» "Mav 26, 1923] 

iy 
i 
NATURE 
793 

understanding of the structure of the atmosphere in 
which the formation takes place. A feature of the 
atmospheric structure which is gradually asserting 
itself is the resilience of stratification due to the 
increase of potential temperature with height. We 
have always recognised isothermal structure as 
stable, still more so an inversion of lapse rate; but, 
when one thinks of it, it is clear that. the datum for 
stability is the lapse rate of the saturation adiabatic 
for upward movement, and that of the dry adiabatic 
for downward movement. Anything on the iso- 
thermal side of these lines implies stability and 
resilience upon displacement, gradually and con- 
tinuously increasing up to the isothermal condition 
and beyond that to inversion. 
Since the stratification is only disturbed by 
saturated air sufficiently warm, the successive layers 
of the atmosphere in ordinary conditions may be 
regarded as independent layers, easily capable of 
motion along surfaces of equipotential temperature, 
_ but unable to move up or down across those surfaces. 
In this respect the layers are like a pack of cards, 
deformable with a certain amount of resilience, very 
slippery, but not interpenetrable. The impenetra- 
bility of one layer by another is quite inexorable at 
the bottom, where there is a discontinuity of density 
between land or water and air, and at the top of the 
troposphere, where there is thermodynamic dis- 
continuity no less effective in the end, though the 
effort involves much greater sacrifice in the way of 
displacement required to produce the necessary 
resilience. Between these two extremes of resilience 
the surfaces of equipotential temperature are nearly 
horizontal. Expression is Gihaally given to the 
principle of resilience by regarding the motion in 
any layer as being limited to the horizontal. Of 
course, the limitation excludes the cases of penetrative 
convection which sometimes occur, and also the eddy 
effects due to the motion of a layer relative to the 
next above or below. But one is as rare as heavy 
rainfall, and the other, though never absent, is very 
small in magnitude. Thus for a first survey both 
may be left out of account. 
Also, to begin with, it is best to think of the atmo- 
sphere as made up of a number of layers of finite 
ickness, and not attempt the gradation of an 
infinite number of layers of infinitesimal thickness. 
There is often a natural sortigg of the structure into 
irregular strata; but, for the moment, let us think 
of twenty layers each half a kilometre thick between 
the ground and the stratosphere, the two boundaries 
the resilience of which must eventually balance the 
internal stresses of a quasi- permanent cyclonic 
vortex. 
The motion in each of the twenty layers except 
the lowest will adjust itself to the distribution of 
pressure in that layer. The law of the lowest stratum 
is different. In consequence of surface friction there 
is a flow across the isobars, with all the disturbing 
consequences thereof. 
As we regard the undisturbed medium as a pile 
of twenty horizontal layers, so we must regard a 
cyclonic system as made up of a number of in- 
dependent layers. We must consider the vortical 
motion produced in the medium as twenty separate 
rotating discs, not as a unified rotating column. 
There is practically nothing in the structure to 
prevent the slipping of one disc of revolving fluid 
over another to any extent. The unit of cyclonic 
‘activity is not a column reaching from earth to 
heaven, but a layer, say half a kilometre thick, with 
a mass of fluid revolving according to its own laws 
between the upper and lower boundaries. 
In these circumstances, if the centres of the revolv- 
NO. 2795, VOL. 111] 
| 
the velocity varied with height, would soon have 
i 
| 
| 
ing masses were accurately superposed to begin with, 
it would be a marvel of Nature if they remained so. 
Since pressure is transmitted, the displacement of 
one would alter the distribution of pressure for all 
beneath, the inflow at the bottom would affect the 
mass-distribution of those immediately above, and 
the relative motion at the surfaces of separation 
would add further complications. 
If all this be correct, what might once have been 
a vertical revolving column, in a stream of which 
become a number of separate and more or less 
degraded revolving discs. If there happened to be 
a thick enough layer in the original medium without 
any height-change of velocity, there might be enough 
organised rotational energy to preserve the original 
identity : otherwise the energy would soon be lost 
in eddies and ultimately in frictional heat. 
I am not sure whether Mr. Banerji (NATURE, 
May 109, p. 668) realises that that was the kind of 
structure which I had vaguely in mind three years 
ago when I wrote the remarks in Geophysical Memoir, 
19, to which he refers. The lapse of time has enabled 
me perhaps to think it out more clearly and to 
develop further the condition for no change of velocity 
with height. If one imagines a number of masses of 
saturated air overcoming the resilience locally, pass- 
ing through a series of superposed layers of air and 
removing therefrom automatically, by eviction, a 
vast quantity of air amounting in the aggregate to 
millions of tons, each layer must set up its own 
scheme of pressure and rotation independently of the 
others. As a means of making a rotating column 
the experiment could scarcely succeed if, by the 
time that the removal was complete, the centres of 
the upper systems were displaced horizontally a long 
way from the lowest; the superposed pressures and 
the incidental relative motions would certainly spoil 
the symmetry and in time destroy the unity of the 
system. 
The monsoon winds, of the peculiar behaviour of 
which Mr. Banerji speaks as being sufficient to 
account for the movements of the cyclonic storms. 
in Indian seas, belong to the lowest kilometre and 
are therefore not properly amenable to the distribu- 
tion of pressure. To my mind surface winds are 
primarily disturbers and destroyers of ordered 
vortical motion. 
succeed in feeding on their energy and thus increasing 
its own vigour in the way which Dr. Fujiwhara 
described recently to the Royal Meteorological 
Society. In Geophysical Memoir, 19, I was not 
dealing with that part of the subject, but only with 
the initial stages of the creation of the vortex. 
NAPIER SHAW. 
April 27. 
The Relation of Actinium to Uranium. 
At present the most likely view of the origin of 
the actinium series is that uranium II undergoes a 
dual change in which about 96 per cent. of atoms 
form ionium and the radium series, and the remainder 
form uranium Y, the product of which, proto-actinium, 
is the parent of the actinium series. An alternative 
view is that uranium I undergoes the dual change. 
In 1917 Piccard from a consideration of the Geiger- 
Nuttall relation put forward the view that the 
actinium series might arise from an isotope of uranium, 
actino-uranium, of atomic weight 240, present in 
ordinary uranium to the extent of about 8 per cent. 
Partly from experimental work carried out by Mr. 
W. P. Widdowson and myself, and partly from a 
| survey of certain general relations in radio-activity, 
A well-organised cyclone may, 
