CLASSIFICATION OF AMERICAN WHEAT VARIETIES. 31 
Varieties classed as having clavate spikes are clubbed, that is, dis- 
tinctly larger and more dense at the apex. This is due to a shorten- 
ing of the rachis internodes in that part of the spike, which results 
in a change from dorsoventral to lateral flattening and a broadening 
of the upper portion of the spike. 
Elliptical spikes are short and quite uniformly rounded at both 
the base and apex, but are flattened on the sides. Most varieties of 
club wheat have spikes of this shape. 
In the descriptions of varieties these designations of spike shapes 
have sometimes been modified to take into account the length and 
width of the spikes and the overlapping of shapes which occurs in 
some varieties. 
Spikes that are unusually long are described as linear-fusiform, 
linear-clavate, etc. If spikes are unusually short that fact is in- 
cluded in the description. Broad spikes may be described as broadly 
fusiform or broadly oblong and narrow spikes as narrowly fusi- 
form, etc. 
Varieties that are nearly intermediate between any of the shapes 
are sometimes described as oblong-fusiform or oblong to subclavate. 
By the use of these compound descriptive terms spike shapes are 
more accurately presented in the description than they can be in the 
keys, where brevity is imperative. 
DENSITY OF THE SPnSIE. 
The differences in shape of spikes shown above are due in part to 
differences in density. All spikes are described as of three density 
classes, viz, lax, middense, and dense. These differences are shown 
in Plate VI. These are minor differences which are used to advan- 
tage in distinguishing varieties. Seringe {17 Jf,) separated the com- 
mon wheats into two groups, having lax and dense spikes, respec- 
tively. Koernicke and Werner {133) described the spikes of many 
varieties according to different degrees of density. Neergaard {lI{-7) 
suggested a formula for use in measuring the density of the spike. 
Eriksson {88) subdivided the botanical groups of Koernicke and 
Werner on the basis of density into subvarieties called laxuvfi^ den- 
sum^ and capitatum^. He measured the density of spikes by de- 
termining the number of spikelets in 100 mm. of rachis length. 
Heuze {112) used the spike density along with spike shape as the 
leading character in separating varieties. Boshnakian (^5) de- 
scribed means of measuring density and suggested the name Triticum 
compacto-capitatuvi for varieties of club wheat having clavate heads. 
Many measurements have been made by the writers to determine 
the difference in density of the spikes of the varieties here described. 
The most definite were found comparable at one station for one year, 
but otherwise were of little value. It was found necessary to estab- 
