LOO REPORTS ON THE STATE OF SCIENCE. 
surface. The difficulty that arises is to discover a means by which air 
can be brought into the anticyclone and out of the cyclone in the upper 
air, and to make these results accord with the results of cirrus observations, 
which imply a definite outward motion over cyclonic regions and an inward 
motion over anticyclonic. At the same time it must be remembered that 
the cirrus observations do not imply that the anticyclone becomes a 
cyclone at the cirrus level or conversely ; the direction of rotation is the 
same for the cirrus as at the surface according to Hildebrandsson’s 
results, and this can be the case only if the direction of the gradient of 
pressure remains the same. 
The results imply that the motion has a component across the isobars 
from the lower to the higher pressure. The difficulty of explaining this 
result was felt by earlier writers. Hann! expressed the opinion that the 
outward motion in cyclones was due to the centrifugal force of the motion 
exceeding the gradient. Although it is difficult to see how the necessary 
wind would be produced to bring about this state of affairs, it is at least 
a possible condition. If cyclones decreased in intensity with increasing 
height, and the air rising from the lower levels retained its angular 
momentum, it would indeed furnish a reasonable explanation. 
The case of anticyclones is more difficult because the effect of centri- 
fugal force is to assist the gradient of pressure in producing flow outwards. 
Gold? showed that in anticyclones there is a limit to the gradient and 
velocity for the motion to be steady and along the isobars. The approxi- 
mate radii of isobars at the earth’s surface differing from the pressure at 
the centre by 1, 2, 3, 4, 5 mm., are 260, 370, 450, 520, 580 km. for this 
limiting case. If the gradients are less than these, there will be a steady 
motion with correspondingly small velocities. If the gradients are greater 
than these, the motion cannot under any circumstances be steady, and 
there will always be an outward component in the wind, because the 
centrifugal force due to the increased velocity will more than counter- 
balance the increase in the force arising from the earth’s rotation. The 
only possible case where there can be flow from low to high pressure for 
anticyclonic motion is when air enters a region where the gradients are 
less than the limiting gradients, with a velocity also less than that corre- 
sponding to the limiting gradient, but greater than that corresponding to 
steady motion for the actual gradient in the region. In that case the 
effect of the earth’s rotation would be to make the air flow inward towards 
the centre. It seems improbable, however, that such a state could persist 
for any time, because the results of observation show that the wind usually 
adjusts jtself to the gradient, provided it is at a sufficient height above the 
earth’s surface to be practically free from the effects of surface friction and 
irregularities. 
It seems more probable either (1) that anticyclones and cyclones 
arriving in the European area are in general dissipating systems, which 
are replaced continually by other systems arriving from what may be 
called productive regions, or (2) that there is interchange with regions in 
which the surface temperature or the temperature gradient is sufficiently 
different to produce mean temperatures greater in low-pressure areas and 
less in high-pressure areas than are found over Europe. 
. The results of observations of pilot balloons at Ditcham,* July 27-30, 
1908, and at Munich,‘ during thesame period, and September 30 to October 2, 
' Lelrbuch, p. 406. © 
2? Barometric Gradient and Wind Force, M.O. No. 190. 
® Quart. Journ. Roy. Met. Soc., 1908. 4 Registrierballonfahrten. 
