WHEAT-GROWING AND ITS PROBLEMS 291 



and finally no further yield is obtained, no matter how much 

 water is added, the limit being now set by some other factor. 

 With excess of water there is an actual fall in the yield, 

 not because the underlying law is different but because of 

 secondary disturbing influences. A large amount of water in 

 the soil necessarily diminishes the supply of air and keeps the 

 temperature down, so that it is necessary to have some sort of 

 balance between these three factors. Further, the plant itself 

 does not always need the same amount of water. More water is 

 wanted in the period of active growth than in the seedling stage 

 or at the time of ripening. Thus, whenever wheat is sown in 

 late autumn it has been known that wet winters are bad for the 

 crop, while dry ones are good. " Under water, famine ; under 

 snow, bread," runs an old proverb. Shaw has recently shown 

 that the relationship between autumn rainfall and crop-depres- 

 sion is almost mathematical. Over a period of years the average 

 crop in England varies above or below a certain value in inverse 

 proportion to the rainfall of September, October and November, 

 the yield in bushels being — 



39' 5 bushels - f x rainfall in inches. • 



The fatal effect of rain when the grain is forming or at 

 harvest time has always vividly impressed itself on the farmer's 

 mind. Anything indeed was better in old days than too much 

 rain, and the dread of wet seasons reveals itself in a score of 

 proverbs such as : " Drought never bred dearth in England." 

 The depressing effect in both cases is complex but open to 

 analysis : winter rains wash out the nitrates from the soil and 

 thus militate against the development of a full root system, 

 without which a heavy crop cannot be secured. Autumn rains 

 keep the plant growing too long and retard the ripening pro- 

 cess. The plant therefore becomes very liable to the attacks of 

 various fungi ; its straw does not stiffen, so that it becomes laid 

 and is expensive to harvest ; the grain, even when reaped, is not 



' Proc. Roy. Soc. February 1905. The rainfall is for the 36th to 48th weeks 

 inclusive. In Proc. Roy. Soc, May 1906 Dr. Shaw gives the formula for the 

 eastern counties as 46 bushels — 2*2 x rainfall in inches over the same period. 

 He further shows that in these counties the yield of wheat for the twenty-one 

 years 1885 to 1905 is very nearly identical with the sequence given by a periodic 

 variation of eleven years and its five harmonic components. This relationship 

 is even closer than that shown by the rainfall. 



