8 BULLETIN 622^ U. S. DEPARTMENT OF AGRICULTURE. 
The purple colors are in the glumes themselves. The black colors 
come from a melaninlike pigment, which in the hulled varieties 
occurs in the lemmas. This may exist in different degrees and has 
been used to subdivide black barleys, although the advisability of 
separating on degree is questionable. 
In the naked varieties the colors are determined more easily. The 
absence of pigment results in a white or yellowish grain, as in the 
hulled. The blue colors are located in the aleurone layer. In mature 
specimens they are unmistakable ; in immature ones the color is slight 
and may give only a bluish cast, as it is deposited late in the develop- 
ment of the kernel. In the white varieties immaturity may result in 
a greenish appearance which might be confused with immature blue 
kernels. The purple color results from a blue aleurone layer beneath 
a red pericarp. In all purple varieties at present known the color is 
well developed except in very immature specimens and can be deter- 
mined without difficulty. Black naked varieties with a true melanin 
pigment are very rare. Purple naked ones are usually referred to as 
black, although in reality they contain no black pigment and are 
separated easily from those having the black pigment. 
DENSITY. 
Density unfortunately has been used as a major distinction in the 
classification of barley and usually has been confused with fertility. 
This is the most unsatisfactory of all the six characters. The behavior 
of hybrids indicates that density is determined by heritable factors, 
however, and further information upon its behavior may increase 
greatly its value as an agronomic character useful in the description 
of biotypes. As may be seen in Plate I, in typical forms, such as 
Peacock, Plumage, and Hanna, representing zeocriton, erectum, and 
nutans^ respectively, the differences are marked. The intergradations 
are so numerous and so complete that any broad application results 
in confusion. Statements in taxonomic publications on barley indi- 
cate that all investigators have had great difficulty in making satis- 
factory separations. The groups have been divided into dense and lax 
varieties. In some cases the dense varieties have been further sub- 
divided into dense and very dense groups. Kornicke in some of his 
treatments has divided even the very dense into two subdivisions. 
Such confusion is inevitable wherever separations are made upon 
degree, and the use of density has been confined in this paper to the 
description of subvarieties. 
SCHEME OF CLASSIFICATION. 
A very superficial survey of the combinations that are possible with 
the variable factors just described is sufficient to indicate the scope 
