PARTICULATE ORGANIC MATTER I N SEA WATER
47
get significant formation of organic carbon in experiments that were
effectively sterilized by filtration through 0.22 p filters.
An evaluation of some of the major problems was published by
Batoosingh et al. (1969). A prerequisite for such work was of course
the development of an air supply that was reasonably free from
contamination and, incidentally, a comparison of this air with the
ordinary supply of laboratory compressed air as an aid in evaluating
previous results. Combustion of the air seemed an unnecessary refinement. Instead, a small volume of air (about 0.5 litre) was recirculated.
One method was simply to carry out the experiment in a tightly
stoppered pyrex bottle with magnetic stirring sufficiently vigorous to
introduce subsurface bubbles. The other was to introduce a bubbling
stone and circulate the air via silicone tubing and an AC vibratory
pump. The latter provided more efficient mixing with little likelihood
of serious contamination.
Results showed that the laboratory air supply in current use was
not objectionable, so that previous experiments such as the one described above on yields obtained with Nova Scotia coastal water were
not invalidated. The use of a polyethylene bottle instead of a glass
one did not introduce a significant error.
A major factor in determining the yield was the pore size of the
filters. Highly significant yields were obtained by bubbling sea water
that had been prefiltered through silver or Millipore filters of 1.2 p pore
size or glass fibre. The yield declined t o a very slightly significant level
when 0.22 p filters were used for pre-filtration. The results implied that
particles in the size range of 0.22-1 -2 p, and possibly larger, are important as nuclei in the formation of larger particles.
Further tests compared triple filtration through 0.8 p filters, as
practised by Menzel (1966), with single filtration, and the latter gave
higher yields, indicating that part of his difficulty in producing particles
was due simply t o the fact that triple filtration, even through filters of
medium pore size, removes a considerable fraction of the small particles
that facilitate aggregation. However, Batoosingh et al. (1969) found
that even after triple filtration the yield was not entirely insignificant.
They used a 24-hour bubbling period, as contrasted with 3-6 hours in
Menzel’s experiments, and the bottle used as a bubbling chamber had
a larger surface : volume ratio than his cylinder. These factors may have
been responsible for the difference.
Barber’s (1966) conclusion that bacteria are necessary for the aggregation process was not confirmed, but it was not categorically denied.
Batoosingh et al. (1969) found that samples filtred through 0.22 p
filters were sterile, as judged by plating test, yet there was a slight
47
get significant formation of organic carbon in experiments that were
effectively sterilized by filtration through 0.22 p filters.
An evaluation of some of the major problems was published by
Batoosingh et al. (1969). A prerequisite for such work was of course
the development of an air supply that was reasonably free from
contamination and, incidentally, a comparison of this air with the
ordinary supply of laboratory compressed air as an aid in evaluating
previous results. Combustion of the air seemed an unnecessary refinement. Instead, a small volume of air (about 0.5 litre) was recirculated.
One method was simply to carry out the experiment in a tightly
stoppered pyrex bottle with magnetic stirring sufficiently vigorous to
introduce subsurface bubbles. The other was to introduce a bubbling
stone and circulate the air via silicone tubing and an AC vibratory
pump. The latter provided more efficient mixing with little likelihood
of serious contamination.
Results showed that the laboratory air supply in current use was
not objectionable, so that previous experiments such as the one described above on yields obtained with Nova Scotia coastal water were
not invalidated. The use of a polyethylene bottle instead of a glass
one did not introduce a significant error.
A major factor in determining the yield was the pore size of the
filters. Highly significant yields were obtained by bubbling sea water
that had been prefiltered through silver or Millipore filters of 1.2 p pore
size or glass fibre. The yield declined t o a very slightly significant level
when 0.22 p filters were used for pre-filtration. The results implied that
particles in the size range of 0.22-1 -2 p, and possibly larger, are important as nuclei in the formation of larger particles.
Further tests compared triple filtration through 0.8 p filters, as
practised by Menzel (1966), with single filtration, and the latter gave
higher yields, indicating that part of his difficulty in producing particles
was due simply t o the fact that triple filtration, even through filters of
medium pore size, removes a considerable fraction of the small particles
that facilitate aggregation. However, Batoosingh et al. (1969) found
that even after triple filtration the yield was not entirely insignificant.
They used a 24-hour bubbling period, as contrasted with 3-6 hours in
Menzel’s experiments, and the bottle used as a bubbling chamber had
a larger surface : volume ratio than his cylinder. These factors may have
been responsible for the difference.
Barber’s (1966) conclusion that bacteria are necessary for the aggregation process was not confirmed, but it was not categorically denied.
Batoosingh et al. (1969) found that samples filtred through 0.22 p
filters were sterile, as judged by plating test, yet there was a slight