OABRIKLLE L. C. HOWARP. 37 



carried out under personal supervision. The boys who do the work 

 have to wash their hands after finishing each box, as otherwise 

 a few ungerminated seeds might be carried to other cultures. 

 After transplanting, the soil in the boxes is throw^n away and 

 the boxes washed. These precautions have proved successful 

 and in no case has any mixture of the cultures been discovered. 



The elimination or rather the reduction of the differences 

 due to environmental influence is much more difficult. There is 

 perhaps ho plant which is so sensitive to changes in soil, climate 

 and external conditions generally as the tobacco. The shortness 

 of the growth period, the large amount of material in the form of 

 leaves and stem formed in a short time are probably the reasons 

 why any check or stimulus has so great an effect. Moreover, 

 the tobacco plant appears to have an infinite capacity to adapt 

 itself to conditions, the same pure type giving rise to plants 

 1^ feet or 8 to 10 feet high according to the cultivation. Even 

 in very adverse circumstances the plant goes through its com- 

 plete cycle and forms seed. Apart from such extremes, a very 

 small difference in cultivation is sufficient to induce a very 

 marked change and to raise a field of plants uniform enough 

 for accurate measurement is not easy. Nevertheless, in all 

 plant-breeding experiments, the absolute necessity of normal 

 and well-growTi plants cannot be emphasized too strongly. 

 The differences induced in tobacco plants apart from 

 size are almost incredible. The general effect of un- 

 favourable environment on N. tabacum is to wipe out all 

 differences and to make the plants appear uniform. The 

 differences between the various types in leaf shape and leaf 

 surface, which are small, almost disappear in under-developed 

 plants. Unless the plants are well growTi, it would be very 

 easy to be misled in observations on such characters as the 

 undulations of the leaf surface. The numbers would probably 

 show far too great a proportion of flat leaves. 



Owing to the large amount of experimental work carried on 

 in the Botanical Area at Pusa, special care has been taken to 



