12 BULLETIN 581, U. S. DEPARTMENT OF AGRICULTURE. 
Table 8 summarizes 39 days’ work under factory conditions on the 
relation, between the amount of rot and the micrescopical counts. 
As in Table 7, the results are arranged in the order of the amount of 
decayed material in the original stock. Under factory conditions it 
is frequently difficult to trace a definite portion of raw stock through 
the whole process of manufacture, with any certainty that the final 
product is from the same stock, unmixed with other lots. It has 
therefore been considered advisable to average the daily results 
obtained on the amount of rot and compare them with corresponding 
daily averages on the microscopic tests. It might be stated in this 
connection that the results thus summarized in ‘Table 8 represent 
experiments conducted in 17 different factories and involve 179 
determinations of percentage by weight of rot in stock and 235 mold 
counts. 
Experience has shown that in factories where good, thorough wash- 
ing is employed and where promptness of handling is observed the 
mold count is of greater importance in judging the condition of the 
raw stock than the counts on the other organisms. High yeast and 
spore, and bacterial counts usually indicate secondary spoilage. In 
order to secure as large a mass of data as possible on the mold con- 
tent, complete counts for the other organisms were made in ouly a 
few cases during the field work of 1916. Hence the blanks in the 
table. A comparison of the field results shows that, as in the case 
of the laboratory tests, high mold counts were not found in the 
products made from stock which had been well handled and sorted. 
The results given in Table 8 emphasize two points also brought 
out in Table 7: First, none of the samples having 1 per cent or less 
of rot give excessively high microscopical counts; second, with in- 
creasing amounts of rot more or less increase in one or another of the 
microorganisms occurs. They also show that while a comparatively 
low count may not always indicate a product made from suitable 
stock, it is clearly demonstrated that a high count on any one or 
more of the three types of organisms indicates bad stock. It appears, 
then, that one weak spot in the micro-counting system is its failure 
to disclose some bad products which should be condemned. 
To facilitate the study, the mold counts given in Table 8 have been 
plotted graphically in Figure 2, in which the “Zone of Possible Mold 
Counts’ is indicated by the shaded portion. Attention should be 
called to the fact that beyond 20 per cent of rot the chart is plotted 
on the basis of the assumption that 100 per cent rot would give a 
mold count of 100 per cent of the fields. As a matter of fact, the 
mold count reaches this maximum of 100 per cent of the fields with 
less than this amount of rot, as will be seen by reference to the 
laboratory series. These high counts, while occurring occasionally in 
commercial samples, arenowrare. In the studies of the relationship 
