326 



DISCOVERY 



found and a pit is dug, which gets so large that 

 pumping often has to be resorted to. The Celestine, 

 being comparatively soft, is then picked out b}^ 

 hand, cleaned and stacked in neat square heaps : 

 some 20,000 tons before the war were exported 

 annually to Germany ; now the bulk of it goes to 

 the United States. Associated with the Celestine, 

 gypsum sometimes occurs, appearing as a pretty 

 pink, opaque crystallised substance. No systematic 

 method of working has been followed, and the 

 ground is left full of pits and mounds which render it 

 practically useless. 



The WTiter wishes to record his thanks to Messrs. 

 Sir Isaac Pitman, Ltd., for the loan of the photographs 

 showing the ballroom and a railway station in the 

 Wieliczka Salt Mines. 



Plant Breeding 



By I. B. N. 



Systematic botanists have marshalled all known plants 

 into divisions and subdivisions of which the " species " 

 is generally the ultimate. But the great total of 

 140,000 recorded species does not adequately represent 

 the diversitj'' of the plants. Only eleven species of 

 wheat are recognised, but they embrace about 3,000 

 different kinds or " forms." There are forms with 

 tall straw and forms with short ; red grain and white 

 grain forms ; early and late ripening forms, and so on 

 in each species through a vast range of differences. 

 New forms of cultivated plants may be introduced into 

 a locality in a variety of ways. They may be brought 

 in from elsewhere ; occasionally they arise in nature 

 by processes to be described later ; and they can be 

 made from old ones by hybridisation — that is to say, 

 by breeding from two distinct species. They are not 

 always better than the old forms, and careful tests of 

 their merits must be made before they can be recom- 

 mended for general cultivation. All methods of ob- 

 taining and testing forms of crop plants new to a 

 locality constitute the science of plant-breeding. 



Scientificall}^ the plant-breeder must try to be 

 " Jack of many trades and master of his own." Like 

 the geneticist, or student of heredity, he must study 

 the laws of transmission of characters from parents to 

 offspring. The geneticist maj^ study those plant 

 characters which are easiest to observe or of greatest 

 scientific interest, but the plant-breeder must deal with 

 those which are of agricultural importance. Of these, 

 in general, the most vital is yielding capacity, the 

 average amount of useful produce furnished per acre. 

 Now the eye is no judge of such a plant attribute as this 



and definite measurement is essential. We maj' sup- 

 pose that a new form of barley has been produced and 

 it is desired to test its yielding capacity in comparison 

 with an old form. If a one-tenth-acre plot of each 

 form of barley were grown, the plots being side by side, 

 and the new gave 15 per cent, more grain than the old, 

 there would appear to be convincing evidence of superi- 

 ority. But experiment has shown that, if instead of 

 one pair of plots many such pairs are grown, the 

 difference between the forms is not constantly 15 per 

 cent. It fluctuates, and may even sink to zero on some 

 pairs. This inconstancy is attributable to a number 

 of causes. The soil may not be equally fertile on 

 both the plots of a pair ; sparrows may damage one 

 plot of a pair more seriously than the other ; the 

 mcidence of a fungoid disease may similarly be heavier 

 on one plot than another. Actual test shows that a 

 difference in crop between two adjacent one-tenth- 

 acre plots cannot be regarded as significant unless it 

 exceeds 20 per cent. Present-day crops in England 

 have reached a level which makes a sudden yield 

 improvement of 20 per cent, almost inconceivable. An 

 improvement of 2 per cent, is much more probable, 

 and small as this seems, it would, if effected in all the 

 cereal crops of England, increase their annual value by 

 1^1,000,000. Methods of detecting such small yield 

 differences have, by degrees, been evoh'ed. The 

 principle is to grow about twenty small plots of every 

 form under test, arranging them so that every form has 

 a plot in every part of the total area of ground employed. 

 Then, if one end of the area be particularly fertile, all 

 forms share the amenity ; and correspondingly all 

 forms would suffer together in a part infested by 

 wireworm. The Chess-board Method, as this system 

 is called, permits of the certain detection of differences 

 of 2 per cent. 



Other agricultural characters vie with yielding 

 capacity in importance and difficulty. Resistance to 

 disease in potatoes, the bread-making quality of 

 wheat, and the suitability of barley for good beer are 

 examples. Often, not to create, but to appreciate the 

 merits of new forms, is the essence of the problem, 

 and it is impossible to understand either the difficulties 

 or the achievements without a knowledge of the com- 

 plexity of the agricultural characteristics of plants. 



Ornamental plants offer problems demanding an 

 eye for the beautiful and an understanding of popular 

 prejudices. The profusion of forms of tulips, primulas, 

 daffodils, carnations, violas, orchids, and many others, 

 which annually astonish us at the shows of the Royal 

 Horticultural Society, are mainly the favoured 

 survivors of great numbers of new forms raised by 

 hybridisation. It is far easier to gauge the opinion 

 of the gardening public upon a new flower than the 

 value to housewife and farmer of a new potato. 



