1899.] NATURAL SCIEXCES OF PHILADELPHIA. 107 



not appear to have been hitherto noticed. The outer five have an 

 incurved character, but the inner three curve outwardly, and the 

 outer surface of the upper and lower anthers meet face to face. The 

 pistils have an outward trend, or we may say recurve, as the three 

 inner stamens do. I say pistils, for though it is customary to look 

 on the flower as having but a single deeply cleft pistil, the three- 

 valved ovarium clearly indicates a tricarpellary origin. 



The fact that but a single seed is found in the capsule must be 

 attributed to abortion, just as we find but a single seed in many 

 pluri-carpellary rosaceous fruits; and to the same principle of abor- 

 tion that has already operated in breaking up the pentamerous 

 character on which, as we have seen, the flower is normally planned. 

 Instead of a relationship to Chenopodiacea", as usually suspected, 

 or with Caryophyllacere, as has been suggested, its proper position 

 is evidently nearer Rosacese than with either. In Sjnrcea, for 

 instance, say Spircea opuUfoUa L., we have the persistent five-cleft 

 calyx, but only three carpels in many instances, with remarkable 

 irregularity in the number of the seeds. Then, proceeding to 

 Nevlusia, we have the permanent calyx doing service for petals just 

 as we find in Fagopynim, and we have a little disc on which are 

 the stamens, corresponding to the glands on which are the Fago- 

 pynim stamens. Further, we have the ovaries reduced to two, 

 with a single ovule. 



This conception of the relationship of Polygonacece to the 

 Rosacese enables us to account for the stipules, characteristic of 

 Rosacese, but which are wanting in Chenopods. That the ochrea 

 in Polygonacese is but a stipule which may become united so as to 

 clasp the stem is universally conceded. 



A feature that has been overlooked in Fagopyrwn is particularly 

 instructive. Even the most recent descriptive work, Britton and 

 Brown's Illustrated Flora of the Northern States and Canada, says 

 of it, * * glabrous except at the nodes. ' ' But one-half the stem and 

 its inner face are hairy, only the exterior half is glabrous. It has 

 been wholly my work to show that leaves do not originate at the 

 node from which the leaf- blade springs. I contend that the whole 

 of axis or stem is made up of the sheathing bases of the leaves, the 

 blade being simply the departure of the upper portiou, when the 

 cohesive power has been reduced in force. The margin of the leaf- 

 blade in Fagopyrum is ciliate. The line of hair on the stems of 



