents in which the cloud exists. 95. This latter fact may be shown as follows— Suppose that E is the position of the head of the cloud at any moment, and that it has been travelling in the direction AE with a uniform velocity represented by AB. Let the spaces between the equidistant horizontal lines (Figs. 5, 6) represent layers of air through which the particles are falling. Let the velocity of the current in each layer be uniform, and let the distances el, fi, gk, hm represent these velocities, so that el is the velocity of the top layer and hm that of the bottom layer. Also layer. Also suppose that the particles are falling with a uniform velocity represented by Ae. Then the particle which started at A will now be found at a point a such that ha= el +fi + gk + hm. The particle which started one interval later, that is, from B, will be found at a point b such that nb=el+fi+gk; and so on. So that the shape of the whole cloud will at this moment be represented by the figure formed by joining the points a, b, c, d, E, or, more strictly, by a curve passing through these points. By a comparison of Fig. 5 and Fig. 6 we can see at once how rapidity of |