l88o.J BOTANICAL GAZETTE. 5 1 



South Twin, James Bay. Collected July 17, 1887, by 

 Mr. James M. Macoun, for whom it is named. 



"A small, compact bush, 2 to 4 feet high, with just the 

 habit of a garden currant, growing in peaty soil by a small 

 pond in company with S. arctica." This association, taken 

 in connection with the resemblance which the leaves bear 

 to certain Canadian forms of S. arctica, might suggest the 

 question whether our plant was due to a cross with that spe- 

 cies, but the aments do not show a trace of any such admix- 

 ture. Had it not been for the intermediate character of the 

 Labrador specimens I should have taken this for an extreme 

 modification, either of S. Richardsonii, or of the Greenland 

 S. lanata, deserving to rank as a good species. Its occur- 

 rence on the level of the plain in hit. 53 20', further south 

 than any form of either species mentioned has ever been 

 found before, is in keeping with the essentially arctic vege- 

 tation by which it is accompanied and the physical conditions 

 described by Mr. Macoun in his interesting " Notes on the 

 Flora of James Bay" (Bot. Gazette, xiii. 113). 



Plate IX. S. Richardsonii Hook., var. Macouniana Bebb. Capsule 

 and stamens X&. 



3. S. Barrattiana Hook. Alpine swamps in the Rocky 

 Mts., Drnmmond! In thickets at high elevations, Kicking- 



Wh 



Prof- yohn 



M 



a single collection and remains thenceforth for more than 

 half a centurv known to science only through the type speci- 

 mens, it is interesting to compare the rediscovered plant with 

 the original description and note in how far, if at all, it may 

 be necessary to modify the characters at first assigned. In 



icoun's specimens, the full-grown leaves, 2\ inches 

 \ inch broad, are elliptic-oblanceolate, pointed at 

 both ends, not at all "blunt and cordate at the base," but 

 with this exception the agreement throughout with the excel- 

 lent description given by Hooker is perfect, even to the "sti- 

 muli's olabriusculisr This character, omitted by Andersson, 

 would appear to be constant. It presents certainly a striking 

 peculiarity, the like of which I have not observed in any 

 other willow. Set upon a hirsute stem, at the base of silky- 

 pubescent leaves, and thus in marked contrast with the sur- 

 faces on either side, we find the stipules yellowish-green and 

 glabrate ! The styles and stigmas are reddish-brown. The 

 dark -brown twigs. " marked with the scars of former years' 

 leaves," are rendered still more shaggy in appearance b} 



