GROWTH AND MOVEMENT 441 
the embryo initial from the other), it is assumed that the two hemi- 
spheres are unlike, even though no structural differences are visible. 
This is expressed by the term polarity, after the analogy of the invisible 
differences in the two ends or poles of the magnet. A like polarity must 
be imputed to all other cells, its progeny, so that the embryo initial, when 
it develops, produces at the one end a root and at the other a shoot. 
Later in life, any piece of the shoot cut away from the rest shows a ten- 
dency to produce shoots at the apical end and roots at the basal end, when 
put under conditions to regenerate lost organs. The conception of 
polarity in the cells is thus extended to aggregates of cells of any size, 
because they show such differences at the apical and basal ends. All 
attempts to ascertain the nature of polarity have so far proved futile, 
so that there is nothing to "explain " the phenomena but the word and 
the assumption for which it stands. 
Correlations. The term correlation designates the reciprocal influ- 
ence of organs. Of this little is known beyond the fact that the "suppres- 
sion or the removal of one organ exercises a marked effect upon some or 
all of the remaining ones. Many examples might be cited, but no ade- 
quate explanation of the effects can be given. It is known, however, 
that at least some of them are not due merely to differences in the food 
or water supply, or to like conditions. Examples will make clear what 
is meant by correlations. 
Quantitative correlations. In the axil of each cotyledon of the bean there is 
present a bud, neither of which develops into a shoot unless the main axis is cut 
off or prevented from developing. If one desires sweet peas and such plants to 
continue flowering, it is necessary to cut away the older flowers or the young pods, 
so as to prevent the formation of fruit. If this is done, the plants go on flowering 
till frost, whereas their season is quickly over when allowed to set seed. The 
gametophyte of ferns is short-lived, as a rule ; but if the fertilization of the egg 
be prevented, its life may be prolonged for months, and it proliferates, forming 
archegonia again and again. The possibility of shaping a tree by judicious prun- 
ing, and of increasing the production of fruits by orchard trees in the same way, 
rests upon like reactions. 
Qualitative correlations. Correlations are not merely quantitative, as the above 
examples might seem to imply ; they are also qualitative. That is, the whole be- 
havior and even the structure of an organ may be altered according as other organs 
are present or absent. Thus, the central axis of most conifers is strictly radial in 
structure and in branching, while the lateral branches are distinctly dorsiventral. 
But if the terminal shoot be cut away, one (or more) of the laterals may become 
erect, losing entirely the dorsiventrality, and becoming radial like the leader. The 
aerial shoots of the potato, which bear foliage leaves and flowers, are very different 
from the subterranean ones, which bear the scales and tubers. But if the aerial 
