View report page

(5) Possibly the most important part of the procedure is to ensure that the aliquot spread out on
the disc is really representative of the whole of the suspension. The writer personally has found
difficulty in securing this by the method prescribed in the A.O.A.C., which is to take a portion
of the suspension on a knife blade or scalpel and spread evenly over the disc to give uniform
distribution; the difficulty lies in removing a small volume of a suspension containing solid
fibrous matter and an aqueous liquid on a scalpel, and again in transferring a portion of this to a
slide, without permitting any separation of liquid from the solid. The use of a dissecting needle
to remove the sample portion from the scalpel to the mount has been suggested, but a simple and
efficient method is to use a straight sided tube of 3-3.5mm. bore and, removing the finger from the
top momentarily, allow a large drop to fall on to the disc; this drop carries down the cellular
matter without separation of liquid, and it should then be spread evenly over the disc with a
needle before the coverslip is placed in position.
Interpretation of the Results
The Trade Specification for Tomato Puree referred to above lays down a Sampling Procedure
and a statistical method for Treatment of Results. A minimum of 10 cans is suggested from a
consignment of 100 tons and the number of fields to be counted on each sample is 50. The True
Mean Value for the consignment, with a 95% probability, is then calculated from the arithmetic
mean and the standard error.
In the writer's experience a count based upon only 50 fields may be very different from the
mean of say 500 fields, and whilst the method of the Trade Specification is designed to provide
the most reliable and economical means of ascertaining the Howard Mould Count of a sizeable
consignment of puree, it cannot usually be adapted to the practice of a Public Analyst,
Under the Food and Drugs Act the sale of any one article of unsound or defective quality
constitutes a presumptive offence, and a Public Analyst may be required to issue a certificate on
one small sample of puree. Generally speaking samples are taken informally in the first instance,
an adverse report being followed up by a formal sample. Hence, before legal action was taken,
more than one sample would have been examined, and in the case of 2-3 oz. cans, perhaps six
might have been purchased and mixed together for the formal sample.
In the case of a sample taken under the Imported Food Regulations being defective, a few
more samples would normally be taken to see if the whole consignment is similar. If these
showed significant differences, further samples would be drawn, but under present circumstances
and within the limits of the financial and laboratory facilities available it would not be practicable
for Public Health Authorities to carry out the regular sampling of consignments upon the
scale recommended in the Trade Specification.
There is however nothing to stop the importer taking as many samples as he wishes. and in
the event of a disagreement between his analyst and the Port analyst a formal sample could be
drawn in the presence of his representatives, mixed and divided into three or four parts to enable
independent tests to be made if necessary. Post-sampling mould growth can be prevented by
addition of formalin and refrigeration.
The Public Analyst has therefore to take greater care to ensure the correctness of his count
than is envisaged in the Trade method. This he can do by counting more than 50 fields and thus
reducing the standard error of his mean. He will not usually be very concerned to determine the
exact count when it is low, and in view of the number of other tests to be applied to manufactured
tomato products, it may be suggested that for routine laboratory purposes one count, i.e. 25 fields,
might be sufficient if the result is below 30% positive; 50 fields if above 30 and below 40%
positive, 100 fields if above 40 and below 50% positive; but that if the count is 50% positive, or
whatever the currently accepted limit may be, or over, counting should be continued until further
counts of 25 fields do not significantly alter the mean of proceding counts. At least two suspensions
of the puree should be made, by different operators, and it will be usually found sufficient
if each operator counts one hundred fields on each suspension making 400 fields in all.
Some purees yield very variable counts, however, and if such a sample is encountered still more
counts may be necessary; occasionally as many as 800 fields.
The above suggestions are based solely upon experience obtained in this laboratory; they
are not part of the official procedure and it may well be that in other laboratories experience may
indicate a preference for some ouner procedure or method of interpretation.
Significance of the Howard Mould Count
No attempt is made when counting to differentiate between the types of mould that may be
present. Unless characteristic sporulation can be found this would indeed be impossible, because
having been sterilised no cultures can be grown. The only value of the count is as a guide to
the extent of fungal rot in the tomatoes used.
Several investigators have studied the relationship between the amount of rot in the raw
tomatoes and the Mould Count of the puree, and Howard's original data are of interest.7 His
. averaged results from 179 determinations of visible cut-out rot and 235 mould counts from 17
factories, summarised by the writer, are shown in Table III.
29