DURATION OF RACES. 343 



tive. This cannot readily be explained away, while 

 the failures may be, by exhaustion of soil, incoming 

 of disease, or alteration of climate or circumstances. 

 On the other hand, it may be urged that, if a variety 

 of this sort is fated to become decrepit and die out, it 

 is not bound to die out all at once, and everywhere at 

 the same time. It would be expected first to give 

 way wherever it is weakest, from whatever cause. 

 This consideration has an important bearing upon the 

 final question, Are old varieties of this kind on the 

 way to die out on account of their age or any inherent 

 limit of vitality ? 



Here, again, Mr. Knight took an extreme view. 

 In his essay in the " Philosophical Transactions," pub- 

 lished in the year 1810, he propounded the theory, 

 not merely of a natural limit to varieties from grafts 

 and cuttings, but even that they would not survive 

 the natural term of the life of the seedling trees from 

 which they were originally taken. Whatever may 

 have been his view of the natural term of the life of 

 a tree, and of a cutting being merely a part of the 

 individual that produced it, there is no doubt that he 

 laid himself open to the effective replies which were 

 made from all sides at the time, and have lost none of 

 their force since. Weeping-willows, bread-fruits, ba- 

 nanas, sugar-cane, tiger-lilies, Jerusalem artichokes, 

 and the like, have been propagated for a long while 

 in this way, without evident decadence. 



Moreover, the analogy upon which his hypothesis 

 is founded will not hold. Whether or not one adopts 

 the present writer's conception, that individuality is 

 not actually reached or maintained in the vegetable 



