100 Rhodora [May 
Scribner ! from Homer, Alaska, was set off, differing in having short 
thick spikes, 3-4 cm. in diameter, and longer floral parts. On dissect- 
ing, it can be seen that the grain is longer and lanceolate in outline, 
instead of oblong as are the shorter grains of E. mollis and E. arenarius. 
Dissection of the grain of E. capitatus shows within, instead of hard 
compact stored food material, a mass of spongy white tissue only 
partially filling the space. Specimens such as Piper's no. 4664 from 
Kenai, Alaska, Aug. 18-20, 1904, bearing from one rootstock two 
flowering culms, one of which is undeniably E. mollis Trin., and the 
other as surely FE. capitatus Scribn., indicate clearly enough that Æ. 
capitatus Scribn. is a pathological state of E. mollis Trin. In a later 
article * published with Merrill, Scribner himself included E. capitatus 
under E. mollis Trin. 
From northwestern North America E. villosissimus Scribner 3 
of this same group was described, but this plant is so very like E. 
arenarius, var. villosus that it is not easy to discuss its differences. 
Scribner and Merrill * in their * Grasses of Alaska" fail to add any new 
characters, although they maintain Æ. villosissimus as a species. Their 
key difference is 
Spike less than 8 cm. in length....................... E. villosissimus. 
Spike exceeding 10 om. in length. ........................ E. mollis, 
In the description following, which is a condensed form of the original 
one, E. villosissimus is allowed to have spikes from 5 to 8 em. in length. 
According to this arrangement what is to be done with plants which 
have spikes less than 5 cm. in length, or more than 8 and less than 10 
cm. in length? Such plants are no less common than those which fall 
between the artificial limits set to define E. villosissimus. The plant 
called E. mollis has been found in the far north with spikes as short as 
4 em. and near its southern limit with spikes as long as 35 em. Any 
large collection will show from intermediate regions a complete series 
of specimens with no appreciable break in the length of spikes. From 
this it is plain to the author that E. villosissimus Scribn. must be 
treated as a synonym of E. arenarius, var. villosus. 
The older herbarium-sheets of E. arenarius, var. villosus show an 
interesting tide of opinion, almost all giving evidence of numerous 
1 Scribner, Bull. U. S. Div. Agrost. xi. 55, pl. xiv (1898). 
? Scribner & Merrill, Contrib. U. S. Nat. Herb. xiii. Pt. 3, 90 (1910). 
3 Scribner, Bull. U. S. Div. Agrost. xvii. 326, fig. 622 (1899). 
‘ Scribner & Merrill, 1. c., 89 (1910). 
