180 MISSOUEI BOTANICAL GAEDEN. 



Regarding the explanation of these cases, I soon found 

 that every polymerous flower had two bracts at its base instead 

 of being in the axil of a single bract. These bracts may be 

 entirely independent or they may be more or less coalesced 

 at the base, but they project in opposite directions. (See 

 photograph by DeVries, 1909, p. 472.) In all these cases 

 the ovary and hypanthium are more or less flattened. An 

 examination of the stems which bore these polymerous flow- 

 ers, disclosed the fact that they exhibited irregularity in the 

 placing of the flowers on the stem, or in other words, varia- 

 tions in phyllotaxy. The flowers and their bracts were not 

 uniformly distributed on the stem, but certain flowers were 

 very close together and others long distances apart. It seems 

 clear that this is the explanation of the phenomenon, which 

 is therefore one of synanthy rather than of polymery. The 

 Anlagen of the flowers are of course laid down and their 

 position determined when the terminal rosette of the stem 

 is very small. Anlagen of successive flowers therefore arise 

 very close together, and if anything leads to variation in 

 their position they will sometimes occur partly in contact or 

 overlapping, giving a flower in which the parts are more or 

 less completely doubled in number. The flower having two 

 independent styles, and the fact that two bracts are always 

 found at the base of polymerous flowers, shows that it must 

 be due to a partial coalescence of primordia, such as I have 

 mentioned. It is interesting to note that flower No. 5, hav- 

 ing its parts in threes, was immediately below No. 4, which is 

 heptamerous, and on the same side of the stem. Similarly, 

 the plant bearing flower No. 9 also bore at the same time 

 (Aug. 20) the two trimerous flowers, Nos. 10 and 11. It 

 may also be noticed that in all the polymerous flowers the 

 number of stigma lobes is less than the number of stamens. 

 The same is true of DeVries' records. These polymerous 

 flowers are much larger than the normal owing to the larger 

 number of parts, the parts themselves retaining their usual 

 size, except that the hypanthium and style are stouter, as 

 might be expected, and the filaments are sometimes thicker. 



In the season of 1910 a number of additional observations 



