280 
LAWS OF ATMOSPHERIC MOVEMENTS.! | 
fee ee motion of the upper layers of the atmosphere 
is discussed in these two papers by Dr. W. N. 
Shaw, recently published. It is difficult in a short 
space to give a clear idea of the conclusions reached, 
but some of the main points may be here summarised. 
In the paper published by the Scottish Meteoro- 
logical Society it is shown that if p denote the pressure 
in millibars, and 6 the absolute temperature (C.), and 
Ap and Aé@ the changes that occur in passing hori- 
zontally from one place to another, then the rate of 
increase of pressure difference in millibars per 
metre of height is 0-0342 p/@ (A@/@—Ap/p). It is then 
shown that from about 1 km. to g km. the values 
of Aé and Ap have in general the same sign, so that 
the term in brackets is small, and hence pressure 
differences, i.e. the barometric gradient, are main- 
tained without much alteration up to g km. Above 
1o or 11 km. A@ and Ap have in general a different 
sign and the magnitude of the gradient rapidly falls 
off. The effect upon the wind at various altitudes is 
then considered, and special cases where simultaneous 
observations over England, Scotland, and Ireland are 
available are taken. 
In the paper published by the Royal Society of Edin- 
burgh, Dr. Shaw gives five axioms or laws of atmo- 
spheric motion, two lemmas or postulates, and six 
propositions. His first law reads thus :— 
‘In the upper layers of the atmosphere the steady 
horizontal motion of the air at any level is along the 
horizontal section of the isobaric surfaces at that level, 
and the velocity is inversely proportional to the separa- 
tion of the isobaric lines in the level of the section.” 
The whole discussion turns upon the truth of this 
law, and Dr. Shaw confesses that observation is at 
present incapable of proving or of disproving it. 
There can be very little doubt that it is approximately 
true, for except near the equator pressure differences 
come into existence and persist for days or even weeks. 
These differences could not continue even for an hour 
if there were not some compensating horizontal 
acceleration acting on the air from the low towards 
the high pressure, for otherwise the inevitable rule 
which makes the surface of a liquid horizontal would 
come into play, and a depression would be filled up 
almost as soon as it was formed. Some opposing 
acceleration must therefore act whenever and wherever 
there is a barometric gradient, and we can conceive 
of no other possible source of this acceleration 
save that given in law I. But in the upper strata 
there must be a certain small amount of flow outward 
across the isobars from low to high pressure to com- 
pensate for the inverse flow that occurs close to the 
earth, where frictional resistances prevent the re- 
quisite velocity along the isobars from being attained. 
The other laws and the two lemmas will probably be 
accepted with the small reservations given by the 
author without demur. i 
The propositions follow from the laws and postu- 
lates. They are of great interest, but are too long to 
be quoted here. It will suflice to say that Dr. Shaw 
finds that a current flowing east or west will be stable, 
but a current flowing north or south is more or less 
unstable, and must lose or gain air as it goes. Also 
his suggestion about the flow of air up or down the 
land slopes from the interior of the continents to the 
sea is very pertinent, and, to my mind, affords a 
better explanation of the winter anticyclone over Asia 
and North America than that commonly given. 
H. Dings. 
1 (x) ‘‘ Upper Air Calculus and the British Soundings during the Inter- 
national Week (May 5-10), ro13." From the Journal of the Scottish 
Meteorological Society. Third series. Vol. xvi. No. xxx. 
(2) “‘ Principia Atmospherica : a Study of the Circulation of the Atmo- 
sphere.” (Proc. Roy. Soc. Edin., vol, xxxiv., 1914, pp. 77-112). 
NO. 2324, VOL. 93] 
NATURE 
| battery. 
| May 14, 1914 
AN ELECTRICAL ANALOGY OF THE 
ZEEMAN EFFECT. 
(p82 discovery, announced by J. Stark in Nature 
of December 4, 1913, that when hydrogen in a 
state of luminescence is placed in an electric field of 
suitable strength and direction, the spectral lines are 
resolved into three or more components, is evidently 
a fact of prime physical importance. It will place in 
the hands of physicists another method of investigat- 
ing the internal structure of the atom, and, in con- 
junction with the Zeeman effect, will no doubt be of 
immense service in the discovery of further regularities 
in spectral series, and in the attempts now being made 
by Bohr, Nicholson, and others to explain the origin 
of spectra on a dynamical basis. In this connection a 
| series of papers in the Rendiconti della R. Accad. dei 
Lincei by Garbasso, Lo Surdo, and Puccianti will 
be of great interest to readers of NATURE. 
The effect appears to have been observed independ- 
ently by Lo Surdo whilst working on the retrograde 
positive rays in the neighbourhood of the kathode. 
An account of his first observations is given in a 
paper read on December 21, 1913. A cylindrical tube 
20 cm. long and 4 mm. in diameter was used. It 
carried disc electrodes which completely filled the sec- 
tion, and it was excited by means of an accumulator 
In these circumstances it was found that the 
electric field in the Crookes dark space was of itself 
sufficient to produce resolution of the lines. The ob- 
servations were made with a four. prism quartz spectro- 
graph. By suitable modifications the tube was varied 
so that the line of sight was either along or per- 
pendicular to the field. The two outer components 
are polarised with the electric vector parallel to the 
field, the remainder in a perpendicular plane. When 
observations are made along the lines of force the 
outer components are missing. In a later paper, using 
a tube 1-5 mm. in diameter, Puccianti finds that the 
effect can readily be seen in the well-known Hilger 
wave-length spectroscope with constant deviation 
prism. The lines of the hydrogen spectrum show an 
interesting series of regularities; these are displayed 
in the following tabie (after Lo Surdo). Writing the 
Balmer formula in the form, 1/A=a—4a/n?, n has 
the values 3, 4, etc., for the different lines. 
H H, ial jas 
n ah fe ane 3 4 5 6 
Total no. of components 3 4 5 6 
Order) of line in the 
SGMESM EM gen fics. as Ist. ead |. Bra 4th 
Component with electric 
vector | to the field... I 2 3 4. 
Appearance of resolved) | | _ | 
Vas! 23) as | | 
Pot Nal | 
I HT 
It is seen that the number of components in a given 
line is the same as the corresponding value of n in the 
Balmer formula, and that the number of internal 
components is the same as the order of the given line 
in the spectral series. According to the measurements 
of Puccianti the separation of the outer components 
for Hg and H, are in the ratio 1-49, or, expressed as 
fractions of the corresponding wave-lengths, 
2s fb. 
In a paper of December 21, 1913, Garbasso dis. 
cusses the matter theoretically, to. arrive at the con- 
clusion that the Thomson model atom is incapable of 
explaining the (earlier) observations, except by the 
introduction of improbable hypotheses. 
R.S W, 
1 The components placed above have the vector parallel to the field. 
