366 
Past Changes in the Universe. 
[July, 
in a straight line, probably as a rule traverse an immense 
distance before an encounter, and thus have time to cool 
down before the renovation of its beat by a collision ; the 
dead or used-up material of former suns being thus avail- 
able for fresh suns. 
There is one point in connection with this subject that 
may be worth noticing : — The sun, as is known, is giving 
off an enormous amount of energy in the form of heat into 
the ether of space, no less than about 1720 foot-tons of 
energy being thrown off from every square foot of the sun 
per second. Can it be imagined that this enormous total of 
energy can be given off in a particular direction — i.e., from 
the sun — without any reaction in the opposite direction ; or 
would it be thought that if the sun were emitting all this 
energy from one side only there would be no reaction in the 
opposite direction ? To us it seems incredible, whatever 
the constitution of the ether may be imagined to be, that 
there should be no reaction at all by this enormous total of 
energy thrown off.* Admitting that there is a slight re- 
action, then it is probable that it would not be absolutely 
equal all round the sun, owing to accidental irregularities 
in the temperature and radiating power of the materials of 
the sun’s surface. If this reaction were only a few grains 
per square foot (due to 1720 foot-tons of energy thrown off 
per square foot, per second), the resultant or unbalanced 
reaction, owing to irregularities in the distribution of the 
sun’s heat and radiating power, might amount to a consi- 
derable total quantity, owing to the enormous area of the 
sun. This reaction, even if relatively small, might, in aCting 
for ages (millions of years), in the end sensibly affeCt the 
proper motion of the sun ; or this might be a cause having 
* We do not mean to infer that the reaction due to radiation is the cause of 
the motion of the radiometer, as we think there are grounds for believing that 
(in the case of the relatively extremely feeble heating of the radiometer) it 
would not be a sufficient cause to produce the motion, or probably any mea- 
surable effe<5t. There can be little doubt that the effective cause of the motion 
of the radiometer has been quite successfully traced to the action of the mole- 
cules of the residual gas, as shown by Mr. Johnstone Stoney. This, how- 
ever, by no means proves that there is no reaction by radiation. In considering 
the reaction in the case of the sun, the relatively enormous value of the energy 
continually thrown off from its surface (amounting to thousands of horse- 
power per square foot ) must be duly realised. It may be estimated that a 
portion of the sun’s surface equal in size to the disk of an ordinary radio- 
meter (say a quarter of a square inch in area) is throwing off a wave energy 
equal to twelve horse-power. A vertical column of ether of this sectional 
area, situated on the sun, is therefore transmitting the same energy as the 
belt of a twelve-horse steam-engine. Can it be imagined that it can do so 
without any reaction at all ? Consider, also, the millions of square miles of 
the sun’s surface, each minute portion of which is throwing off this same 
energy. 
