136 Rhodora [JULY 
In closing this general discussion, the writer must call attention to 
the desirability of much further field work in connection with the genus 
Amelanchier. Our knowledge of the region around Lake Superior, 
Lake Michigan, and Lake Huron, and of the Dakotas, Iowa, Minne- 
sota, and Lake Winnipeg is especially weak, and in regard to the 
Southern States, also, information is sadly lacking. The present 
treatment can be considered only tentative, therefore, so far as these 
regions are concerned. 
In the future much better specimens of Amelanchier should be made. 
The herbaria are now clogged and overloaded with flowering specimens 
unaccompanied by fruit or leaves. For accurate work these are 
worthless, and the writer cannot guarantee to name specimens of that 
sort accurately. To prepare good specimens collections should be 
made from the same plant at flowering time just as the first petals 
begin to fall, at the time when the fruit is half grown, and at the matur- 
ity of the leaves. Mature ripe fruit is nearly useless. The mature 
leaves are not absolutely necessary, but the other two collections are 
indispensable. If time and opportunity will allow, still other stages 
in development should be obtained, and in this respect Mr. W. M. 
Blanchard cannot be too highly praised for setting an excellent example. 
Full notes should be taken in regard to the habitat, habit, time of 
flowering and fruiting, and these should be appended to the label. 
The following is a synopsis of the more important characteristics 
of the various species. For the drawings of leaves and young and 
old fruits of the different species the writer is indebted to the skill and 
generosity of Mr. F. Schuyler Mathews. 
A. Flowers racemose: leaves from acute to rounded or cordate at the base, 
conduplicate when young!; petioles slender, 8-25 mm. long: hypan- 
thium and summit of the ovary various: base of the style abruptly 
inserted on the nearly flat summit of the ovary: fruit globular, ovoid, or 
Lag arr the leaves coarse (on average leaves 3-5 (6) per em.): veins 
conspicuous, usually straight, parallel, and close together, short 
intermediate ones few or none (Ulmus- or Alnus-like): summit of 
the ovary woolly: hypanthium open, saucer-shaped, constricted 
1 In the buds of the first seven species the leaves are conduplicate, and on emerging 
from the bud remain for some time closely folded together. Instead of being equit- 
ant, however, they are placed side by side in the peculiar manner shown in Fra. 5-H. 
In the eighth species, A. Bartramiana, the leaves are not folded, but more or less 
imbricated in the bud, as in Fra. 8-H. The leaves of this species, on emerging from 
the bud, are nearly flat (see Gray's Man., ed. 7, p. 459). This isa strange difference 
to be found between species of the same genus, but à similar condition is found in 
plums and cherries. 
