384 The N.Z. Journal of Science and Technology. [Dec. 
are drawn inwards and strike its edge. Under the action of gravity, wdiere 
two bodies only are concerned, no impact can possibly be direct unless either 
(i) the two bodies are initially at rest, or (ii) the initial relative velocity of 
the two is exactly in the line joining their centres. In any particular case 
of impact it is exceedingly improbable that either of these conditions will 
be fulfilled, and, if the latter is, no attraction is needed to bring about the 
collision, and its occurrence is merely hastened and not caused by gravi¬ 
tation. In this case we must consider the relative motion to be one of 
approach, for a motion of recession would imply that the bodies had 
previously passed through one another. 
To sum up, then, we may say that practically every collision which is 
brought about by gravitation is of a grazing character, and of those which 
would have occurred without its action, and are therefcre only modified by 
it, the vast majority are also oblique. 
Grazing collisions, then, being the prevailing type, are the most important 
to study. We have seen that when two suns meet they possess, in addition 
to any heat-energy that corresponds to their temperature, a vast store of 
kinetic energy equivalent to tens of millions of calories per gramme. The por¬ 
tions of the two suns that come in one another’s way have the greater part 
of this energy of motion changed suddenly into molecular energy or heat. 
But what happens to the remaining portions of the grazing suns ? It 
seems evident that if the graze is a slight one these parts pass on. There 
is nothing to stop them. The energy required to raise the elements from 
the solid state at absolute zero to the critical temperature is probably 
less than a thousand calories per gramme. What is this in comparison 
with the tens of millions of calories available ? Moreover, the energy 
required to shear the bodies must be far less even than that required to 
vaporize them. The expenditure of this infinitesimal fraction of the avail¬ 
able kinetic energy will not therefore diminish the velocity of the bodies 
by any appreciable amount. If they had not grazed, they would have gone 
on in their orbits. As it is, they go on with a slight change in velocity and 
a more decided change in the eccentricity of their paths. New rotations 
are given to them, and each has a fiery wound gouged out of one 
side. Brilliant as the light must be that comes from these blazing scars, 
it is faint and obscure whilst masked by the intense radiance of the arrested 
masses, and plays a very minor part in the visible phenomena of novae. 
It is not until the dazzling brilliance of their evanescent companion has 
faded away that the wounded stars can appear to a distant observer as a 
variable spectroscopic binary with two maxima and two minima, a system 
wdiich may perhaps show different spectral lines at the two maxima, and 
which probably appears nebulous at each minimum. 
The portions struck off from the stars form suddenly an incandescent 
mass beside which the splendour of the wounded stars is invisible. This 
evanescent body is the new star. Whilst its brilliant career lasts it 
monopolizes all attention. It is its light which shows the astonishing 
variations. It is its spectrum which is characteristic of all these tem¬ 
porary additions to the stellar hosts. It is its dissipation into space that 
causes the otherwise inexplicably rapid diminution of the nova’s light. 
We must therefore give special consideration to this third body. When 
two equal stars meet, if it were not for tidal distortion their centres would 
be two radii apart at the beginning of the impact. Owing to the distortion 
the distance will be rather greater than this. But, however oblique the 
impact may be, at the moment when the encounter begins the centres will 
be at least as near as when the surfaces meet in direct collision. In the 
case of a graze, then, the velocity destroyed is as great as in direct impact. 
