PISTIMI'.ITIOX IN TIIK ATM' >Sl'l!Kl;K (VKU i:\dl. AND. 275 



.-mil \\:is falling with unusii.-il rapidity. The wind was east, and it was raining 

 heavily. Tin- (litre of tin- depmnon, which came from W.N.'W., passed directly 

 over in the evening. A balloon was sent up in the evening, but not recovered. At 

 7 a.m. on Noveml)er 4th a third balloon was sent up, the l>;m>meter being at 742 nun. 

 and rising rapidly. The sky was dear save for a little cirrus, and the wind S.W. 

 It had shifted through N.E. and N.W. The two balloons, 2.30 p.m. and 7 a.m., 

 were recovered, and both gave specially clear records, but, excepting that up to 5 km., 

 the afternoon temperatures were a few degrees the higher, the temperatures recorded 

 \\ere practically identical. In this instance, at least above 5 km., the front and back 

 segments of a cyclone were alike. Both balloons reached 16 km. and travelled 

 (JO miles to tin- eastward, one falling at Chelmsford and the other at Gravesend. 



In discussing the mechanics of a cyclone, it is absolutely necessary to consider the 

 question of barometric pressure. It will be seen from the diagram how the low 

 pressure of a cyclone lies under a column of cold and therefore heavy air, and this 

 shows how fallacious it may lie to explain high or low pressures by the presence of 

 cold or warm air above them. Also it must be remembered that the atmosphere is 

 perfectly free and unconfined save at its lower boundary. If two large sealed vessels 

 communicate with each other by even a small pipe, it is impossible to maintain 

 any difference of pressure between their gaseous contents unless some force acts 

 tangentially on the gas in the connecting pipe, as, for example, when one vessel is 

 above the other and gravity alters the pressure. Much more so, therefore, with the 

 atmosphere. Suppose a certain pressure at a point A and a lower pressure at a point 

 C, and to exclude gravity let A and C be on the same level. If this difference of 

 pressure is to be maintained a sufficient force must act along every possible line of 

 communication between C and A, that is to say, along every line that can be drawn 

 from A to C without passing through the earth or leaving the upper limit of the 

 atmosphere. Failing this force the pressures at A and C will be equalised in some- 

 what the same time that it will take for sound to travel along the unopposed line of 

 communication between A and C. At all events, the velocity should be of the same 

 order as that of sound. This statement hardly seems to me to require proof, but if it 

 does the following facts may IKJ quoted. The wave of pressure produced by the erup- 

 tion of Krakatoa passed round the earth with the velocity of sound. On the equator, 

 where there is no possibility of any opposing force, there are hardly any appreciabfe 

 differences of pressure, save those due to the semi-diurnal pressure wave which travels 

 with a velocity of about four-thirds of that of sound. The nature of the force that 

 acts between A and C in parts of the earth away from the equator was pointed out 

 by FERRBL. If C be the centre of a well-developed cyclone, it is largely supplied by 

 the centrifugal force of the curvilinear winds that blow round C, but in moat cases it 

 is due to the tendency of a moving body to turn to the right hand (in the northern 

 hemisphere), which again is due to the earth's rotation. (See a paper by Mr. GOLD, 

 ' Barometric Gradient and Wind Force,' M.O. 190.) Briefly, we may say that a 



2 N 2 



