MUTATION. 163 



ficulty in the way of this hypothesis has been the fact, that 

 such novelties were often produced only after the parent- 

 species had been grown for a number of generations without 

 producing any aberrant forms. 



Now however, we know that plants of Oenothera Lamarckiana 

 may be complex hybrids, that they may be impure for quite a 

 number of genes, that their pollen may be geno-typically differ- 

 ent from their ovules, but that they may nevertheless breed 

 true, as the result of a peculiar mechanism, which is as yet 

 obscure. This mechanism, this mysterious set of conditions, 

 produces the remarkable result that individuals may produce 

 pollen like the pollen-cell furnished to their make-up by their 

 father, and ovules like that from which they developed. What 

 happens in these plants must be some sort of a suppression 

 of the normal synthesis of the germ-cells. We might conceive of 

 some process analogous to that which takes place in the pro- 

 duction of periclinal chimeras, in which two geno-typically 

 different individuals live in intimate contact, one inside the 

 other, and yet retain their geno- typical identity. We can imag- 

 ine how a plant of Oenothera grows up as a composition of tis- 

 sues of different identity, partly composed of cells directly pro- 

 duced from the original male gamete, and partly of cells de- 

 rived directly from the original ovule. If then we have sufficient 

 imagination to think of a way in which this male individual, 

 derived from the pollen can furnish pollen, whereas the other 

 individual, derived from the ovule, in its turn will furnish the 

 ovules of what looks like one plant, we could make some work- 

 ing-hypothesis, which may prove of some use in the investiga- 

 tion of the peculiar mechanism. We would not be surprised if 

 some of these at first sight sexually produced hybrid Oenothe- 

 ras proved to be a sort of periclinal chimera, formed in a quasi- 

 sexual but really asexual way, composite of the two parent- 

 forms, such as Muricata plants in Biennis epiderm, just as 

 Cyticus adami is Cyticus laburnum with an epiderm of Cyticus 

 purpureum. 



One thing is certain, namely that the process by which Oeno- 



