1911] Fernald,— Book Review 219 
in Macdougal, Vail and Shull, Mutations, Variations and Relationships oj 
Oenotheras, 70; Gray, Man. ed. 7); Goodyera pubescens is unknown east of 
central Maine (see RHODORA, i. 5); and the plants which in New Brunswick 
have passed, by error of determination, as the “King Devil Weed,” Hieracium 
praealtum, are H. florentinum, H. pratense, and H. floribundum, all European 
weeds recently introduced into our flora and now unfortunately too common 
(rather than rare) in the regions they infest. Rarity in a plant is purely 
relative, but to those who have explored the regions of Bic and of the Gaspé 
Peninsula in Quebec (small portions of the “New Brunswick Area”) it will be 
a surprise to learn that Dryopteris fragrans and Woodsia hyperborea are “rare.” 
But this term might without question be applied to some scores of species 
which Harshberger does not mention in his list of rare plants of the area: such 
species as Cystopteris montana, Phegopteris Robertiana, Woodsia scopulina, 
Calamagrostis purpurascens, Carex lagopina, Sagina occidentalis, Arabis Hol- 
boellii, Draba aurea, Dracocephalum parviflorum, Physalis grandiflora, and 
Arnica gaspensis, species which as yet are known from only one or two stations 
in the New Brunswick Area.] 
[p. 376] 
[In the list of Alpine plants of New Hampshire Dryas integrifolia Vahl 
appears just as if it had full claim to the title, although it was shown in 1903 
that its record as a New Hampshire plant was a mistake (see RHODORA, v. 
281). And surely Professor Harshberger can find no European support 
for his assertion that D. integrifolia grows in Europe or that Betula glandulosa 
is European — unless “Europe” is stretched to cover Greenland, Kamt- 
schatka and the Altai Mountains.] 
[p. 378] “b) New England Area.” 
“The flora of this area is characterized by the absence of Picea 
alba, Pinus Banksiana and by the presence of Pinus s[S]trobus and 
other trees found in the New Brunswick area, but its chief charac- 
terization depends on the presence of such trees as Quercus alba, Q. 
prinoides, Q. coccinea, Juniperus virginiana and Castanea dentata." 
[This definition is practically as ineffective as that of the New Brunswick 
Area; for, since the author begins his first two subdivisions (Sea Islands and 
Sea Coasts) with Mt. Desert Island and eastern Maine, it is difficult to recon- 
cile the discussion with the definition. Surely Picea canadensis (alba) is 
abundant on Mt. Desert Island and follows the coast to Casco Bay (see 
Rand & Redfield, Fl. Mt. Desert, Sargent, Dame & Brooks, etc.), and Pinus 
Banksiana, though not common on Mt. Desert, is abundant on the adjacent 
mainland (see Rand, Rnmopona, i. 135; Sargent, Dame & Brooks, etc.). 
But singularly enough none of the trees which characterize the New England 
area as defined are found on Mt. Desert and the adjacent coast (See Rand & 
Redfield; also A. H. Graves, Ruopora, xii. 173). In fact, the northeastern 
limit of Quercus alba is in south-central Maine (see Sargent, Dame & Brooks, 
etc.), the most northerly station being at or near Waterville; Quercus pri- 
noides (never a tree in New England) is unknown in Maine, its northeastern 
limit being in the region of Manchester, New Hampshire; Quercus coccinea is 
