ABNORMALITIES IN OENOTHERA.. 



181 



made 



The records of these were kept 



by my assistant, Mr. V. Follenius, during my 

 I had the oDnortunitv of examining the 



before the end 



g the most interesting 

 Cases of polyphylly or 



synanthy were found in a much wider range of Oenothera 

 forms than in the previous year. The following is the list : 



Formula 



1. 



2. 

 3. 

 4. 



K7C7S14N11 



KeCeSioNs 



KeCeS^Ns 



K4C4S8N4 



K4C4S8N5 



5. . K3C3S6N6 



6..K 3 C 3 S 6 N4 

 7..K3C3S8N5 



8..K 6 C6Si2N 8 



9..K 5 C 5 S?N? 

 IO-.KsCsSsNs 



H..K5C5S8N8 



12. . K5C5S8N8 



I3..K5C5S8N7 



14. . KsCsSsNs 



I5..K5C5S8N5 



I6..K5C5S8N5 



17. . K5C5S8N5 



I8..K5C5S8N5 

 19,.K5 1 C4S 9 N4 



Remarks 



O. multiflora 

 O. multiflora 

 O. multiflora 

 Race No. 25, from 



near Liverpool, 



England 



0. biennis X Lamar c- 



kiana 



a 



« i 



< • 



"(same 

 plant as No. 6) 

 O. grandiflora from 



Alabama 

 Race 54 X 40 

 O. biennis, Chelsea 

 Physic Garden 



Two bracts at base of flower. 



Two bracts partly coalesced. 



Two bracts at base. 



Two perfect tetramerous flowers, 

 with their hypanthia in contact 

 throughout their length and 

 partly fused. Ovaries in contact 

 and partly fused. Two bracts. 



One bract. 



One bract. 

 One bract. 



Hypanthium and ovary flattened. 

 One bract. 



One bract. 



Only one bract at base of each 

 flower in this race. In one 

 case the bract had two tips, as 

 though resulting from the in- 

 complete coalescence of two 

 bracts. 



Two sepals of normal width, 

 3 narrower, occupying about the 

 width of the other two. 



The case of No. 4, in which two tetramerous flowers were 

 found, each with its bract, and only partly fused by their 

 ovaries and hypanthia, is particularly instructive and con- 

 firmatory of my explanation. One hexamerous (No. 8) and 

 one pentamerous (No. 9) flower, however, had each but one 

 bract at its base, as well as the three trimerous flowers in the 

 race 0. biennis X 0. Lamarckiana. It is therefore prob- 



