726 INHERITANCE 



parental shell, and takes on the form of the protoplasmic body. A mass 

 of silicious particles, which have earlier been collected within the parent, 

 passes to the surface of the new protoplasmic body and is there molded 

 into the final form of the shell of the offspring. Meantime the nucleus 

 has divided and one of the products passes into the young individual. 

 Parent and offspring now separate. 



During this process are formed the spines in their final form, length, 

 and number. The parent shell may have four spines, the shell of the 

 new individual five spines or some other number; there may be differences 

 in the length of the spines. The differences between the characteristics of 

 parent and offspring might conceivably be due to slight and obscure 

 differences in the environmental conditions at the time of reproduction. 

 Or they might be due to diversities in the internal condition or consti- 

 tution of the individuals at the time of reproduction, however such 

 diversities had been produced. 



Results of long- continued selection. — From the results thus far set 

 forth, it appears that the individual diversities within a clone of Dij- 

 jiugia are not usually or strongly inherited. Is there any tendency what- 

 ever for such inheritance.'^ In other words, does the difference of condi- 

 tion or constitution that results in diversity of characters ever or in any 

 degree persist through vegetative reproduction? 



This may be tested by long-continued selective breeding within a 

 clone. Suppose that the basis of selection is the number of spines. From 

 a single clone two groups are segregated, one containing only individ- 

 uals with high numbers of spines, the other only individuals with low 

 numbers of spines. In the former group are retained only descendants 

 with high numbers of spines; from the second group only descendants 

 with low numbers of spines. In this way there are obtained in later gen- 

 erations from the first group individuals descended for many generations 

 from ancestors all of which had high numbers of spines, and from the 

 second group individuals descended for many generations from ancestors 

 with low numbers of spines. Selective breeding of this kind may be car- 

 ried on with any of the varying characteristics. 



When such selection is practiced for many generations, usually for 

 considerable periods no difference in the offspring of the two groups is 

 to be discovered. But as the number of generations becomes greater, an 



