326 PLANT-BREEDING 



Exactly similar conclusions may be derived from a dis- 

 cussion of the (Enothera lata, which has often been seen to 

 be produced from the parent species by a single leap. There 

 is even less connection between the various marks by which 

 it is distinguished from Lamarck's primrose. It strikes us 

 through all the periods of its life as quite another type. The 

 very first leaves of the young seedlings differ, being broader 

 and more rounded at the tip. This type of leaves is pre- 

 served during the whole life history, and the rosettes, the 

 young stems, and the branches are distinguished by this 

 same mark. There can be no doubt that the form of the 

 leaves during the whole lifetime is regulated by one single 

 unit-character. This unit probably causes still another 

 mark, the extremely sinuous surface of the leaves. Sinuos- 

 ities, although not lacking in the parent species, are much 

 more numerous in this mutant form. The weakness of the 

 stems and the consequent bending of their tips is more diffi- 

 cult to explain as an effect of the same cause, but it is as 

 constant a mark as the leaves. More curious is the behavior 

 of the flowers. These have only one sex, producing no 

 pollen at all. The anthers are developed and of the normal 

 size, but in their cavities the pollen is sometimes entirely 

 wanting, and sometimes sterile, their place being occupied 

 by the outgrowth of the cells of the inner layer of the wall. 

 These cells commonly collapse and are absorbed, and so it 

 is in Lamarck's primrose and in all its other derivatives, but 

 in the lata they thrive and increase their size until the time 

 of the shriveling of the anthers. 



The correlation between the broad and sinuous leaves 

 and this inability to produce pollen is a phenomenon which 

 it is at present far beyond our power to explain. But it is 

 absolutely constant. The lata has been produced anew by 

 the main strain in my garden more than three hundred times. 

 A large number of these plants have flowered, and the flowers 



