2 26 Flowers and their Pedigrees. 



probably they have developed their beautiful scarlet 

 colour. 



So far, all the members of the rose family with 

 which we have dealt have exhibited a progressive 

 advance upon the common simple type, whose em- 

 bodiment we found in the little wayside potentillas. 

 Their flowers, their fruits, their stems, their branches, 

 have all shown a regular and steady improvement, a 

 constant increase in adaptation to the visits of insects 

 or birds, and to the necessities for defence and pro- 

 tection. I should be giving a false conception of 

 evolution in the roses, however, if I did not briefly 

 illustrate the opposite fact of retrogressive develop- 

 ment or degeneration which is found in some mem- 

 bers of the class ; and though these members are 

 therefore almost necessarily less familiar to us, 

 because their flowers and fruits are inconspicuous, 

 while their stems are for the most part mere trailing 

 creepers, I Aust find room to say a few words about 

 two or three of the most noteworthy cases, in order 

 to complete our hasty review of the commonest rosa- 

 ceous tribes. For, as we all know, development is 

 not always all upward. Among plants and animals 

 there are usually some which fall behind in the race, 

 and which manage nevertheless to eke out a liveli- 

 hood for themselves in some less honourable and 



