July 7, 1904 j 



NA TURE 



227 



from a suitable datum temperature or pressure, the 

 Jines might be interchanged, a step of one degree of 

 temperature (Fig. A corresponding to a step of one 

 millimetre of pressure in the same direction for the 

 unner layer (Fig. 2), and in the opposite direction for 

 the lower layer (Fig. 3). 



The complexity of the surface pressure arises, there- 

 fore, not from the upper layer alone, nor from the 

 lower layer alone, but from the superposition of the 

 two. We can resolve the surface pressure into two 

 components, one due to the upper stratum above 4000 

 metres which, if it acted alone, would produce a 

 general circulation from west to east around minima 

 of pressure near the poles ; the other, due to the lower 

 stratum, which, if it acted alone, would produce a 

 circulation from east to west. Both circulations would 

 correspond closely with the surface distribution of 

 isotherms. Where the one is predominant, in the 

 lower middle latitudes, we get resultant westerly circu- 

 lation ; where the other is predominant, near the poles 

 of cold, we get an easterly circulation. Between the 

 two we get a region of minimum pressure and a 

 merging of the two circulations which gives rise to 

 the circular storms of the northern and southern 

 temperate zones. 



It appears, therefore, that we ought to regard the 

 surface distribution of temperature as giving rise to 

 a distribution of pressure in the lower stratum tend- 

 ing to maintain a circulation of air from east to west 

 round the poles of cold. Extending this idea, a region 

 of cold in the northern hemisphere should tend to 

 maintain a clockwise circulation round its centre in 

 the lower atmosphere, and a region of heat a counter- 

 clockwise circulation. 



The reciprocity between the pressure distribution of 

 the upper and lower layers is of course not fortuitous. 

 Hann has shown that the expansion of the lower layer 

 by heat increases the pressure at a given level in the 

 upper regions, without altering the pressure at the 

 surface, by the mere thrusting of part of the air 

 upwards ; so that the observed effect of expansion over 

 a large area is to diminish the pressure of the lower 

 stratum and increase, by an equal amount, that of the 

 upper. Referring to the diagrams again, the effect of 

 increased surface temperature upon the isobars of 

 Fig. 3 would be a bulge of the isobars towards the 

 region of low pressure — the equatorial regions ; upon 

 the upper isobars there would be a corresponding 

 bulge towards the region of higher pressure, again the 

 equatorial regions. Thus the lines of both diagrams 

 would be affected geographically in an exactly similar 

 way and to the same extent ; they would thus preserve 

 their similarity in spite of temperature variations at 

 the surface. 



It would be interesting to consider what the effect 

 of the daily solarisation of the earth should be from 

 this point of view. Primarily it should produce no 

 pressure variation at the surface; but inequalities of 

 motion in the upper and lower air would probably 

 alter the relative phase or magnitude of the disturb- 

 ance of the two components, and hence give rise to 

 daily variations of pressure at the' surface, and thus 

 necessarily produce a diurnal variation of the baro- 

 meter. 



Other consequences follow from the treatment of 

 the distribution of pressure due to the weight of the 

 lower layer as producing, or rather maintaining, a 

 circulation in the one direction or the other about the 

 colder or warmer regions, as the case may be, instead 

 of flow from cold regions to hot. 



One important result as regards the formation of 

 circular storms in our latitudes mav be inferred from 

 this method of analysing the distribution of surface 

 pressure. Friends have frequently suggested to me 



NO. 18 10, VOL. 70] 



that our circular storms are like the eddies formed 

 when water flows through a bridge ; and to them I 

 have alwavs put the question. What in the atmosphere 

 stands for the bridge? I am now prepared to re- 

 cognise that the caps of relatively cold air in the north 

 and south polar regions form an adequate represent- 

 ation of the piers of the bridge. In the lower air, 

 where the pressures of the polar caps are. dominant, 

 thev stop the westerly currents which still flow in 

 lower latitudes, and replace them by currents from 

 the east. Between these two currents is a field where 

 mixing must take place, and circular eddies may be 

 formed. 



What happens in the equatorial regions is another 

 story. Buys Ballot's law shows that the equator is 

 subject to a peculiar meteorological condition. If you 

 stand with vour back to the wind north of the equator, 

 the low barometer is on your left ; south of the 

 equator, it is on your right. There must be a 

 transition region where the law ceases to apply, as, 

 indeed, one would expect if Buys Ballot's law is the 

 practical expression of motion with an acceleration 

 due to the rotation of the earth, and varying as the 

 sine of the latitude. 



In the upper air of the equatorial regions there is 

 probably a persistent flow from the east, as shown by 

 observations of clouds and of the Krakatoa dust. In 

 respect of the formation of eddies, this current will 

 act like an intermediate pier of a bridge. Hence, in 

 January, the river in which, upon this analogy, atmo- 

 spheric eddies mav be expected is a stream of air flow- 

 ing round the earth in middle latitudes, divided by the 

 equatorial belt with its region of doldrums below and 

 easterly current above, and bounded north and south 

 by easterly currents which correspond with the circula- 

 tion of the lower atmosphere induced by the pre- 

 dominant influence of the polar caps of cold air. 

 Eddies may be looked for between the easterly and 

 westerlv currents, and they are sometimes found there. 



W. N. Sh.\w. 



nOTES. 



The trustees of the Carnegie Institution met on May 18 

 and transacted the necessary business to provide for the 

 transfer of all matters to the Carnegie Institution of 

 Washington, a charter for which passed Congress and was 

 approved on April 28. The trustees named in the Act met 

 at once and reorganised under the new charter. The 

 by-laws of the Carnegie Institution were adopted as the 

 by-laws of the new organisation, and the officers of the 

 old organisation were elected. General resolutions adopt- 

 ing all the obligations, &c., of the old institution were 

 passed. Under the new charter no questions can be raised 

 as to the competency of the institution to carry on the 

 operations outlined in the deed of gift of the founder. The 

 executive committee of the Carnegie Institution of Washing- 

 ton met after the reorganisation, and practically completed 

 the making of grants for the year 1904. It will greatly 

 facilitate the work of the executive committee if all those 

 thinking of making applications for grants for 1905 will 

 do so not later than September, as applications for grants 

 for 1905 will then be taken up. 



A SLIGHT but decided earthquake, which lasted about a 

 second, was e.xperienced at Derby at 3.22 p.m. on July 3. 

 The vibration was not nearly so pronounced as on the 

 occasion of the seismic disturbance a year ago. A similar 

 shock was felt in the mid-Cheshire district about the same 

 time. In Leftwich, near Northwich, the tremors were very 

 distinct. The shock, which was also felt in Northwich, lasted 

 two seconds, and at Sandbach four seconds. Two shocks 



