i86 CONSTITUTIONAL TENDENCY TO VARIATION. 



should be made the sulbject of repeated experiment ; year after 

 year seeds should be saved, seed-beds " rogued," and attempts 

 made to secure fixity of character. If they end in nothing, as 

 they often will, such experiments have the advantage of also 

 costing nothing ; but if they lead to a good result a permanent 

 ■ gain is secured. 



The tendency of plants to variation being so general, and in many 

 cases so remarkably great, we may reasonably expect that by taking 

 proper advantage of it we may obtain much, if not all that we would 

 wish. " We see every day the wide range of seminal diversities in our 

 gardens," said Dean Herbert, in the Journal of the Horticultural 

 Society. " We have known Dahlias from a poor single dull-coloured 

 flower break into superior forms and brilliant colours ; we have seen a 

 Carnation, by the reduplication of its calyx, acquire almost the appear- 

 ance of an ear of Wheat, and look like a glumaoeous plant ; we have 

 seen Hollyhocks in their generations branch into a variety of colours, 

 which are reproduced by the several descendants with tolerable certainty. 

 We cannot, therefore, say that the order to multiply after their kind 

 meant that the produce should be precisely similar to the original type ; 

 and, if the type was allowed to reproduce itself with variation, who 

 can pretend to say how much variation the Almighty allowed ? Who 

 can say that this glorious scheme for clothing the earth was not the 

 creation of a certain number of original plants, predestined by Him in 

 their reproduction to exhibit certain variations, which should hereafter 

 become fixed characters, as well as those variations which even now 

 frequently arise, and become nearly fixed characters, but not absolutely 

 so, and those which are more variable, and very subject to relapse in 

 reproduction?" 



But we are by no means destitute of the power of procuring, 

 with some certainty, improved varieties, by an application to 

 practice of physiological principles. In the last chapter has 

 been shown the importance of securing the production of seed 

 by plants in the niost healthy state possible, because a robust 

 parent is likely to afford a progeny of similar habits to itself. 

 In annuals, however, this is apparently restrained withia 

 narrower limits than in woody plants, from the great difficulty 

 of fixing a new peculiarity in the former, and the facility with 

 which it may be effected in the latter case, by means of buds, 

 cuttings, grafts, and similar modes of propagation. The object 

 of the scientific gardener who desires to improve the varieties 



