METEOROLOGY 7 



has been shown, the reader is referred to the original memoir, 

 for space does not permit of more than a brief summary of a 

 few of the resuhs obtained. Thunderstorms in which Hghtning 

 discharges took place immediately overhead were not experi- 

 enced, but several took place in the immediate vicinity and 

 were studied in considerable detail. It was observed that the 

 sudden changes of potential gradient near the ground that 

 accompanied each flash of lightning were more often + than — , 

 that is to say a + gradient before the flash was increased, 

 while a — gradient was either diminished or changed into a 

 + gradient. The ratio of + as compared with — changes was 

 3 to 2. The electric moment of a discharge varied very much, 

 the mean value found being of the order of 100 coulomb- 

 kilometres ; assuming that the discharge took place across a 

 distance of 2 km. on the average, there results the value 20 

 coulombs as representing the amount of electricity discharged 

 in a typical flash. The writer concludes that under these 

 conditions the potential difference across the two kilometres 

 may be considerably less than 6 x 10® volts, but is probably 

 greater than 10^ volts. The energy dissipated in a flash (| QV) 

 would be of the order of magnitude of io^° joules, and assuming 

 that such a flash occurs every 10 seconds, it is interesting to 

 compare this rate of dissipation of energy with the power 

 available, assuming that the rain of the thunderstorm is caught 

 at a height of i kilometre. Equality between the two is 

 obtained by assuming that rain falls at the rate of 10 cms. in 

 an hour over an area of 3 square kilometres, which is pro- 

 bably not in excess of what often takes place in actual storms. 

 The Quarterly Journal of the Royal Meteorological Society for 

 January 192 1 contains an interesting paper by Captain C. K. M. 

 Douglas, dealing with the variations of temperature in the 

 lowest four kilometres of the atmosphere. The author made 

 numerous observations of the temperature and humidity aloft 

 when flying in Northern France in 191 8 and 1919, and found 

 that the temperature was very much affected by the previous 

 course of the current in which the observations were made. 

 An investigation made by W. H. Dines some years ago showed 

 that whereas there exists a close connection between the 

 temperature and pressure in the free air, especially above 4 

 kilometres, the connection between the temperature and the 

 direction of the wind is insignificantly small. Douglas confirms 

 this result by a statistical treatment of his observations, but finds 

 that if the air is traced back along its course there is seen to 

 be a very considerable connection between the temperature of 

 the air and its position some days previously, currents from the 

 north being cold and those from the south warm. The latter 

 often curve round an anticyclone and so reach France as N.W. 



