8 9 o 



NA TURE 



[December 30, 1922 



number of periods — no attempt had been made to 

 make the table complete — it would not be difficult 

 to find a period near any specified period, especially 

 if one were allowed to consider multiples and sub- 

 multiples. He then discussed the two chief meteoro- 

 logical cycles, the Bruckner cycle and the sunspot 

 cycle. The hundred-year record of London rainfall 

 has been analysed for a 35-year period, and a curve 

 added to a diagram of monthly totals to show on 

 the same scale the contribution of the cycle to the 

 total rainfall. The amplitude of the cycle is absolutely 

 insignificant in comparison with the monthly varia- 

 tions. Dr. Simpson admitted that there is an 

 appreciable correlation between sunspots and meteoro- 

 logical factors, but as sunspots have no true periodicity 

 they cannot introduce a periodic term into meteoro- 

 logical phenomena. 



Turning to Sir William Beveridge's results Dr. 

 Simpson regretted that he had not seen Sir William's 

 recent paper in the Journal of the Statistical Society 

 but only his papers in the 'Economic Journal, because 

 the periods on which Sir William appears now to 

 rely are different. He was prepared to admit that 

 Sir William Beveridge had discovered certain 

 periodicities in his curve of prices of wheat which 

 were many times greater than one would expect 

 bv chance, but he strongly contested that these were 

 meteorological periodicities. Sir William Beveridge 

 laid great stress on a periodicity discovered by Capt. 

 Brunt in Greenwich temperature, 5-1 years, which 

 coincided with one of his cycles, but it was pointed 

 out that Capt. Brunt discovered 9 cycles, four of 

 which had greater and four smaller amplitudes than 

 this particular cycle. Also Capt. Brunt's cycle of 

 5-1 years reduced the standard deviation of mean 

 monthly temperatures at Greenwich only from 2-80 

 to 2-77, an insignificant change. 



Dr. Simpson also criticised Sir William Beveridge's 

 synthetic curve and asked why that curve should be 

 compared with rainfall. There appeared to him no 

 more reason why it should apply to rainfall than to 

 any other meteorological or economic or even bio- 

 logical factor which might conceivably affect a 

 harvest. In conclusion, admitting all that Sir William 

 claimed to have done, he did not think that a predic- 

 tion which gave a correlation coefficient with actuality 

 of only 0-38 had any practical value. When Sir 

 William had increased his correlation coefficient to 

 abouf 083 he would be a valuable forecaster, but 

 not until then. 



Mr. Udny Yule said that the comments of Dr. 

 Simpson seemed to him unfair. It must certainly 



be recognised that mere inspection of data was 

 wholly inadequate and might lead to unfounded 

 ideas as to the existence of periodicities, but this 

 criticism had no bearing on work carried out by the 

 periodogram method. He felt a good deal more 

 doubt than some previous speakers on the question 

 whether crop cycles were or were not a vital factor 

 in the general economic cycle, which required far 

 more study. Front the statistical side the most 

 important work now to be done is the determination 

 of the crop cycles in areas other than Western Europe, 

 e.g. South America and India : in so far as crop 

 cycles are an important factor in the economic cycle, 

 the resultant in any one country must be a complex 

 effect dependent on the sources of its raw materials. 

 On the side of economic theory it seemed to him 

 there is also work to be done. The treatment of 

 economics is in general static. The economist is 

 too apt to tell us that " in the long run " a pendulum 

 will hang vertically, whereas the whole interest of 

 the pendulum is that it swings, and the problem is 

 why it swings and how it swings. The treatment 

 of economics should be dynamic. The question 

 might be asked, for example, whether there is not 

 an equation relating production not merely to price 

 but to price and its time differentials, an equation 

 which might (or in given circumstances might not) 

 have a periodic solution. 



Prof. H. H. Turner considered that we should be 

 grateful to Sir William Beveridge, first, for producing 

 a long series of annual values, going back much 

 further than our longest rainfall record ; secondly, 

 for having himself analysed them completely by the 

 periodogram method, so that others can profit by 

 his analysis ; and thirdly, for two considerable 

 successes in the outcome. One of these is remarkable. 

 He had succeeded in forecasting the weather in some 

 sort — a rare, if not unique, achievement up to the 

 present. The other success consisted in isolating 

 several periods which must be further investigated. 

 The periodogram gives us only the beginning, not 

 the end of an investigation. Having obtained, for 

 example, the definite suggestion of a 15-day period 

 we must then see how it behaves throughout the 

 series ; the maximum phase seems to oscillate in 

 this case. Such oscillations frequently occur in 

 manifestations of periodicity which may itself be 

 quite regular ; thus, the rotation of the earth is 

 quite regular, but one of its manifestations is sunrise, 

 which swings to and fro. Sir William Beveridge had 

 given us a good start, which it is to be hoped will be 

 followed up. 



Geology of the North Sea Basin. 



'"TMTE long-standing custom of devoting at least 

 ■*• part of a session of the Geological Section of 

 the British Association to matters pertaining to the 

 geology of the district in which the meeting is being 

 held, was extended this year to the consideration of 

 the wider question of the geological history of the 

 North Sea basin, the discussion on this subject being 

 the first of a series held in different sections on various 

 aspects of the North Sea. 



The discussion was opened by Prof. P. F. Kendall, 

 president of the section, and was continued by Mr. 

 J. O. Borley, of the Fisheries Research Laboratory, 

 Lowestoft, who described the nature and distribution 

 of the deposits now being laid down. Mr. Thomas 

 Sheppard dealt with the geology of the Hull district, 

 and Mr. C. Thompson contributed an interesting paper 

 on the present rate of erosion of the coast of Holderness. 

 The main tectonic lines of the British Isles and of 



NO. 2774, VOL. I IO] 



the neighbouring area, the North Sea, were produced 

 prior to the formation of the Permian rocks, the 

 three main axes of folding being the Caledonian 

 (N.E. and S.W.), the Pennine (N. and S.), and the 

 Armorican (W. and E. approximately). Later 

 movements, for the most part along these old lines, 

 were responsible for the changes in the distribution 

 of land and water which have taken place. 



The region now occupied by the North Sea appears 

 to have been an area of depression since a very remote 

 period. Thus it is found that movements which took 

 place during late Carboniferous times and during 

 the period, unrepresented by any deposits in Europe, 

 that elapsed before the deposition of the Permian 

 rocks, caused the coal measures to dip into the basin 

 in Holland and Belgium, in Northumberland and 

 Durham, and probably also in Lincolnshire, to re- 

 emerge at Ibbenburen. 



