40 The N.Z. Journal of Science and Technology. [Jan, 
Bickerton. Stated very briefly, this theory is that in the great majority 
of cases a new star is due to a grazing collision between two suns. In the 
case of a slight graze the two wounded suns pass on, but the portions 
actually coming into impact form an intensely heated and absolutely 
unstable third body. It is this third body which, in the dramatic early 
stages, gives practically all the light; it is its spectrum which is seen 
and recognized as the spectrum characteristic of a nova, and it is the dissi¬ 
pation of this dynamically unstable third body into space that causes the 
otherwise inexplicable sudden fall in brightness. 
Theories which attribute the sudden decline to cooling are incredible. 
The third body is too hot to hold together. If two dead stars met directly, 
moving in the same straight line, there would be no third body and no two 
wounded suns. A new star would be produced in an hour, would increase 
in brightness with astonishing rapidity, and would shine as a permanent 
star for a hundred million years. But this, of course, could only happen 
if the two stars were initially at rest, or else moving initially in the same 
straight line. Collisions may, and doubtless do, occur between all the 
different kinds of bodies in the Universe, and if a collision theory of 
novae is adopted the question to be considered is which, if any, of all 
the possible forms of collision would give rise to exactly those phenomena 
which are observed in the case of a typical new star. 
A. C. G. 
On the Character of the Spectral Lines due to a Sphere 
or Spherical Shell of Transparent Luminous Gas ex¬ 
panding uniformly. 
Let AB be the diameter of the sphere which points directly from the 
observer. Divide AB into a very large number of equal parts, AC, CD, 
DE, EF, &c. Draw planes perpendicular to OA through the points of 
division, thus dividing the surface of the sphere up into zones of equal 
area. Taking the centre of the sphere as vertex, construct the cones 
OPAQ, OXBY, and the conical shells 
PQSR,0, &c. The cones and shells 
are all of equal altitude — viz., the 
radius of the sphere—and stand on 
equal bases, therefore they are all 
equal in volume. 
If we represent the radial velocity 
of the expanding gas by OA, the re¬ 
solved part, in the direction of the 
observer, of all the gas within the shell 
PQSR.,0 is represented by lines vary¬ 
ing in length from OD to OC, and 
that within the shell BSUT,0 by lines 
varying in length from OE to OD. 
We have thus divided the whole quan¬ 
tity of gas into portions which are 
equal in mass and whose velocities 
towards or from the observer differ by 
equal amounts. Now, the displacement of a spectral line is proportional to 
the relative velocity of approach or recession of the source of light and the 
observer; so the shell or sphere of luminous gas will give rise to a spectrum 
in which each band will be equally bright all over and sharply defined at 
each side. This fact is useful in discussing the spectra of new stars. 
A. C. G. 
A 
