The Hypothesis of a Premutation Period. 493 



rate account for the comparative rarity of their appear- 

 ance. 



We should therefore have to determine, experimen- 

 tally, the effect of the combination of such extremes. I 

 do not regard this as in any way impossible. For ex- 

 ample, take very weak buds and their shoots, or very 

 weak flowers and supply them with as much nutriment 

 as possible. Just as saplings or the strong shoots pro- 

 duced by resting buds, often develop hitherto latent char- 

 acters (such, e. g., as the well-known intemiediate forms 

 between leaves and thorns in the common barberry) ; so 

 perhaps they might also be induced to give rise to muta- 

 tions. A very rapid multiplication is generally regarded 

 as an effective inducement to the production of muta- 

 tions ; the reason for this being probably that the seeds 

 which would otherwise perish at or immediately after 

 germination find the necessary conditions for full growth. 

 The seeds in question are those which have suffered from 

 unfavorable circumstances ; and this gives us the contrast 

 we spoke of above. For the sake of experiment, there- 

 fore, we should collect the seed from the small, late, 

 lateral twigs of the higher orders, and sow it with all 

 possible care.^ 



Let us now suppose that a genuine premutation period 

 has been induced or at any rate discovered. \\'hat must 

 be expected? In other words let us suppose that the 

 potentialities for a whole series of mutations have arisen 

 in a plant or a group of plants. Will each potentiality 

 actually give rise to a mutation, and, hence, to a new 

 species (fit or unfit) ? Chance will obviously have a great 

 deal to do with it. The latent characters are evidently 



^ Experiments of this kind would obviously have to he continued 

 through some years; the greatest difficulty is the choice of suitable 

 plants. 



