OBITUARY NOTICES. 
453 
tropics of Cancer and Capricorn. In these works he deals 
principally with the general circulation of the atmosphere, 
which is a slow motion compared with the rapid movements 
within a cyclone, hurricane, or a tornado. Ferrel considers 
only the so-called steady motion of a fluid and not the vortex 
motions or surfaces of discontinuity that must attend the 
more rapid movements. He shows that there is a close 
agreement between the observed motions of the winds and 
clouds and the observed temperatures and pressures through¬ 
out the globe on the one hand and his own deductions from 
his adopted principles on the other, and that all the phe¬ 
nomena of rains and storms also confirm his reasoning. 
These views are applied by him to the ocean as well to the 
atmosphere. A steady, slow interchange of oceanic water 
goes on between the equator and the poles due to differences 
of density. This is by the earth’s rotation converted into 
a westward motion near the equator and an eastward motion 
at high latitudes, so that the prevailing winds actually har¬ 
monize with and help to maintain this oceanic circulation, 
although the} 7 cannot be said to produce it. As a conse¬ 
quence of the oceanic circulation the surface of the ocean 
must be depressed in high latitudes and at the equator, and 
be elevated under the tropics relative to the ellipsoidal sur¬ 
face. In the case of enclosed oceans, like the North Atlantic, 
which have a slight gyratory motion with the hands of a 
watch, the surface of the ocean in its central region must 
be above its surface at the borders of the ocean. 
(9.) As a departure from the usual course of his investiga¬ 
tions, I note Ferrel’s remarkably convenient formulae for 
interpolation published in 1861 and 1865, and which I have 
had frequent occasion to use w T ith a great saving of time 
and labor. Undoubtedly he was led to these labor-saving 
inventions by the irksomeness of his daily computations 
for the Nautical Almanac office. 
(10.) In 1865 Ferrel was successful in inventing one of the 
most rapidly converging series as yet discovered for the 
computation of the ratio between the circumference and 
