THE VITAL FABRIC OF DESCENT 313 



suggestion of a more primitive or ancestral type. A variety 

 narrowly selected in one country to secure the accentuation of 

 its peculiar characters, may deteriorate, or fail to reach the 

 same degree of specialization when tlie cultural conditions of 

 growth are changed. Through degeneration, or loss of com- 

 plexity, a species may appear more primitive or less evolved 

 than it really is. To reversion is also ascribed the occasional 

 cropping-out in the individual of some ancestral peculiarity 

 (atavism), but these minor fluctuations of form minister to the 

 healthful diversity of the species, and are far from proving that 

 evolution has turned backward. The transformation of pistils 

 and stamens into petals, as in the formation of double flowers 

 and similar mutative changes, is not, as sometimes supposed, a 

 reversal of evolutionary processes, but is in the direction of 

 developmental history — an over-shooting of the mark, as it 

 were. Reversion would change petals back to stamens ; this 

 seldom happens, and when it does we recognize it as a recovery 

 of normal form and function. It is now coming to be appreci- 

 ated that the evolutionary history of the higher plants has 

 involved a progressive sterilization and vegetative specializa- 

 tion of parts which were once devoted to reproductive purposes. 

 Even the cells of which the bodies of the higher organisms are 

 composed are sexual in their origin and represent a condition of 

 prolonged conjugation. 



The final inconsistency in terms is reached by those who have 

 suggested reversion as the cause of the same phenomena which 

 it is held to obliterate, that is, the mutations themselves. This 

 is to use the one word reversion in two directly opposite senses. 



Mutations often suggest other species of the genus, as in 

 Coffea, and have been termed "reversions" to an ancestral 

 character;^ but just such "reversions" are said, also, to " re- 



' The " ISIaragogipe " mutation of Coffea arabica, for example, has a super- 

 ficial resemblance to Coffea liberica, and has been held by some to be a cross be- 

 tween the two. Other mutations of coffee originating in Central America share 

 features of several of the wild African species. 



Mr. Luther Burbank has found that hybrids also are sometimes more obviously 

 similar to other members of the genus than to their own parents. Thus the 

 Wickson plum, a hybrid between Japanese varieties of Prutius trijiora, was be- 

 lieved by Professor L. H. Bailey to be descended from P. siinoni, a Chinese 



