142
B. M. SHAFFER
that the inducer has a threshold. Experimental treatment that causes a
large stippled aggregation to break up more or less permanently into
hundreds of small clumps and rings very likely induces the cells to sustain maximal secretion while still separate, and thus be unable to propagate a pulse. The amount a cell secretes at each pulse perhaps depends
on the amount of acrasin or precursor accumulated in the interval, and
possibly increases with successive pulses as the metabolic apparatus
matures. Certainly, in Arndt's film, the pulses follow one another at
shorter intervals and become more vigorous; and cells in a stippled
aggregation move faster and probably become more adhesive at each of
at least the first few pulses. Waves transmitted across young populations
could vary in strength by recruiting different proportions of the cells.
As a secreting cell can be oriented by the attractor, this must act
differently when outside the cell; the foregoing scheme makes it likely
that this is true of the inducer also.
Jaffe (1958) has tried to account for certain wave forms in streams,
for example, the alternate concentric zones of sparse and crowded
cells sometimes seen in D. polycephalum (Raper, 1956b), without invoking cyclical changes in the cells. He suggests they may be due to the
lag with which a single chemotactic pulse is relayed from one group of
cells to another, the lag determining the wave length; but additional
assumptions seem to be needed to account for the cells' reacting in
groups and for the dimensions of the group. This hypothesis apparently
predicts that the waves should move inwards; whereas if they are
multiple pulses of increased activity, whether of the kind already discussed or produced mechanically, they should move outwards.
As the aggregating cells induce one another not only to secrete acrasin
but to become mutually adhesive too, they form chains of increasing
length; and the cells within them are then guided by contact following.
The waves of rapid inward movement sometimes propagated along
continuous streams of the large Dictyostelium species may well be produced by increased acrasin secretion, though so far it has not been shown
even to be associated with them; or their origin may be mechanical.
Possibly both factors are responsible, because of a causal interrelation
between secretion and movement. Such pulses may reorient cells within
broad flat sheets. In all continuous streams, the marginal cells must
experience a transverse gradient tending to prevent them from escaping.
This influence, in conjunction with contact following, explains why the
width of a stream, though it may vary enormously from one section to
another according to the number of cells locally available, does not
fluctuate violently between the points at which adjacent tributaries
join it, but either is uniform or tapers gently. If a stream breaks, the tip
of the outer section is free to be attracted towards the stump or an ad ja-
B. M. SHAFFER
that the inducer has a threshold. Experimental treatment that causes a
large stippled aggregation to break up more or less permanently into
hundreds of small clumps and rings very likely induces the cells to sustain maximal secretion while still separate, and thus be unable to propagate a pulse. The amount a cell secretes at each pulse perhaps depends
on the amount of acrasin or precursor accumulated in the interval, and
possibly increases with successive pulses as the metabolic apparatus
matures. Certainly, in Arndt's film, the pulses follow one another at
shorter intervals and become more vigorous; and cells in a stippled
aggregation move faster and probably become more adhesive at each of
at least the first few pulses. Waves transmitted across young populations
could vary in strength by recruiting different proportions of the cells.
As a secreting cell can be oriented by the attractor, this must act
differently when outside the cell; the foregoing scheme makes it likely
that this is true of the inducer also.
Jaffe (1958) has tried to account for certain wave forms in streams,
for example, the alternate concentric zones of sparse and crowded
cells sometimes seen in D. polycephalum (Raper, 1956b), without invoking cyclical changes in the cells. He suggests they may be due to the
lag with which a single chemotactic pulse is relayed from one group of
cells to another, the lag determining the wave length; but additional
assumptions seem to be needed to account for the cells' reacting in
groups and for the dimensions of the group. This hypothesis apparently
predicts that the waves should move inwards; whereas if they are
multiple pulses of increased activity, whether of the kind already discussed or produced mechanically, they should move outwards.
As the aggregating cells induce one another not only to secrete acrasin
but to become mutually adhesive too, they form chains of increasing
length; and the cells within them are then guided by contact following.
The waves of rapid inward movement sometimes propagated along
continuous streams of the large Dictyostelium species may well be produced by increased acrasin secretion, though so far it has not been shown
even to be associated with them; or their origin may be mechanical.
Possibly both factors are responsible, because of a causal interrelation
between secretion and movement. Such pulses may reorient cells within
broad flat sheets. In all continuous streams, the marginal cells must
experience a transverse gradient tending to prevent them from escaping.
This influence, in conjunction with contact following, explains why the
width of a stream, though it may vary enormously from one section to
another according to the number of cells locally available, does not
fluctuate violently between the points at which adjacent tributaries
join it, but either is uniform or tapers gently. If a stream breaks, the tip
of the outer section is free to be attracted towards the stump or an ad ja-