OEIGIN OF DOMESTICATED CARROTS. 465 



inferior flowers, but would select from the best specimens ; and it is by 

 following up this system (even without more crossing than is performed 

 by Nature, and the bees), that great improvements have been made. 

 Thinking the same effects would accrue from a more careful selection of 

 culinary seeds, and that a much greater degree of piioductiveness might 

 be attained, about three years ago I began an experiment with long-pod 

 Beans ; I carefully selected the finest and fullest pods for seed, taking 

 none with fewer than five Beans in each. Next year I had a good 

 sprinkling of pods with six seeds in each ; these were saved for seed. 

 The following year there were many six-seeded pods and some with 

 seven. Following up the same plan, I find this season many more six 

 and seven-seeded pods, than of a less number, and some with eight 

 seeds ; there are still a few plants which produce five-seeded pods, and 

 it is wortTiy of remark, that the five-seeded plants have seldom a six- 

 seeded pod upon them, but all fives ; on the contrary, a six-seeded 

 plant generally has nearly all the pods bearing six Beans or more." 



By a similar process M. Vilmorin obtained domesticated Carrots from, 

 wild ones in a few generations {Hort. Trans., 2nd ser., vol. ii., p. 348), 

 and the curious experiments of M. Esprit Fabre upon fixing the cha-, 

 raeter of Wheat in plants derived from an -SJgilops, were conducted upon 

 the same principles {Journ. Agr. Soc, vol. xv., p. 167). In fact it is 

 thus, and thus only, that in aimual plants any improvement in g[uality 

 can be rendered permanent. 



There is a class of facts apparently opposed to these views. It is said 

 that the fruit of Apple- and Pear-trees, raised from the seeds of varieties 

 of the highest exeeUenoe, wiU. often be little better than that of 

 wildings. This obscure subject wiU. be considered in the next chapter, 

 in connection with hybridity. 



But to fix a new habit in annual plants is not tlie only care 

 of the cultivator, whose patience and skill would be iU employed 

 if it could not be preserved. If a plant has some tendency to 

 vary from its original condition, it has much more to revert to 

 its wild state ; and there can be no doubt that, if the arts of 

 cultivation were abandoned for only a very few years, all the 

 annual varieties of our gardens would disappear, and be replaced 

 by a few original wild forms. 



For. the means of preserving the races of plants pure, the 

 means vary according to the nature of the variety. As far as 

 concerns early and late varieties, it often happens that, as in 

 Peas, the tendency in such plants to advance or retard their 

 season of ripening was originally connected with the soil or 

 climate in which they grew. A plant which for years is 



