No. 443] MUTATION IN PLANTS. 747 



But little definite evidence is at hand as to the time at which 

 the changes antecedent to mutation, constituting pre-mutation 

 occur, although certain stages of development may be designated, 

 previously to which they must come about. Mutations of the 

 higher plants are first apparent in the seedling but the actual 

 alterations or departure from the hereditary behavior must have 

 taken place at least as far back as the formation of the se.xual 

 elements the union of which produced the embryo, and may 

 have occurred even earlier. In any case the mutants are per- 

 fectly formed in the embr)'o and influence of any kind upon the 

 germinating seed may not alter their nature (see page 746). It 

 may be seen from the foregoing that the mutative processes may 

 be connected with either the vegetative body or the sexual ele- 

 ments, and may be found within the sporophyte, or be confined 

 to the gametophyte. 



If the pre-mutative alterations occur in the vegetative proto- 

 plasts of a self-fertilized individual both gametes would presuma- 

 bly carry the same characters to the union. If, on the other 

 hand, premutation occur in one of the sexual elements, or if it 

 occur in the vegetative cells of species which are cross fertilized 

 only, the embryos formed would be the result of the union of one 

 mutant gamete and one of the regular inherited form. In a sense 

 such mutants might be considered as hybrids. This theoretical 

 aspect of the question seems to find a reflection in the behavior 

 of Gi. lata, one of the mutants with pistillate flowers only. When 

 pollinated by the parent form, Gi. lamarckiana, it produces CE. 

 lata and CE. lamarckiana . 



DeVries conjectures that the causes inducing mutation are 

 partly internal, and partly external to the organism. The state 

 of external factors necessary to the process probably occur only 

 at uncertain intervals, and is supposed to embrace a combination 

 of extremely favorable and unfavorable conditions. 



Probably no more profitable subject for research in the whole 

 realm of natural history offers itself to the investigator than the 

 problem of the causes which produce new species. The above 

 supposition deserves early attention from the experimentalist 

 since it is one that is comparatively easily capable of proof and 

 disproof. 



