96 



Baldicci placed it near Menispermaceae, Wettstein near Lauraceae. 

 Small, in his Flora of the Southeastern I nited States, 1903 (and 

 in this followed by Rydberg), placed Aristolochiaceae near Cucur- 

 bitaceae. Many botanists have pointed out the great similarities 

 of Cucurbitaceae and Passifloraceae. Habit, kidney-shaped leaves, 

 basal flowers, and the long projecting connective oi the anthers are 

 suggestive of Viola; the young buds oi Viola and Asarum are strik- 

 ingly similar. Asarum is one ol the many genera showing parietal 

 placentation at the top of the ovary of the hud, axile placentation 

 in the Bower. If we place Aristolochiaceae and related families 

 not far from Yiolaccae and Passifloraceae, the group is no longer 

 in an isolated position, because its characteristics occur more or 

 less in neighboring groups. 



The Stachyuraceae with the single genus Stachyurus were in- 

 cluded in Ternstroemiaceae by Bentham & Hooker. In the Engler 

 system it is next to Flacourtiaceae ; Hutchinson places it by I lama- 

 melidaceae. Stachyurus also has parietal placentae in the upper 

 part of the ovary, axile placentation in the lower part. It the 

 group Parietales is placed early in the system, the relationships of 

 Stachyurus are more readily understood. < )ne ot the earliest 

 shrubs to flower in our Garden, with long hanging spikes and small 

 four-parted flowers, it would seem to fit well, both in appearance 

 and in details of structure, near Corylopsis and other I lama- 

 melidaceae. 



( )n thi' opposite page is a comparison ol four systems ol di- 

 cotyledons beginning with Ranales, but not including Sympetalae. 

 There is evidently much similarity in the arrangement in these 

 systems. Marly botanists should not be given too much credit for 

 beginning with Ranales, when we remember that they placed ferns 

 after the seed-bearing plants. We see that Wettstein and Rendle 

 agree in bringing the parietal placentation groups together and 

 not far from Ranales. Two more groups should perhaps be added, 

 Cactaceae and Salicaceae. In any case there appears to be a num- 

 ber of groups of families with many connecting characteristics, 

 nearly all with parietal placentation, and not far distant Irom Ra- 

 nales. " When a group is assigned to its true place, it becomes 

 inextricable," writes 11. h". Copeland, "every character studied in- 

 creases the certainty of the assignment." 



