THE ACRASINA
151
evidence (Section II, A, 1), dominates that of the attractor made by
the aggregations. When cells themselves produce the stimulus to crowd
together and yet remain separate (Figs. 11 and 12), again at least two
factors must be in balance. For example, if the cells started secreting
or reacting to a diffusible metabolite that slowed them down, chemokinetic condensation would eventually be halted by the resultant repellent gradient.
D. purpureum must make attractor long before any clumps or centres
are visible, because Samuel (1961) has found that waves of oriented
inward and outward movement alternately, spread out from the middle
of a cloud as it forms. Phases of random motion separate them. The inward movement may reasonably be assumed to be due to a propagated
induction of acrasin secretion. The resultant crowding of the cells may
be expected to create a difference in repellent concentration, which
might account for outward movement after the attractor front had
passed. However, the cells can move out again from the middle only in
so far as they pack still closer together. If therefore seems probable that
they are guided by a reversed acrasin gradient. This is further evidence
that acrasin secretion does decline again after each front (Section IV, B,
1). That the decline has such an effect in this species may be attributed
to the particular characteristics of the pulse. As there are some grounds
for supposing that the attractor increases speed (Section IV,. A, 2), it is
most interesting that this steadily rises during inward movement and
falls during outward movement. If the attractor is responsible, it would
appear that concentration rather than gradient is the determinant,
because speed remains high during the random phase coinciding with
the crest, and low during that coinciding with the trough. But this is
not the only interpretation of the cycle, and not, indeed, Samuel's.
A young mucoroides cloud is more difficult to explain. In strain S2,
there may in fact be transient periods of cell elongation towards one or
more points within the cloud. Such periods are especially common under
a layer of agar, for this delays the appearance of definite centres (Shaffer,
1961b). But in Bonner's strain 11, Samuel has been unable to detect any
such phases. The suggestion that a cloud of unoriented cells may result
from chemokinesis (Shaffer, 1959) cannot apply here, because the cells
just outside a cloud are continuously oriented towards it; it too must
therefore make attractor. Possibly all the cells as they mature start to
secrete a small amount of it without needing the external inducer. But
it may be that its secretion is induced, and the cells are unoriented
inside the cloud because, before a definite centre appears, they yield
either only one weak pulse that is sustained too long for there to be any
general internal gradient, or multiple pulses that are, on average, so
brief relative to the variation in the time course of secretion, and
151
evidence (Section II, A, 1), dominates that of the attractor made by
the aggregations. When cells themselves produce the stimulus to crowd
together and yet remain separate (Figs. 11 and 12), again at least two
factors must be in balance. For example, if the cells started secreting
or reacting to a diffusible metabolite that slowed them down, chemokinetic condensation would eventually be halted by the resultant repellent gradient.
D. purpureum must make attractor long before any clumps or centres
are visible, because Samuel (1961) has found that waves of oriented
inward and outward movement alternately, spread out from the middle
of a cloud as it forms. Phases of random motion separate them. The inward movement may reasonably be assumed to be due to a propagated
induction of acrasin secretion. The resultant crowding of the cells may
be expected to create a difference in repellent concentration, which
might account for outward movement after the attractor front had
passed. However, the cells can move out again from the middle only in
so far as they pack still closer together. If therefore seems probable that
they are guided by a reversed acrasin gradient. This is further evidence
that acrasin secretion does decline again after each front (Section IV, B,
1). That the decline has such an effect in this species may be attributed
to the particular characteristics of the pulse. As there are some grounds
for supposing that the attractor increases speed (Section IV,. A, 2), it is
most interesting that this steadily rises during inward movement and
falls during outward movement. If the attractor is responsible, it would
appear that concentration rather than gradient is the determinant,
because speed remains high during the random phase coinciding with
the crest, and low during that coinciding with the trough. But this is
not the only interpretation of the cycle, and not, indeed, Samuel's.
A young mucoroides cloud is more difficult to explain. In strain S2,
there may in fact be transient periods of cell elongation towards one or
more points within the cloud. Such periods are especially common under
a layer of agar, for this delays the appearance of definite centres (Shaffer,
1961b). But in Bonner's strain 11, Samuel has been unable to detect any
such phases. The suggestion that a cloud of unoriented cells may result
from chemokinesis (Shaffer, 1959) cannot apply here, because the cells
just outside a cloud are continuously oriented towards it; it too must
therefore make attractor. Possibly all the cells as they mature start to
secrete a small amount of it without needing the external inducer. But
it may be that its secretion is induced, and the cells are unoriented
inside the cloud because, before a definite centre appears, they yield
either only one weak pulse that is sustained too long for there to be any
general internal gradient, or multiple pulses that are, on average, so
brief relative to the variation in the time course of secretion, and