62 JOURNAL OF THE WASHINGTON ACADEMY OF SCIENCES VOL. 13, No, 4 
The explanation of the roar of the mountain, most pronounced 
during winter when the trees are bare and the winds strongest, is now 
obvious. It is only the feeble notes of the myriads of forest twigs 
at and near the mountain top merged into a mighty volume of sound 
of a more or less average pitch, all focussed, as shown by fig. 4, on 
the valley beyond, but so crudely and variably focussed as to cause 
great fluctuations in the intensity. 
Clouding of mountain crest.—At the same time, approximately, 
that the mountain begins to roar, its crest usually becomes cloud 
capped, even when all other portions of the sky are absolutely clear. 
As already explained, this roaring commonly is in response to the 
strong and persistent winds on the forward side of a cyclonic storm, 
the side where, because the wind is from the southeast, roughly, 
the clouds thicken and rain begins. As also previously explained, the 
wind is necessarily from the southeast over Peters Mountain when its 
roar is heard in the adjacent northwest valley—the particular cir- 
cumstance here under discussion. Of course this mountain, and 
others similarly oriented, roars to its opposite, or southeast, valley 
in response to northwest winds; but these winds generally are relatively 
weak or of comparatively short duration and the roaring they produce 
correspondingly inconspicuous. They, therefore, will not be further 
considered in what follows. 
The southeast winds, then, over this mountain, coming, at least 
in their greater distances, from relatively warm and humid regions, 
progressively approach saturation. Furthermore, it is the under 
layers, in general, of the air that contain the largest amount of moisture, 
partly because of their higher temperature and consequent greater 
moisture capacity, as we elliptically (if not even cryptically) express 
it, and partly because the source of this humidity—bodies of water, 
and growing vegetation, chiefly—is at the surface of the earth. 
After a time, then, the lower layers of the atmosphere become sQ 
humid that their cooling by expansion as they rise over the mountain 
carries their temperature below the dew point, and thereby produces 
a cloud along the mountain crest. At first this cloud, contimuously 
formed by condensation on its windward side, is continuously evapo- 
rated on its leeward side. That is, it is a stationary cloud though 
its every droplet is in rapid motion and of short duration. It is a 
local condition in the flowing air just as a waterfall is a local condi- 
tion in the passing stream. 
Drifting of scud.—The leeward side of the crest cloud, just explained, 
is dragged out in irregular fragments by the gusts and swirls of the 
