i8o Changing Plant and Animal Nature 



ing to different genotypes (so far as concerns the one charac- 

 ter, stature) . 



If we select plants or animals because of some particular 

 quality that we value we may succeed in establishing a pure 

 line that reproduces the quality in question consistently; or 

 we may fail after repeated selection to get a stable popula- 

 tion. It all depends on whether in selecting we are dealing 

 with a phenotype or a genotype. 



Jennings in this country experimented with a one-celled 

 slipper-animalcule, the Paramecium. This lives in stagnant 

 waters and is easily cultivated in hay infusions, where it feeds 

 upon bacteria. It multiplies rapidly by simple cell-division, 

 each individual giving rise to two. Starting with a single 

 Paramecium in a dish, Jennings allowed the animal to multi- 

 ply until there was a colony of a large number of individuals. 

 In this colony there was a relatively wide range in size (see 

 Fig. 92) . Some of the smallest were taken out and allowed to 

 multiply in one series of dishes. Similarly, some of the 

 largest were separated and allowed to multiply in another 

 culture. After many generations Jennings found that the 

 individual variation among the offspring of the smallest was 

 exactly the same as the variation among the offspring of the 

 largest. Selection produced no effect whatever. On the 

 other hand, if several paramecia were picked out of a ** wild " 

 lot or mixed population, it was possible to get, by selecting 

 the smallest in one direction and the largest in another direc- 

 tion, Paramecium populations that were consistently small 

 in the one case and consistently large in the other case 

 (although of course there was the individual variation in 

 each group) . That is, each series had a distinct modal 

 dimension. 



Negative Results of Selection 



Other controlled experiments in which the effort has 

 been made to establish pure strains by selecting for a par- 

 ticular quality, such as pigment pattern or egg production, 



