264 The Pedigree Families. 



For comparison I give in column B the mutants pro- 

 duced by the five other mothers per similar area under 

 cultivation or per 4 square meters. 



If we reckon the percentage of mutants among the 

 seedlings derived from this' particular mother we find it 

 to be about 40 % instead of the 6 % of the wdiole culture 

 (or the 5 % of the 5 other mother plants). 



But if we calculate this figure per cubic centimeter 

 of seeds we find it to be for this plant 1.8, but for the 

 rest 3.2. 



The absolute total of mutants is therefore reduced to 

 one-half through failure to germinate, wdiilst the absolute 

 number of seeds that germinated w^as reduced from 70 

 to 5 per cubic centimeter. The percentage of mutants 

 from the seeds which did germinate rose, therefore, from 

 5 to 40 %. 



The failure to germinate was in all probability due 

 to the fact that the seed had been kept for 5 years ; but it 

 is difficult to see why this should have affected one sample 

 of seed and not the five others. When seeds are kept 

 they gradually die off, some sooner, some later. 



It follows from the facts described above that mu- 

 tating seeds do not die off so soon as, or remain capable 

 of germinating longer than normal Lamarckiana seeds. 

 It is only the seeds of the dwarf form that appear w^eaker 

 than the parent species and perhaps also those of O. 

 clliptica. 



Should this phenomenon prove to be a general one 

 it should be possible by artificially hastening the death 

 of the seeds to materially increase the percentage of 

 mutants in a given sample. Such a discovery would im- 

 mensely facilitate the search for mutations in the vege- 

 table kingdom. 



