No. 518] GENETIC AL STUDIES ON OENOTHERA 



1 15 



cross, were so blended that as regards the measurements 

 of parts, habit, texture of foliage, etc., the average for 

 each set of hybrids would probably present a fair mean 

 between the two parents concerned. There was. how- 

 ever, a wide variation in the resemblance of the hybrids 

 to one or the other of the parents. 



2. No character of either parent was discovered which 

 appeared as dominant in these hybrids of the F, genera- 

 tion, after the manner which has been described for cer- 

 tain forms (e. g., Pisum) that illustrate most conspicu- 

 ously Mendelian dominance in the first generation. 



3. Some of the hybrids of each cross presented a greater 

 resemblance to one parent and some to the other, and the 

 forms could therefore be arranged in two groups (twin 

 hybrids) in one of which the maternal characters were 

 most evident and in the other the paternal. There was 

 no clear evidence that the hybrids of these cultures car- 

 ried in marked preponderance the paternal characters 

 (patroclinous), or on the other hand that maternal char- 

 acters were more prominent. The range of variation 

 among the hybrids was too great to permit of such con- 

 clusions. 



It becomes a matter of interest to determine how 

 plastic these hybrids will prove to be in later generations, 

 and whether or not they will exhibit variations that may 

 be fixed and accentuated by artificial selection. This line 

 of enquiry will be especially pertinent to the behavior 

 of those hybrid plants of biennis and grcmdiflora which 

 in this Fj generation presented LamarckianaAike flowers 

 and inflorescences, although differing markedly from 

 LamarcJciana in habit and foliage. 



I am greatly indebted to the Harvard botanic garden 

 for the facilities placed at my disposal for this work. 



December. 1909. 



