128 THE DEVELOPMENT OF RAW COTTON 



might consist of a dozen different strains, and yet all the 

 plants might produce exactly the same lint. Perfectly 

 Independence definite differences in petal colour and such- 

 of Inherited like can be quite independent of lint prop- 

 Characters. er ties. Similarly, so far as our knowledge 

 goes, white and brown lints may be otherwise identical, 

 or two entirely different lints be borne on plant bodies 

 which are externally indistinguishable. There is no limit 

 yet known to the shuffling of characters which may take 

 place in this way through deliberate or natural inter- 

 crossing within the main groups of the genus. 



The obvious characters make only half the story, how- 

 ever. We are beginning to realize, and in some ways to 

 understand, how two kinds of cotton may 



Growth S ] 10W entirely different reactions to their 

 Habits. 



environment, so that such a complication 



as the following example presents is well within proba- 

 bility : Two strains of cotton are growing a hundred miles 

 apart, one on the sea -coast, the other in the interior, and 

 appear to be exactly the same. When grown side by side 

 in either site, the imported one is conspicuously unhappy, 

 and plots of it can be recognized at the other side 

 of the field. These differentiating characters which are 

 not obvious include such reactions as tolerance of salt, 

 liability to shedding, stage at which senescence sets in, 

 and velocity of growth at given temperatures. In 

 separate species of the genus these features may be most 

 clearly seen, but in a less obvious degree they may be 

 found in the strains isolated from the same variety. One 

 of the most interesting sights of cotton-growing in the 



