314 CULTIVATED PLANTS AND THEIR ORIGIN 



bring about changes in the structure and function of various 

 organs of plants. Generally it may be said that variations of 

 this kind, induced by changes in the amount of food-constituents 

 in the soil, or by alterations of season and climate, are rarely, if 

 ever, hereditary ; they appear under certain conditions, but when 

 the conditions are altered the variations disappear. 



For instance, by growing tall varieties of peas, beans, or any 

 other plants upon poor soil, successive generations of short 

 individuals may be obtained so long as the poverty of the soil 

 is maintained ; the seeds of such plants, however, when grown 

 upon good soil at once give rise to tall plants, showing that the 

 dwarfness of habit induced by such soil conditions is not a 

 permanent hereditary modification. 



Wheat, oats, and other cereals, when grown upon good garden 

 soils, at wide intervals apart, as has been done by some propagators, 

 develop tall straw, long ears, and large grain, but no new permanent 

 variety can be produced in this manner. 



By growing beets possessing ' fanged ' roots close together, they 

 have no room to develop their disfiguring branches, and may thus 

 be made to assume a good form ; nevertheless, seeds raised from 

 such plants, when grown under ordinary conditions of cultivation, 

 immediately give rise to plants with 'fanged' roots like their 

 ancestors. When attempting to develop a new race of any kind 

 of plant, it is therefore important that the modification taken as 

 a basis upon which the selective process is carried out, should 

 not have arisen merely as the result of external conditions. 



Where increased size of certain organs is the feature desired 

 in a new race, it is perhaps best to raise the successive genera- 

 tions of plants from which the selection is to be made upon a 

 moderately poor soil, rather than a specially rich one ; any in- 

 creased size of one plant over that of another under such circum- 

 stances would be less likely to be due to an accidental excess of 

 manure, and more likely to be due to an innate hereditary quality 

 of the plant. 



