"PURE LINES" 379 



about by a " mutation " in the pure line, not by selection among 

 the individual differences. 



Jennings has illustrated the significance of " pure lines " in 

 reference to the slipper-animalcule (Paramoccium). The figure 

 shows a population made up of eight pure lines, each of which 

 is marked by a certain range of size. The line x — x indicates 

 the mean of the populate ; the crosses indicate the means of 

 the several lines. If a giant be isolated from the first line, 

 its progeny, kept in the same conditions, keep up the characters 

 of that line. The large-sized stock thus arising can hardly be 

 called the result of selection from the population in question ; it 

 is the result of the isolation of an individual of a particular 

 pure line. And the other point is, that no amount of selection 

 will get anything out of the isolation beyond the limits of the 

 pure line from which it came. 



The conclusion that experiments on " pure lines " suggest 

 is one towards which many lines of modern experimentation 

 point, namely that in certain sets of cases the variations 

 that count are mutations, not fluctuations. By a germinal 

 stride a most excellent ear of wheat is formed ; there is 

 more to be got out of that single ear than out of years of 

 selection of smaller fluctuations. The fact that, although plus 

 and minus fluctuations occur in the pure line, selection can 

 make nothing more of them, seems to show that these small 

 fluctuations are often not transmitted. It is probable indeed 

 that many of them are not germinal variations at all, but ac- 

 quired modifications due to diversities of nurture. There is a 

 question, however, which must not be left out of account, namely, 

 whether the individual new departures, which are often far 

 from abrupt or startlingly discontinuous, may not be the outcrop 

 of the long-continued selection of forms showing small fluctua- 

 tions. This is, indeed, suggested by the conclusion of some 

 investigators, though not of De Vries, that a mutation is a 

 stride in the same direction as that of the majority of the flue- 



