38 



ROBERT W. HEGNER 



and 29. This result is what the studies of various investigators 

 of pure Hues and my own studies of A. dentata would lead one 

 to expect. 



Table 11 gives the correlation in diameter of parents and 

 offspring within the twenty-five small families. 



e. Relations between nuclear number and diameter within a large 



family 



Family ap. 5. The members of family ap. 5 were more viable 

 and multiplied more rapidly than those in the other families, 

 hence this line was chosen for further investigation. Nuclei 

 were counted only occasionally during the first two periods 

 described below ; they varied in number from 5 to 7. 



1. Correlation in diameter between pareMs and immediate 

 offspring. Period 1. This period extended from March 18 to 

 May 10; 1918. The 173 parents and offspring are recorded in 

 correlation table 12, and the first part of the pedigree is pre- 

 sented in figure 32. 



2. Results of selection for lines with large and small diameters. 

 Periods. On May 11th selection for diameter was begun on 

 the basis of past performance, and the following specimens were 

 chosen as progenitors of the high and low lines, respectively. 



All parents and offspring within the high line with a diameter 

 of from 28 to 32 units were kept unless they persisted in pro- 

 ducing young with a diameter less than 28; and likewise in the 

 low line, all parents and offspring with a diameter of 24 units 

 or less were retained unless they habitually gave rise to progeny 



