126
B. M. SHAFFER
of centres they had observed in both large and small populations.
Comment. The I-cells, which were initially picked out by their size,
were reported to be on average very much bigger than the R-cells. But
the fact that their ranges did not overlap is of doubtful significance,
because the I-cells were compared not with a whole population sample
but with R-cells chosen 'at random', and because the cells 'display an
unfortunate propensity towards continual alterations of shape and size
. . . which makes any inference based upon size weakly founded'
(Sussman, 1955b). Measurements were taken from photomicrographs
(R. R. Sussman et al., 1960) at densities at which, because of intercellular
contact, even living I-cells could often be recognized only after sustained
observation (Ennis and Sussman, 1958b). I-cells also differed from Rcells in more than half-a-dozen aspects of morphology and behaviour;
but as many of the characteristics of I-cells were observed only at
higher magnification after the cells had been selected by size, it is possible that their occurrence in other cells may have been underestimated.
In their earlier experiments, Ennis and Sussman confused cells slightly
larger than average with I-cells, and to eliminate this 'subjective bias'
they subsequently chose populations that definitely either did or did not
contain I-cells; but since they ignored doubtful cases, this did not help
to decide whether the population was discontinuous or not. Recently,
the possibility of continuity has been admitted (R. R. Sussman et al.,
1960), which would seem to agree better with the postulated spectrum
of initiating power.
Konijn and Raper (1961) concluded that I-cells were not a separate
class; and Gerisch (1961a) was unable to detect even one large cell in his
cultures, all of which were grown in suspension. They suggested that
large cells might be produced by unfavourable conditions. The importance of culture conditions has been demonstrated also in D. mucoroides
and purpureum, in which Bonner (1960) found giant cells on nutrient
agar but not on non-nutrient agar. Quite apart from the question of the
I-cells' role in aggregation, a steady-state ratio of cells of different type
or size would be of the greatest interest, but these workers have been
unable to detect one.
Dr. Sussman has been kind enough to send me his haploid D. discoideum NC4. In clones expanding over an ordinary agar plate spread with
bacteria, large cells and all intermediate sizes eventually become numerous in some sections of the feeding front, especially at very low densities.
Perhaps they are the descendants of 'anomalous I-cells', like those described by R. R. Sussman et al. (1960). Their proportion declines steeply
behind the front and is thus highly dependent on the developmental
stage. And whereas in Ennis and Sussman's experience, all cells in the
I-cell size range possessed a number of additional, supposedly diagnostic
B. M. SHAFFER
of centres they had observed in both large and small populations.
Comment. The I-cells, which were initially picked out by their size,
were reported to be on average very much bigger than the R-cells. But
the fact that their ranges did not overlap is of doubtful significance,
because the I-cells were compared not with a whole population sample
but with R-cells chosen 'at random', and because the cells 'display an
unfortunate propensity towards continual alterations of shape and size
. . . which makes any inference based upon size weakly founded'
(Sussman, 1955b). Measurements were taken from photomicrographs
(R. R. Sussman et al., 1960) at densities at which, because of intercellular
contact, even living I-cells could often be recognized only after sustained
observation (Ennis and Sussman, 1958b). I-cells also differed from Rcells in more than half-a-dozen aspects of morphology and behaviour;
but as many of the characteristics of I-cells were observed only at
higher magnification after the cells had been selected by size, it is possible that their occurrence in other cells may have been underestimated.
In their earlier experiments, Ennis and Sussman confused cells slightly
larger than average with I-cells, and to eliminate this 'subjective bias'
they subsequently chose populations that definitely either did or did not
contain I-cells; but since they ignored doubtful cases, this did not help
to decide whether the population was discontinuous or not. Recently,
the possibility of continuity has been admitted (R. R. Sussman et al.,
1960), which would seem to agree better with the postulated spectrum
of initiating power.
Konijn and Raper (1961) concluded that I-cells were not a separate
class; and Gerisch (1961a) was unable to detect even one large cell in his
cultures, all of which were grown in suspension. They suggested that
large cells might be produced by unfavourable conditions. The importance of culture conditions has been demonstrated also in D. mucoroides
and purpureum, in which Bonner (1960) found giant cells on nutrient
agar but not on non-nutrient agar. Quite apart from the question of the
I-cells' role in aggregation, a steady-state ratio of cells of different type
or size would be of the greatest interest, but these workers have been
unable to detect one.
Dr. Sussman has been kind enough to send me his haploid D. discoideum NC4. In clones expanding over an ordinary agar plate spread with
bacteria, large cells and all intermediate sizes eventually become numerous in some sections of the feeding front, especially at very low densities.
Perhaps they are the descendants of 'anomalous I-cells', like those described by R. R. Sussman et al. (1960). Their proportion declines steeply
behind the front and is thus highly dependent on the developmental
stage. And whereas in Ennis and Sussman's experience, all cells in the
I-cell size range possessed a number of additional, supposedly diagnostic