354 THE SURVIVAL OF THE UNLIKE. [XXI. 



are impotent while others are not is proof that fruits 

 vary or differ in this respect when grown from seeds. 

 Perhaps there are as few impotent fruit trees now as 

 there ever were, and that our attention is now called 

 to them simply because they have been propagated or 

 multiplied extensively and because we are now in- 

 quiring carefully into all horticultural problems; but I 

 am inclined to think, from reasons already advanced, 

 that there must be a general (though very slow) tend- 

 ency towards self -sterility in highly cultivated plants. 

 The natural check to this self- sterility is the raising 

 of plants from seeds, by which means a considera- 

 ble amount of variation is secured in sexual characters. 

 In proof of this, I will cite the case of garden vegeta- 

 bles, in which the various individuals of a variety are 

 fertile with each other, even when a given individual 

 is sterile with itself. Thus blocks of the same variety 

 of tomato or bean fertilize freely. But while this 

 same inter -varietal fertility would undoubtedly result 

 from growing only unbudded or ungrafted fruit trees, 

 the disadvantage, as every one knows, would be so 

 great as to make the practice unprofitable. But the 

 same result can be obtained by planting different 

 named varieties together, for these varieties represent 

 different seed -parents. And this is the conclusion 

 which the best practice enforces, for mixed orchards 

 are, as a rule, the most successful ones. 



A broad epitome of the whole problem seems to 

 run something like this : There is a general tendency 

 in nature toward a separation of the sexes, or uni- 

 sexuality, and this tendency is probably hastened 

 among plants by high cultivation. The first signs 

 of separation — and beyond which most plants may 



