8 Rhodora [JANUARY 
SoLipaco LEPIDA DC. In the region including Newfoundland and 
southern Labrador, the Gaspé Peninsula and northern New Brunswick 
and northern Maine, Solidago canadensis L. is less common than 
farther south, and S. serotina Ait. is unknown. Here their places are 
occupied chiefly by a very common goldenrod, in foliage and size of 
heads somewhat intermediate between those two species, but with 
the branches of the panicle strongly ascending or but slightly secund. 
In this northeastern area, centering about the Gulf of St. Lawrence, 
this common goldenrod, with heads much larger (involucre 3-5 mm. 
high) than in S. canadensis (involucre 2-2.8 mm. high) ! but with the 
stems puberulent or minutely pilose much as in that species, is quite 
as variable as others of the genus, but it is possible to recognize four 
somewhat distinct trends of the species. 
About the Straits of Belle Isle and in some parts of eastern New- 
foundland the plant has broadly lanceolate to oblong coarsely serrate 
leaves and a short compact thyrsus which is almost overtopped by the 
upper leaves. This plant is a good match for authentic Alaskan 
material and in foliage is closely matched by a fragment of the type of 
Solidago lepida DC. Prod. v. 339 (1836). 
On the limestone detritus of Pereé, in Gaspé County, Quebec, 
occurs a plant strongly resembling S. lepida, but with the leaves very 
densely cinereous-pilose. This cinereous extreme of the series has, 
so far as the writer can determine, received no na 
A third variant is a plant with Sosy iene eae leaves 
and elongate or somewhat rhombic scarcely leafy thyrsus. This 
plant, abounding on the gravels of the River Ste. Anne des Monts in 
Gaspé County, Quebec, seems quite inseparable from extreme speci- 
mens of the northwestern S. elongata Nutt. Trans. Am. Phil. Soc. vil. 
327 (1840). 
But by far the commonest trend of the species in the Northeast is a 
tall plant which is clearly a very large extreme of S. elongata with an 
ample panicle. Often as large but never as secund as in S. serotina, 
but with large leaves running well into the inflorescence much as in 
S. rugosa, var. villosa (Pursh) Fernald, this variety, in foliage and in 
the long branches of the inflorescence, passes directly into S. elongata 
and in the leafy character of its panicle clearly connects with the most 
characteristic S. lepida. It occurs in good development, not only in 
1See Fernald, Raopora, x. 92 (1908). 
2? Fernald, Ruopora, x. 91 (1908). 
