806 
NATURE 
[JuNE 16, 1923 

Letters to the Editor. 
[The Editor does not hold himself responsible for 
opinions expressed by his correspondents. Neither 
can he undertake to return, nor to correspond with 
the wiiters of, rejected manuscripts intended for 
this or any other part of NATURE. No notice is 
taken of anonymous communications. | 
Gravitation and Light-pressure in Nebulz. 
Dors not Prof. Lindemann’s theory of spiral 
nebule (discussed by Sir Oliver Lodge in Nature of 
May 26, p. 702) fail through disregarding the absorp- 
tion or reflection of radiation which must necessarily 
accompany any mechanical action of light-pressure ? 
Prof. Lindemann’s typical nebula has a mass of 
about oI gram per sq. cm. of area, which is 
probably something like the true value; but to get 
this value, Prof. Lindemann’s assumed particles of 
diameters 10-4 or r0-* cm. must lie behind one another 
some thousands deep. The particles in the outer- 
most layer are, no doubt, acted on by light-pressure in 
the way supposed, but not so those in the inner 
layers; these are shielded from light-pressure but 
not from gravitation—and here the whole theory 
seems to fail. Incidentally, a nebula formed of solid 
particles lying thousands deep would surely be too 
opaque for nove formed in its interior to be seen as 
nove. 
The whole question seems to be governed by a 
calculation much simpler than any given by Prof. 
Lindemann. Ifa star or group of stars emit radiant 
energy E per unit time, the flow of momentum 
through any cross-section of a cone of solid angle w 
will be Ew/4rC per unit time. Thus the maximum 
mass m which light-pressure can possibly support (or 
repel) in this cone at a distance y from the light is 
given by 
where y is the gravitation constant, and M the mass 
of the star or stars. Thus the maximum mass per 
unit area, m/wy?, is 
Tie 
47yC M 
This is of course independent of v, because gravita- 
tion and light pressure both fall off as 1/y?; it does 
not depend on the number of stars at work, since E 
and M are each proportional to that number. For 
the aggregate of matter in the universe we may 
perhaps put E/M equal to unity, this being about 
half its value for our sun. The maximum mass per 
unit area which light-pressure can support or repel is 
now 1/47yC, or about one gram per 25,000 sq. cm., 
whereas the spiral nebule, on Prof. Lindemann’s 
own estimate, have one gram per Io sq. cm. No 
doubt it may be argued that the nebule in the past 
were more tenuous than now, but the nebule reduced 
to a surface-density of one gram per 25,000 sq. cm. 
would cover the heavens many times over at any 
reasonable distance. 
I have long wondered whether a true example of 
suspension in equilibrium between gravitation and 
light-pressure may not possibly be found in the shells 
which constitute the outer surfaces of the planetary 
nebule. That there is suspension in equilibrium 
seems scarcely open to question ; the hypothesis that 
gravitation and light-pressure are the equilibrating 
agencies satisfies all the numerical tests I have been 
able to apply. But it is necessary for stability that 
the matter should be gaseous, as is in fact spectro- 
scopically found to be the case. A hollow shell of 
NO. 2798, VOL. 111] 

solid particles can be suspended in equilibrium, but 
this equilibrium cannot be stable for radial displace- 
ments. Press the shell closer in to the star and its 
particles shield one another more from light-pressure, 
so that gravitation becomes relatively more potent 
than light- pressure and the shell falls in. But a 
hollow shell of gas may be stable if its temperature is 
such that its degrees of dissociation and ionisation 
are sufficiently sensitive to changes of temperature. 
Press such a shell closer in to the star and its 
constituents break up more; assuming that this 
increases its opacity, light-pressure becomes relatively 
more potent and the shell is driven back. Draw the 
shell out to a radius greater than the equilibrium 
radius and the converse occurs: the equilibrium is 
stable. J. H. JEANs. 
May 28. 
In the issue of Nature for May 26, p. 702, there is 
an interesting letter by Sir Oliver fdee about a 
suggestion I had put forward on the nature of the 
spiral nebule. Sir Oliver Lodge suggests that the 
recombination of electrons ejected photoelectrically 
might well be more important as a source of light 
than simple reflection. This does not seem to me 
probable. In spite of many efforts to improve 
photoelectric cells, the best type only gives an 
efficiency of about 2 per cent. Even if the spirals 
were equally efficient, therefore, only this fraction of 
the incident energy could be emitted in the form of 
light on recombination of the electrons. It is probable 
that any material of which spirals may be composed 
has a reflection coefficient of at least 20 per cent. and 
very likely 50 per cent., so that ordinary reflection 
must be at least ten times and is probably some 
hundreds of times as effective as the process imagined 
by Sir Oliver Lodge. F. A. LINDEMANN. 
Clarendon Laboratory, 
University Museum, Oxford, 
June 4, 1923. 

Selection and Segregation. 
I THINK by “ recent discussions in the columns of 
Nature” Prof. Arthur Willey (NATURE, May 5, p. 
602) alludes to discussions started by me. I gather 
that he believes that evolution is founded on muta- 
tions the inheritance of which is Mendelian, and, 
therefore, that natural selection preserves but does 
not create racial change. Experiment furnishes him 
with justification, but not with proof. On the same 
evidence, and a great deal more, divergent opinions 
have been founded. Crucial tests are required— 
tests which, from the nature of the case, experiment 
cannot furnish, but which may be found in abundance 
among facts that are within common knowledge 
and are not disputed. So far as I am able to judge, 
they do not support Mendelian theory, and to them 
I invite Prof. Willey’s attention. If I be mistaken, 
it should be easy to indicate my errors. If I be 
right (I am not alluding to Prof. Willey), is it worth 
while to ignore evidence which is common property 
and on which attention is more or less riveted, or to 
hint that I am ignorant of the latest work of the 
truest biologists—and then to scuttle from my 
challenge ? In spite of Mendelians, the Mendelian 
facts are. so illuminating that they are sure, when 
linked with the rest of our knowledge, to influence 
immensely our conceptions of, and control of, life. 
They can be so linked only by means of crucial in- 
stances. May I, therefore, iterate a few of the latter ? 
I have some morals to draw or imply. Can a flaw 
of fact or reasoning be found in the following ? 
