THE INITIATION OF STAMTNAL ZYGOMORPHY. 693 



to this question seems important, for, if the observer delays his consideration until the 

 petals are expanded or the anthers dehisce, he cannot claim to have fully determined 

 the symmetry of the flower. 



The time of anther-dehiscence or receptiveness of stigma may be given a prominent 

 place in a consideration of symmetry, but from the developmental standpoint all stages 

 in the development of a flower as a mechanism are important. 



On the one hand, then, a flower may be considered Actinomorphic, when, in the 

 expanded condition, all members of each cycle of floral parts are^ — for all practical 

 purposes — of the same size and form, and in the same condition, at any point of time. 



On the other hand, the student of floral development will not regard as actinomorphic 

 a flower, in the development of which it has been found that all members of each cycle 

 are not of the same size and form and not in the same condition at any point of time. 



It is to be remembered, however, that all the primordia of a given cycle of floral 

 parts may not arise simultaneously on the receptacle, and that in the later stages 

 inequalities of part size in the perianth may be associated with the aestivation. 

 Segregation of spore-mother-cells in equivalent parts in all the anthers of a stamen- 

 cycle may not begin simultaneously, and the formation of the fusion nucleus in all 

 ovules at one level may not be accomplished at the same moment. 



It is not proposed to consider as other than actinomorphic a flower in which are 

 observed only those departures from absolute precision ; but when it is shown that there 

 is initiated a definite tendency in any member of a cycle of floral parts to reach 

 a developmental stage in advance of the remaining members, or to have the attainment 

 of any developmental stage delayed, the flower in which this is observed is, in my 

 opinion, to be considered Zygomorphic. 



Even when the results of such tendencies are counterbalanced at a later stage, the 

 flower is to be regarded as zygomorphic. 



The grounds on which these opinions are based will now be given. 



In November 1911 the Melianthaceous (Eng. and Prnl.) plant Greyia Sutherlandii 

 (Hook, and Harv.) flowered under glass in Glasgow Botanic Gardens. The expanded 

 flowers were for the first time examined by the writer on account of the great length of 

 the stamen-filaments. It was at once evident that, although the ten stamens formed a 

 single cycle with a very slight tendency to difi"erentiate into two alternating cycles of 

 five, no two undehisced stamens, in any one expanded flower examined, were of the 

 same length (fig. 1). 



For all practical purposes the filament-bases were on one level, and the anthers 

 were of uniform size. The diff"erences in stamen-length were, then, due to inequalities 

 of filament-length. But although the stamens were not of equal length in any young 

 expanded flower, no corresponding marked inequality was observed in the perianth. In 

 some flowers the right anterior lateral sepal was slightly longer than its neighbours, 

 and the right anterior lateral portions of the corolla and staminal disc were slightly 

 larger than the remainder. The flowers were viewed from the adaxial side. 



