View report page

Unfortunately different types of mould have different effects on the mould count. This has
been well shown by Eisenberg, who determined the mould counts of purees made from tomatoes
affected by specific moulds.8 Reference should be made to his paper for fuller information, but
briefly it may be inferred that although the averages of his counts accorded with Howard's data
individual samples showed widely scattered results, sometimes exceeding ± 50% from the mean.
Also of interest in this paper is the relative area of the rot; a cut-out rot of 10% by weight, with
a mould count of about 68% positive, corresponding roughly to an area of visible rot of 1½ inches
diameter in an average size tomato; but again, variations due to the kind of rot were very great.
One of the difficulties in relating the mould count to the percentage of rot is that some
portions of rot are soft and completely broken up in the extraction but others are relatively hard
and may be discharged with the cores and skins. Hence the presence of high proportion of visible
rot can result in a low mould count.
After thorough studies by the U.S.A. Department of Agriculture it was concluded that in
properly sorted stock the percentage of decay should not exceed one per cent.9
"Whatever the deficiencies of the Howard technique as a means of providing a true indication
of the state of the fruit and manufacturing conditions, it is the best available for indicating contamination
by mould and all that this implies" (J.C. Dakin10). The general inference, as originally
expressed by Howard and Stephenson7, appears to be universally accepted, — "A low mould
count does not necessarily indicate sound stock, but a high mould count always indicates bad
stock or improper handling." This is the real value of the test to the Public Analyst.
In addition to the variations between the kind of rot and the mould count, criticism has been
made of the accuracy of the technique itself. Only small quantities of sample are taken for each
count and as emphasized in the Trade Specification "it must be recognised that any mould present
in tomato puree is very unlikely to be uniformly distributed throughout the consignment, or even
throughout a given sample". H.R. Smith9 states that it is quite possible for properly qualified
analysts to obtain satisfactory checks on the same sample and quotes a trial in which eight
analysts representing different organisations found no significant difference between their counts;
but a recent collaborative trial between eight analysts in independent laboratories in this country
showed very wide differences. The procedure in the latter trial, however, specified 50 counts,
and as already indicated it is the writer's opinion that far more counts are necessary to obtain a
reliable figure.
A further criticism of the Howard method is that, since it is based on the recognition of mould
filaments, rots which are caused by bacteria, yeasts, viruses and physiological causes unaccompanied
by mould are not detected8. All evidence of spoilage should, however, be taken into
account in considering whether a product is sound or unsound; and a Mould Count of less than
50% positive if accompanied by excessive bacterial growth might afford evidence for condemnation.
The taste, preferably after dilution in the case of a puree, may also be of evidential value;
but not always, because the degree of mustiness depends upon the type of infection, a high
mould count sometimes being found when the flavour is normal and vice versa.
Furthermore, the mould count does not differentiate between bad stock and improper handling.
Conditions in the factory may be conducive to deterioration even after the tomatoes have been
sorted, cleaned and any visible rot cut out. Any accumulation of tomato slime, inadequate cleansing
of equipment or delay before sterilisation might cause a high mould count.
Although the final product might seem equally objectionable to a consumer whatever the
origin of the mould, it would be useful for the purposes of the Food Standards (Tomato Ketchup)
Order, 1949, to be able to ascertain if a sample of Ketchup has in fact been "made from clean
and wholesome tomatoes, or from the tomato puree made from clean and wholesome tomatoes." In
this connection the "Rot Fragment Count" may be of use.
30