203 
thing, upon the angle which the leaves assume with refer- 
ence to the stalk. It is also probably influenced somewhat 
by the form of the leaves. In those cabbages of which the 
leaves tend to spread out upon the ground, those forming 
_ the head must make a complete semi-circle in order to meet 
at the top. If the leaves grow very spreading and are 
roundish, as in Fottler’s Brunswick, the ends donot always 
come together, hence the top of the head does not grow so 
rapidly as the sides, and a flat or oblate head is the result. 
On the other hand, if the leaves grow nearly erect, as in 
the Pomeranian cabbage, a slight curvature of the lower 
part of the midrib is sufficient to bring the ends together, 
and a conical head is the result. 
In each of the divisions based upon the form of the head, 
Wwe may assume to be possible both red and green varieties; 
hence the last subdivision is that based upon the color of 
the foliage. There are as yet no red Savoys, but we do not 
know that such varieties are impossible, as some of the 
red cabbages show a tendency to blister. A further sub- 
division might be made, based upon the degree of undula 
tion of the borders of the leaves, but as this is often extremely 
variable, it has not seemed wise to extend the scheme. The 
complete classification is therefore as follows: 
