What did Kenneth Arnold actually see, that June afternoon in 1947? No absolutely certain answer is possible after so long a time. The disks were probably a mirage (see Figure 3) in which the peaks of the mountains seemed to float above the mountain chain[II-25]. An alternative but much less probable explanation is that he observed orographic clouds, a type unique to mountainous country, which often appear to stand more or less motionless and can assume dramatic shapes. “Grindstone” clouds, shaped like thick, solid disks (see Plate Ia), are common phenomena in the valleys just east of the Sierra Nevada in California and in the mountainous regions of Washington, Colorado, and New Mexico--areas where flying-saucer reports have tended to concentrate[II-26]. One of the most spectacular types of mountain cloud, it closely resembles the “_pile d’assiettes_” or “stack of plates” formation in which the cloud assumes a flat, round shape like a plate or a saucer, and two or more are piled together in a neat stack, as in Plate Ib[II-27]. Another picture of a “stack of plates” (which some observers reported as a hovering flying saucer) was made on May 31, 1953, near Jindabyna, Snowy Mountains, New South Wales, and reproduced in _Weather_ in November 1954 Plate 47. The cloud formed over a tub-shaped depression in the mountains and remained stationary for more than an hour[II-28]. [Illustration: _Figure 3._ Mirage of mountain peaks. Top, normal view of mountain chain; bottom, mirage in which some of the peaks seem to be detached and above the peaks, like saucers.] Such clouds reflect the undulations of lee waves formed in the atmosphere when stable currents of air flow over obstacles such as hills or mountains. An up-and-down wave motion may be impressed upon the air, provided that temperature and wind conditions are suitable. As the air describes its wavelike path, it alternately warms and cools, the warming taking place as it sinks into the wave trough and the cooling as it ascends to the wave crest. If the air is very dry, the undulating current will not be visible to the eye, although the updrafts and downdrafts will readily be felt by aircraft that chance to pass through them. On the other hand, if the air before entering the wave is moist enough, the cooling in the wave crest will cause water droplets to condense and a cloud to appear. In the vicinity of an isolated peak the cloud may assume the form of a cap covering the summit, or it may be displaced slightly downwind and resemble a lens or disk. Not infrequently a series of lenticular clouds will appear, trailing downwind at regular intervals of a few miles. Although these wave clouds are usually stationary, they sometimes move at great speed, especially when the air temperature is changing rapidly. From a study of a remarkable photograph made in 1956, R. J. Reed of the University of Washington has offered the suggestion that the disks Arnold saw were actually wave clouds in rapid motion.