TRANSACTIONS OF SECTION A. 641 



(3) Piling up of water at one end of a lake by winds. In deep lakes piling up 

 of water by wind is small, owing to the readiness with which the return current 

 takes place, and may be neglected in a consideration of the causes of seiches. 



(4) Rapid flooding at one portion of a lake may in exceptional circumstances 

 produce a seiche. 



(<■)) A heavy rainfall may produce a seiche. 



(ff) By the gravitational effect of the precipitation. 

 (b) By the force of its impact. 



(6) The impact of wind-gusts on portions of the lake may also cause a 

 seiche. 



The foregoing are cau.ses which might produce seiches suddenly. The fol- 

 lowing causes might account for a gradual increment in the seiche amplitude: — 



(i) Microbarometric fluctuations over portions of the lake surface of a 

 period approximating to one of the seiche periods of the lake. The 

 disturbing effect may be considerable even where there is no con- 

 siderable disparity between the barometric period and the seiche 

 period. 



(ii) Periodic fluctuations in the direction and pressure of the wind. It is 

 difficult in any individual case to say what is the precise effective 

 cause of a seicue, as Avind, rain, and barometric variations usually 

 occur together. 



7. Difference of Temjierature in the Tlfpei- Atmosphere between Equatorial 

 and Polar Regions. By L. Tejsserenc de Bort. 



In former discussions of the observations made over the Atlantic on the 

 ' Otaria,' and those made at Kiruna, near the Arctic Circle, by the French- 

 Swedish expedition, I have shown that the difference of temperature between the 

 Equatorial and the Arctic regions is very different at different altitudes. In the 

 lower layers the temperature of the equatorial zone exceeds that of the Arctic 

 regions by 25° C. This excess of temperature decreases as the altitude increases 

 imtil at a height of 10 or 11 km. it is as warm over the Arctic Circle as it is over 

 the equatorial zone. With further increase in height there is no further decrease 

 in temperature over the Arctic Circle, because the isothermal region is reached. 

 But in the neighbourhood of the equator the rapid decrease of temperature with 

 increasing altitude still continues ;. the temperature becomes less therefore over 

 the equator than it is at the same height over the Arctic Circle. 



Recent balloon ascents made this year at Kiruna at the end of the winter give 

 a general confirmation of these conclusions. They show that at a very cold period 

 with temperatures below —15° C. near the surface, the temperature in the upper 

 regions is nearly the same as that found over the equatorial zone. There are 

 indeed some days when the polar temperatures are lower than the equatorial, but 

 this is a temporary phenomenon which occurs at the coldest time of the year. In 

 equatorial regions, on the contrary, if one considers as the point of departure the 

 region where the trade winds meet, near the thermal equator, there is, strictly 

 speaking, neither winter nor summer. 



No ascent to great heights has hitherto been made in Arctic regions during 

 the warm season : we propose finally to fill up this gap. But what we know 

 already of the decrease of temperature at different places, together with the obser- 

 vations made at St. Petersburg, enables us to count on temperatures sufficiently 

 high in the upper layers, probably higher than —50° C. 



Thus during the greater part of the year it is sensibly colder by 10° or 20° C. 

 in the equatorial regions at altitudes of 15 or 16 km. than in Arctic regions. This 

 fact, anticipated by me some years ago, deserves to be taken into serious con- 

 sideration in theories relating to the general circulation. 



1908. T T 



