HEREDITY IN PURE LINES 377 



Johanssen pointed out that each pure line had a different mean 

 size of bean and a different distribution about the mean. This was a 

 real hereditary difference due to differences hi the germinal content 

 of the original parent-beans. Two beans of exactly the same size, one 

 an average individual of a larger stock and one a large individual of a 

 smaller stock, were planted and then* offspring varied about two dis- 

 tinct means. This leads to the idea that an individual produces off- 

 spring not hi accord with its somatic appearance, but according to its 

 germinal content. So this idea of the difference between what an 

 individual is somatically, and what it is germinally led Johanssen to 

 introduce the terms "phenotypic" and "genotypic" "phenotype" 

 and "genotype." 



Thus, if one selected all the beans of a given size he would have a 

 group of phenotypes that would be identical phenotypically, but would 

 be very different genotypically; for each might be germinally different 

 and would therefore have different groups of offspring. All of the 

 individuals hi one pure line, however, whether they differ somatically 

 (phenotypically) or not, would belong to the same genotype and would 

 be genotypically equivalent. 



The appreciation of this distinction at this place, before the treat- 

 ment of Mendelian heredity, will be of great advantage. Few more 

 useful terms have been devised in connection with genetics than geno- 

 type and phenotype. 



W. L. Tower hi a long series of experiments on the potato beetle 

 (Leptinotarsa decemlineata] came to similar conclusions as the result 

 of his attempts to modify a character by selection. Instead of using 

 a self-fertilizing type, he chose a long inbred stock that was probably 

 all identical germinally, but varied considerably hi shade of color, etc. 

 He selected for twelve generations the darkest specimens and, instead 

 of getting all dark offspring, he got an array of all shades fluctuating 

 evenly about the average. At the end of twelve generations of 

 selection there was no change in the proportion of light and dark 

 individuals. 



Jennings tried another type of pure-line work, using one-celled 

 organisms which reproduce by binary fission, i.e., by the division of the 

 parent into two nearly equal halves, thus forming two offspring. He 

 chose a considerable number of Paramecia and isolated each in a 

 separate small aquarium where it was allowed to breed for some gen- 

 erations. The original individuals differed quite markedly in size and 

 hi other structural characters. The various sets of progeny were 



