NOTES FROM A GARDEN HERBARIUM— 11. 



83 



The general habit of the dewberries and blackberries 

 is clearly different. The dewberries are trailing, with 

 a tendency to root at the joints, while the blackberries 

 are erect or ascending. But these characters seem to 



Fig. I. Bartell Dewberry — Rubus Canadensis, 



VAR. INVISUS. 



be elusive in some forms, of which the dewberries I am 

 about to describe are examples. 



Rubus Canadensis is distinguished from its varieties 

 by its comparatively weak stems, low habit, ovate elliptic 



and finely toothed leaflets, short and often divergent 

 peduncles, small flowers and fruit. In cultivation, it 

 appears to be represented only by the variety once sent 

 me as " Lucretia's Sister," Fig. 2. I do not know that 

 this variety possesses any merit. 



The Bartell and Mammoth dewberries are dis- 

 tinguished from the species by a different habit, by 

 long, erect and smoother peduncles, and by very dif- 

 ferent^ leaves. This variety I shall call Rtilnis Cana- 

 densis, var. inz'isns, new variety [in-'isns, "overlooked"). 

 Stem stout and stiff, often partially ascending; leaflets 

 much larger than in the type, broad and thin, smooth 

 or nearly so, the teeth very large, simple and rounded, 

 with a minute point at the end of each : peduncles 

 long and straight ; young flower buds bearing a promi- 

 nent tip, formed by the connivent slender ends of the 

 sepals ; flowers and fruits much larger. All these dif- 

 ferences are well shown in Figs, i and 3. The de- 

 tached leaflet in Fig. i is an admirable likeness, con- 

 siderably reduced, of the leaflets of this variety ; the 



Fig. 2. " Lucretia's Sister " — Rubus Canadensis. 



same figure also shows the ascending character of the 

 stem. 



This variety invisus grows in open woods here at 

 Ithaca, where it looks very different from either Ridnis 

 Canadensis or R. villosiis. Professor Dudley, who has 

 given it considerable attention, thinks that it may be 

 found to possees sufficient distinctness to make it worth 

 specific rank. It is certainly more unlike the type of 

 R. Canadensis than is the var. 7-orihaceus. 



My specimens of the Bartell and Mammoth varieties 



