APPLICATION OF MATHEMATICS IN METEOROLOGY. 219 
other on the same levels instead of above one another at dif¬ 
ferent levels. The canal theory found a laboratory experi¬ 
ment to match it by heating water at one end of a long box, 
when the resulting motion apparently satisfied the mathe¬ 
matical analysis, though that was before the international 
observations were made. The reconstruction process is quite 
simple in conception, but intricate in its details. For the local 
vertical central convection current is substituted a general 
system of horizontal currents flowing from the tropics and 
polar zones respectively into the middle zones. For the 
general canal theory of the overflowing strata is substituted 
a counterflow of currents in the lower strata, and on the same 
levels. The cyclones and anticyclones are due to the inter¬ 
action of these horizontal currents of air of different temper¬ 
atures, which transport the enormous energy derived from 
the solar radiation in the tropics, and expend it in raising 
the air in the polar zones to a higher temperature, the 
cyclones being the mechanical products of this thermo¬ 
dynamic process. The observed stream lines and the com¬ 
puted isobars in the higher levels point to this view in the 
most positive manner, and it is itself in harmony with the 
requirements of thermodynamics as well as hydrodynamics, 
assuming a type of engine which is constituted like that of 
an atmosphere heated in the equatorial regions of a rotating 
globe. Unfortunately for meteorology, from this statement 
it now appears to be necessary to reconstruct a great portion 
of the old theory of the general cyclone, and to reject entirely 
the theories which have been proposed to explain local 
cyclones and anticyclones. 
The reversal of important scientific researches in the pro¬ 
gress of investigation is so common in the history of science, 
that it brings no discredit upon students who have explained 
matters according to the data in their possession. Indeed, 
Lord Kelvin considers it to be a “point of honor” to make 
such reversals for one’s self in the interests of perfect scien¬ 
tific truth, and he has set the world an illustrious example 
of this highminded candor and self-effacement. Thus, a half 
