Differentiation in the Stem of the Sunflower. 507 
matous sheaths that surround the vessels and hold the primary xylem 
together. Ultimate partial or complete failure of accommodation to 
mechanical strains may account for the fact that the lower leaves wilt more 
readily, although they are nearer the water-supply, and finally shrivel . 1 
Correlations and their causal interpretation. -Secondly, the variations 
in degree of activity of the cambium in the leaf-trace bundle and elsewhere 
demand, and should ultimately be susceptible of, a causal explanation. 
A brief consideration of the observed correlations from this point of view 
may perhaps be not unprofitable. 
The sluggishness of the cambium in the upper part of the leaf-trace 
bundles is sufficient evidence that the greater cambial activity elsewhere is 
not determined by an abundance of food. 
The sectors which show vigorous tangential growth are synthetic traces 
always in communication with young expanding leaves near the apex. 
There is a close correlation between the growth of these sectors and the 
expansion of the shoot as a whole. 
The growth of the independent leaf-trace bundle is, on the other hand, 
closely correlated with the expansion of the leaf it serves, and tends to cease 
when the leaf has reached its full size-. 
In the lower part of the bundle, however, where primary growth comes 
to an end before the leaf is fully expanded, the primary xylem is supple- 
mented by secondary xylem. It is an interesting fact that this secondary 
growth does not cease when the growth of the leaf comes to an end, but 
continues until the flowering season brings a diminution in the rate of 
growth of the stem as a whole. Thus, although at its initiation this 
secondary growth is correlated with the growth of the leaf, after initiation 
it ceases to show the same correlation, but is correlated instead with the 
growth of the plant as a whole. 
The upper parts of the bundle reach their full size before primary 
growth has ceased around them. Secondary growth is then only slowly 
resumed, and the tissues produced reflect qualitatively as well as quantita- 
tively the relative inactivity of the cambium, which is the more pronounced 
the nearer the node. From this point of view the so-called ‘ parenchyma- 
tization ’ of the xylem may mean simply arrested differentiation. 
In the final stage of the resumption of activity, the cambium through- 
out the bundle produces typical secondary wood and behaves uniformly 
with the rest of the cambium at the same level, in correlation with the upper 
part of the shoot. This stage, however, is only represented in the older 
part of the stem, and is probably reached only when the leaves have 
fallen. 
These correlations strongly suggest that each leaf has a sphere of 
influence in the vascular zone which is localized tangentially and longi- 
1 Cf. Thoday, Proc. Roy. Soc., B. , lxxxii, 1909, p. 25. 
N n 2 
