48
GORDON A. RILEY
amount of particle formation. However, if samples treated in this
manner proved not to be entirely sterile, growth of bacteria might
provide nuclei for further aggregation and therefore might lead to
larger yield of particles. There is no doubt that bacteria become incorporated in particles, and indeed prolonged bubbling is an effective
means for sterilizing sea water, but there is no evidence that bacteria
play a physiological role in the process.
Thus it was suspected that bacteria or any organic particles of a
similar size might be equally effective as nuclei for aggregation. This
matter can be tested further by coarse filtration and some kind of
sterilization of the filtrate. However, most kinds of sterilization are
suspect on the grounds that they might introduce artifacts into the
experiment. Treatment with cyanide seemed to be one of the least
objectionable methods. This was tried in a series of replicates using
1.2 p pre-filters in which one set was bubbled for 24 hours in the usual
way, and replicate pairs were given identical treatment except for an
addition of 50 mg KCN/litre. There was no significant difference in the
results.
The experiments did not fully prove the point because the KCN
treatment did not establish complete sterility. Plate counts indicated
that there could be as many as 100-200 bacterial cells/ml at the end
of the experiment. However, this was one to two orders of magnitude
less than the quantity in untreated controls, and if living bacteria were
very important one would expect to find some difference in experimental results.
Having demonstrated that air supplies did not contribute significantly to particle formation, Batoosingh et al. (1969) carried out a few
experiments of longer duration in further examination of time relations
involved in particle formation. One experiment involved continuous
bubbling for 15 days, with daily removal of an aliquot for carbon
analysis. A concentration of 227 pg C/litre was attained on the second
day, and no further increase was observed thereafter.
Earlier experiments on alternate bubbling and filtering were repeated with better controls and more thorough replication and with
essentially the same kind of results that had been obtained earlier. I n
a series of nine experiments the mean total yield for four successive oneday bubbling periods was 492 pg C/litre and this was 3.3 times the mean
yield for the first day, indicating that particle formation can continue
for some time with only slight diminution if the particles are removed
at intervals. Another experiment was devised to provide continuous
yield, and the amount obtained during one day was larger than in four
days of alternate bubbling and filtering. I n short, the results reaffirmed
GORDON A. RILEY
amount of particle formation. However, if samples treated in this
manner proved not to be entirely sterile, growth of bacteria might
provide nuclei for further aggregation and therefore might lead to
larger yield of particles. There is no doubt that bacteria become incorporated in particles, and indeed prolonged bubbling is an effective
means for sterilizing sea water, but there is no evidence that bacteria
play a physiological role in the process.
Thus it was suspected that bacteria or any organic particles of a
similar size might be equally effective as nuclei for aggregation. This
matter can be tested further by coarse filtration and some kind of
sterilization of the filtrate. However, most kinds of sterilization are
suspect on the grounds that they might introduce artifacts into the
experiment. Treatment with cyanide seemed to be one of the least
objectionable methods. This was tried in a series of replicates using
1.2 p pre-filters in which one set was bubbled for 24 hours in the usual
way, and replicate pairs were given identical treatment except for an
addition of 50 mg KCN/litre. There was no significant difference in the
results.
The experiments did not fully prove the point because the KCN
treatment did not establish complete sterility. Plate counts indicated
that there could be as many as 100-200 bacterial cells/ml at the end
of the experiment. However, this was one to two orders of magnitude
less than the quantity in untreated controls, and if living bacteria were
very important one would expect to find some difference in experimental results.
Having demonstrated that air supplies did not contribute significantly to particle formation, Batoosingh et al. (1969) carried out a few
experiments of longer duration in further examination of time relations
involved in particle formation. One experiment involved continuous
bubbling for 15 days, with daily removal of an aliquot for carbon
analysis. A concentration of 227 pg C/litre was attained on the second
day, and no further increase was observed thereafter.
Earlier experiments on alternate bubbling and filtering were repeated with better controls and more thorough replication and with
essentially the same kind of results that had been obtained earlier. I n
a series of nine experiments the mean total yield for four successive oneday bubbling periods was 492 pg C/litre and this was 3.3 times the mean
yield for the first day, indicating that particle formation can continue
for some time with only slight diminution if the particles are removed
at intervals. Another experiment was devised to provide continuous
yield, and the amount obtained during one day was larger than in four
days of alternate bubbling and filtering. I n short, the results reaffirmed