182 MISSOUSI BOTANICAL GARDEN. 



able that the latter are real cases of polyphylly and not of 

 synanthy, in which,, instead of the partial coalescence of two 

 primordia or their failure to separate, there is a variation 

 in the division of the Anlagen of the various cycles of organs 

 in the flower, resulting in a flower having a larger or smaller 

 number of parts than normal. When, as in flowers 10-18, 

 the androecium is normal while the calyx and corolla show 

 an increase of paffs, this may be considered to be due to 

 polyphylly ratheir than synanthy. 



The ten pentanrerotis flowers in 0. biennis from the 

 Chelsea Physic Garden were found in a culture of 

 33 plants, and caifeful search would doTlbtless have 

 revealed further cases. Evidently the phenomenon is rela- 

 tively common in this race.^ The characters of the race 

 ate very constant and are remarkably different from those of 

 any other race of O. biennis I have seen. They will be de- 

 scribed elsewhere. These pentamerous flowers had invaria- 

 bly but one bract at their base, which would indicate that 

 the phenomenon of pentamery is here due to polyphylly 

 father than to synanthy, although the fact that one of these 

 basal bracts had a double tip might be considered to favor 

 the interpretation of this also as due to synanthy. 



It would seem' therefore that while most of these are cases 

 of synanthy, or coalescence of two primordia, the trimerous 

 flowers and also evidently some at least of the other cases 

 with only one bract at base, are real instances of polyphylly, 

 due to variations in the divisions which the primordia of a 

 flow6r normally tnidergc*. 



My conception oi the process of synanthy is that, owing 

 to variations in phyllotaxy, two independent floWer primordia 

 become so closely approximated that they partly coalesce, 

 and develop harmoniously into a single flower Ih a some- 

 iviiat similar fashiotf to tlie growth of a plant chimera (sec- 

 torial chinlera) as described by Baur and by Winklei'. 



Teicotyly. — ^A number of cases of tricotyly and other 

 abnormalities of the cotyledons ha.ve been observed in my 



' Penzig (1890) states that in O. biennis pentamerous flowers ate 

 common, the number of "carpels" often running up to 9. 



