SELECTED ADAPTATIONS 451 



ing clematises on the other, according as the individuals of 

 successive generations were born with the ability to make 

 the most of the little light, or the ability to climb into the 

 sunshine above. Now, since we are deriving these climbing 

 plants from rosette-bearing ancestors it is easy to see how 

 rosettes might pass into whorls of a few leaves and the few 

 be reduced to two. • We should then have the opposite leaves, 

 characteristic of clematis, arising as an incidental result of 

 the development of the climbing habit. So likewise the 

 peculiarity of having four sepals, which characterizes the 

 genus, may be an incidental result of the opposite leaf- 

 arrangement, for we have only to suppose that such a funda- 

 mental change of habit in the growth of the foliage-leaves 

 would show also in the arrangement of the floral leaves since 

 these would be in a primitive condition readily permitting 

 the change. Other changes involved in passing from the 

 marsh-marigold-type to the clematis-type of flower and fruit, 

 which we have already seen to be advantageous under the 

 drier, sunnier, and windier environment, may be easily imag- 

 ined to have been effected by the accumulation through 

 heredity of innumerable spontaneous variations, each very 

 slight but of vital importance to its possessor in the struggle 

 for existence. 



According to the view above outlined, more and more 

 highly developed groups would be separated from relatively 

 primitive ones through the perpetuation of slight yet favor- 

 able fluctuating variations, fitting their possessors to occupy 

 the more and more trying environment. At the same time 

 the more primitive forms would survive under the less exact- 

 ing conditions and perfect their simpler structure. Inci- 

 dentally there might be preserved through successive genera- 

 tions acquired peculiarities of an adaptive sort or even 

 nonadaptive ones, provided they were not seriously injurious; 

 or, under the same proviso, there might be perpetuated such 

 a characteristic feature as the plan of four in the calyx of 

 clematis, which had arisen as a merely incidental resuH of 

 some other modification. A pecuharity depending in this 

 way upon another peculiarity is called a correlated character. 

 Occasionally features of no conceivable use to their possessors, 



