September i, 1921] 



NATURE 



tion, 1913), is that in practice the energy changes 

 do not take place by scattered radiation alone, but 

 also by corpuscular radiation and characteristic or 

 fluorescent radiation. It does not seem permis- 

 sible to consider scattered radiation by itself. The 

 genesis of radiation must involve the mutual play 

 of both corpuscular radiation and waves. When 

 X-rays fall on a body some of the incident energy 

 reappears as scattered radiation, some as corpus- 

 cular radiation, and some as characteristic radia- 

 tion. Consequently Planck's original oscillators 

 formed an artificial body which has no counter- 

 part in reality. He was, of course, aware of this, 

 for on p. 133 he states that " it does not matter 

 whether such a body exists anywhere in Nature, 

 it is only necessary that its existence and proper- 

 ties should be compatible with the laws of elec- 

 trodynamics and thermodynamics." 



As a result of the difficulties associated with the 

 form of the theorv described in the book referred 



to above Planck made an important modification 

 of his hypothesis (" Eine veranderte Formulierung 

 der Quantenhypothese," Freiiss. Akad. Wiss. 

 Berlin, Ber. 34, pp. 918-23, 1914). This paper 

 assumes that radiation and absorption take place 

 continuously, and that the quantum action is not 

 between the oscillators and the radiation, but 

 takes place between the oscillators and the free 

 particles (molecules, ions, and electrons), which 

 exchange energy by impacts with the oscillators. 

 The laws of classical electrodynamics then hold 

 good for every interchange between the oscillators 

 and free radiation. At the same time the radiat- 

 ing substance becomes more like its counterpart 

 in Nature, and the feeling of artificiality which the 

 former theory produced is removed. Also the diffi- 

 culty connected with the use of Hertz's expres- 

 sion for calculating the density of the radiation 

 disappears. 



(To he continued.) 



The Extent of the Recent Drought. 



THE recent prolonged drought in the British 

 Isles has directed attention to an interesting 

 aspect of meteorological science. It is natural to 

 inquire how far the drought has been confined to 

 our immediate neighbourhood, or how far it has 

 been general. With the exception of Hilde- 

 brandsson's pioneer work on action centres, no 

 systematic research dealing with the extent to 

 which drought has affected considerable areas of 

 the earth's surface at one time has yet been 

 carried out. A basis for detailed study of this 

 character will be provided by the "Reseau Mon- 

 dial," published by the Meteorological Office, five 

 annual volumes of which have now been issued. 

 This publication gives pressure, temperature, and 

 rainfall for about 400 stations distributed over the 

 globe, the month being taken as a unit. In the 

 present article it is proposed to make a pre- 

 liminary survey, so far as material is already 

 available, of the world's weather this year, par- 

 ticularly during the months May, June, and July. 

 As no system of telegraphic reporting from 

 "Reseau Mondial" stations has yet been estab- 

 lished, we have to rely in making such a survey 

 on the most recent monthly, weekly, or daily 

 weather reports obtainable from the various 

 countries, and, largely, upon general newspaper 

 reports. 



Table I. shows the percentage of normal rainfall 

 which has fallen in various parts of the British 

 Isles since the beginning of the vear : — 



Table II. gives the percentage of normal rain- 

 fall for the various districts into which the 

 British Isles are subdivided : — 



Table II.- — Percentage of Normal Rainfall by 

 Districts. 





■^ ■ — . -7 



go o a 



C Tt V O 



caZ tnW WZ fc^ «s:_/ ,i X t/ifi fcZ i^ 



Januar)- 164 162 144 107 128 1 18 165 174 123 132 



February 64 26 16 26 15 21 37 14 9 38 



March... 160 109 45 53 73 62 176 124 92 127 



April ... 61 51 69 87 57 63 56 62 43 51 



May ... 117 89 86 59 70 71 100 83 89 89 



June ... 46 41 27 20 18 6 30 15 13 17 



Average percentage February to June : — 



90 63 49 49 47 45 80 60 49 64 



— j: 



99 

 60 

 114 

 30 

 «3 

 10 



95 



'4 



53 

 73 

 36 



59 49 



It should be noted that Tables I. and II. are 

 not based on identical stations. 



Table I. shows that January was a month of 

 excess rainfall in all regions. Previous to this we 

 have to go back to July, 1920, to find another 

 month with rainfall above normal for the whole 

 British Isles, the percentages for August to 

 December, 1920, varying between 68 and 96. It 

 is evident from the table that the drought has 

 been much more conspicuous in England and 

 Wales than in Scotland and Ireland, where it has 

 not been so remarkable. This is well shown in the 

 map (Fig. i), which has been prepared by the 

 British Rainfall Organization. The area of great- 

 est drought is the southern and eastern midlands, 

 the amount of rainfall increasing outwards from 

 this centre, particularly to the north and west. 

 February, April, and June were the months of 

 greatest deficiency. March, which appears to be 

 normal (loi per cent.), was a month of drought in 

 most places in the eastern and midland counties, 

 but wet in the west and north-west. 



NO. 2705, VOL. 108] 



