106 PROCEEDINGS OF THE ACADEMY OF [1899.. 



style is really made up of two, covered by a membraue evidently 

 of staminal origin. 



X. PhYLLOTAXIS IX CONNECTION WITH ChESOPODIACE^E 

 AND POLYGONACE.E. 



Phyllotaxis has not been made much use of by systematists. In 

 a general Avay recognition has been given to the alternate or oppo- 

 site arrangement of the foliage in connection with orders or genera. 

 We do not expect to find alternate leaves in Rubiacete, and 

 opposite leaves in Rosacese would be a surprise. But we may go 

 further. In a general way we may look for a pentamerous arrange- 

 ment of the floral organs, whenever we find a f arrangement of the 

 leaves. In Chenopodiacese and genuine allies, we find the ^ plan 

 prevail. We may doubt the relationship with Chenopodiacese of 

 any plant with a | arrangement. 



In Polygonacese we find the f arrangement the rule, and a close 

 study of some species of the order shows that, though there are' 

 great deviations from a pentamerous structure, the pentamerous is 

 the main plan, and the apparent deviations are the results of abor- 

 tion, or of a union of parts usually separate. 



FcKjopyrum escnlentum well illustrates this. The outer series 

 which we take to be the calyx is pentasepalous. There are eight 

 stamens, but if we examine the flower closely we find that there 

 are only five in the verticil succeeding the calycine whorl, and 

 these are alternate Avith the sepals as they should be. The next 

 verticil appears to be of eight glands; but on close inspection there 

 are but five, and these are alternate with the stamens. Then we 

 have three rather smaller glands, just above the line of the larger 

 ones, and making the eight which come under casual notice, 

 and these three are at the base of the three stamens, which make 

 the eight usually credited to the species. Here a puzzle presents 

 itself. There should be two more glands to make a complete ver- 

 ticil of five, and two more stamens for another perfect verticil, and 

 how does it come about that the stamen is just above the gland and 

 not alternate therewith ? We cannot answer ; but with a clear 

 case of three perfect verticils on the plan of five, we may assume 

 that in some way, as yet unaccounted for, abortion and disarrange- 

 ment of parts at this point have occurred. 



The three upper stamens present an interesting feature that does 



