MEMOIRS FROM THE DEPARTMENT OF BOTANY OF COLUMBIA COLLEGE. 
No) 
generic validity of such individuals or groups as Thysanella, Polygonella and Fagopyrum. 
Linnaeus included the latter two groups under Polygonum, the former not being known 
in his time. In my opinion all three form perfectly good genera. We have excellent 
characters in the case of Fagopyrum. This type resembles Polygonum in its habit more 
than the other two forms in question, but its tissues have a characteristic texture and its 
leaves a distinctive shape. The most important differences, however, exist in the flower 
and fruit. The calyx withers, or remains unchanged in fruit, the faces of the achene 
are pinnately striate and the angles more or less margined or crested, and the embryo 
is central, the broad cotyledons almost dividing the mealy albumen into halves by an 5- 
shaped curve. None of these characters are possessed by Polygonum. In Polygonella 
we find a very strong character in the habit of the species, and they all possess it in about 
the same degree. Here again the flowers furnish characters to separate this type from 
Polygonum. In the first place, the pedicels are normally solitary, whereas in Polygonum 
they are fascicled, and secondly the calyx either remains unchanged or the three inner 
segments develop into conspicuous wings in fruit, two characters not belonging to Poly- 
gonum. Thysanella has good characters, but as we are not directly concerned with it, 
suffice it to say that its affinities are closer to Polygonella than to Polygonum, and as the 
former genus is distinct from the latter, if Thysanella could not stand generically distinct 
it would naturally fall into Polygonella and not into Polygonum. 
Polygonum proper therefore consists of herbaceous or shrubby plants, often con- 
spicuous on account of their prominently jointed stems. They are either terrestrial, 
amphibious or aquatic, and their vegetative organs may be glabrous, glaucous, variously 
pubescent or glandular. Some species are annual, others are perennial by means of large 
roots, creeping stems or rootstocks. The roots are variable in size, sometimes slender and 
fibrous, often large, fusiform and woody. In dry clay soil the development is usually 
restricted, while in sand they often elongate to a remarkable degree. ‘The subgenus Bis- 
torta furnishes two kinds of fleshy rootstocks; the one is bulb-like and the other more or 
less elongated, either creeping or horizontal. The texture of the tissues of the stems 
varies from herbaceous to woody, and the habits it has assumed are manifold. Some are 
erect or prostrate, some float in or on the water, others climb by means of recurved prick- 
les, and still others are scandent by twining stems. They may be strict or flexuous, in 
most cases unarmed, in a few furnished with strong recurved prickles, and the internodes 
appear as terete in some cases, in others they are channeled or ridged, while in many 
species they are gradually or abruptly enlarged towards the nodes, a character so prom- 
inent in this genus. The leaves are alternate and follow in a @ or rarely in a 3 spiral. 
They vary much, like the other vegetative organs, in their texture and shape, the bases 
