196 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1909. 
formed into the thermal energy of a gas—it is just the transference of 
kinetic energy from one body to another. 
If we regard potential energy as the kinetic energy of portions of 
the ether attached to the system, then all energy is kinetic energy, 
due to the motion of matter or of portions of ether attached to the 
matter. I showed, many years ago, in my “Applications of Dynam-: 
ics to Physics and Chemistry,” that we could imitate the effects of 
the potential energy of a system by means of the kinetic energy of 
invisible systems connected in an appropriate manner with the main 
system, and that the potential energy of the visible universe may in 
reality be the kinetic energy of an invisible one connected up with it. 
We naturally suppose that this invisible universe is the luminiferous 
ether, that portions of the ether in rapid motion are connected with 
the visible systems, and that their kinetic energy is the potential 
energy of the systems. 
We may thus regard the ether as a bank in which we may deposit 
energy and withdraw it at our convenience. The mass of the ether 
attached to the system will change as the potential energy changes, 
and thus the mass of a system whose potential energy is changing can 
not be constant; the fluctuations in mass under ordinary conditions 
are, however, so small that they can not be detected by any means at 
present at our disposal. Inasmuch as the various forms of potential 
energy are continually being changed into heat energy, which is the 
kinetic energy of the molecules of matter, there is a constant tendency 
for the mass of a system such as the earth or the sun to diminish, and 
thus as time goes on for the mass of ether gripped by the material 
universe to become smaller and smaller; the rate at which it would 
diminish would, however, get slower as time went on, and there is no 
reason to think that it would ever get below a very large value. 
Radiation of light and heat from an incandescent body like the sun 
involves a constant loss of mass by the body. Each unit of energy 
radiated carries off its quota of mass, but as the mass ejected from 
the sun per year is only one part in 20 billionths (1 in 2X10") of the 
mass of the sun, and as this diminution in mass is not necessarily 
accompanied by any decrease in its gravitational attraction, we can 
not expect to be able to get any evidence of this effect. 
As our knowledge of the properties of light has progressed, we 
have been driven to recognize that the ether, when transmitting light, 
possesses properties which, before the introduction of the electro- 
magnetic theory, would have been thought to be peculiar to an emis- 
sion theory of light and to be fatal to the theory that light consists of 
undulations. 
Take, for example, the pressure exerted by light. This would fol- 
low as a matter of course if we supposed light to be small particles 
moving with great velocities, for these, if they struck against a body, 
