206 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1926 
apparatus in which a radiocompass coil, tuned to about 30,000 
meters, is rotated slowly and continuously by clockwork, the atmos- 
pheric crashes being recorded on a drum attached to the coil. 
It should be said in this connection that it has been very common 
in Europe to estimate the strength of atmospherics by the number of 
disturbances occurring in a given time. This method, of course, 
would hardly seem to be applicable to our Washington summer 
conditions, or to the conditions during the disturbance season in the 
tropics where often in the afternoons and evenings the noise in the 
telephones forms an almost continuous rumbling through which no 
signal can be heard unless it is strong enough to rise above the 
background of disturbing sounds. 
If, indeed, there is a physical difference between the atmospherics, 
crashes, grinders, etc., it is not at all certain that what is being 
measured in Europe by the counting method is the same thing that 
is being measured in America, either by direct estimates of the 
average disturbance strength, or by measuring the strength of signal 
which can be read through the disturbances. 
On the Atlantic and Pacific coasts of the United States, except 
for occasional local thunderstorms, very little certain connection has 
been noticed between the direction of the atmospheric disturbances 
and rain areas. On the Atlantic coast, the main disturbances seem 
to come roughly from the Southwest, but it seems uncertain whether 
the sources are in the Allegheny Mountains or much farther re- 
moved, perhaps in Yucatan. Experiments reported by the Navy 
Department in New Orleans have indicated the more southerly 
origin. 
Unfortunately, very few triangulation experiments have been 
made in America for fixing the exact positions of sources of atmos- 
pherics. In most cases, therefore, the direction is all that is known. 
Observations made at Madison, Wisconsin, by Professor Terry of the 
University of Wisconsin, covering the last two years, show conditions 
in the Middle West which are similar to those described by the con- 
tinental European observers; that is, there is no single prevailing 
direction of the atmospherics, but a more or less definite connection 
with thunderstorms and other rain areas. This absence of any pre- 
vailing southerly source of atmospherics in the central portion of the 
country casts doubt on the Mexican origin of those observed in the 
Atlantic coast region, since the distance from Yucatan to Madison, 
Wisconsin, is about the same as from Yucatan to Washington. 
On the Pacific coast of the United States it is pretty well estab- 
lished that at least at San Francisco and San Diego the sources of 
disturbances are largely local, lying in the mountain ranges not far 
from the coast. These centers seem to be permanently fixed, 
resulting in very constant directional conditions. 
