48 
ABBE. 
the first steps proper to be taken in the reduction of the ob- 
servations with balloons or kites to any given level in the 
atmosphere. These authors make the point that as the sur- 
face of the ocean is an equipotential surface, therefore obser- 
vations reduced from it upward to some other equipotential 
surface of gravity possess simpler relations to each other than 
when reduced to a uniform height above sea-level. With 
the help of the Carnegie Institution, Bjerknes and Sand- 
strom are now still further developing and improving their 
method with the assurance of throwing new light upon at- 
mospheric motions. 
Even when we study the motions of gases on a small scale 
in our laboratories, we have the greatest difficulty in under- 
standing the processes that go on right before our eyes ; still 
more is this the case with those that go on in the atmosphere. 
The smooth flow of air, like that of water, made visible by 
some fine floating particles, suddenly changes without ap- 
parent cause into a series of whirls and vortices, and then 
from whirls hack to a steady, smooth flow. A vortex ring of 
air traverses a large room to a distance far greater than a fine 
straight jet can do, as though its large front surface ex- 
perienced less resistance than a small jet. 
Under hydrodynamics proper I may mention the works 
of Chree, Bigelow, Bjerknes, J. J. Thomson, Ekholm, Mar- 
gules, and Rayleigh; also the discussion between Airy, Fer- 
rel, and Kelvin as to the tides in the atmosphere, resting on 
the interpretation of a certain formula in the memoirs of 
Laplace, a subject that was finally elucidated by Dr. Ling, of 
Columbia University. Not only do we owe the mathematical 
theory of heat to Fourier and Poisson, but especially to the 
former a posthumous memoir on the motions of fluids in 
which the internal motions and the distribution of heat are 
mutually interdependent. This memoir is but a fragment, 
establishing certain differential equations, the solution of 
which is rarely possible when the boundary conditions are 
given ; so that in general we must at present rely upon partial 
solutions and suggestions derived from experiments or obser- 
vation. To Rayleigh and Stokes we owe a number of mem- 
