Going much higher up into the region of cirrus, we meet with the most minute and delicate ripple clouds. Some of these have already been referred to. They are connected with either cirro-macula, cirro-cumulus, or cirro-stratus, just as the coarser textured waves we have been considering are connected with alto-cumulus or alto-stratus. In Plate 56 we have an example in which we can see the stages in the process. Nearest to the zenith we have cirro-cumulus, which is here and there irregularly distributed, but is generally arranged in delicate ripples, which are variously curved. Nearer the horizon the troughs of the waves are filled in, and sheets of cirro-stratus are the result. Here, again, the wave-form is evidently not typical. It is an arrangement of either cirro-cumulus or cirro-stratus, produced by the intersection of the plane of condensation by a series of wave movements. [Illustration: Plate 56. CIRRO RIPPLES. (_Cirro-cumulus Undatus._)] The arrangement is, however, so striking a feature when it is well shown that any description of the cloud which contains no reference to the waves is manifestly incomplete, and this would be best effected by adding the word undatus or waved to the name of the cloud. Plate 54 will then be alto-cumulus undatus, Plate 55 alto-stratus undatus, and Plate 56 would be described as cirro-cumulus undatus, passing into cirro-stratus undatus and cirro-stratus. In popular language Plate 55 might be called alto waves, Plate 54 crested alto waves, and Plate 56 cirro ripples. If we are satisfied that the wave clouds are due to a wave movement intersecting a plane of incipient cloud formation, the whole question of their mode of production resolves itself into two parts--how is that plane of incipient condensation produced? and how can we account for the intersecting waves?