6 BULLETIN 569, U. S. DEPARTMENT OF AGRICULTURE. 



men the high count was immediately lowered and remained low dur- 

 ing the rest of the visit. 



Tests of the ketchup manufactured by this factory were made 

 periodically during the stay and were distributed over 8 days. 

 The first four tests were made prior to the general cleanup. The 

 sample used in the fifth test was taken from the first batch of ketchup 

 made after the cleaning was completed. About one-third of the 

 molds in this sample were distinctly of the form which indicates 

 dirty apparatus. They were the filaments dislodged during the 

 cleaning process which had not been fully rinsed out during the wash- 

 ing and hence were carried over in the first kettleful of the product. 

 The last four tests were made during the next three days. 



Table 1. — Microscopical examination of tomato ketchup before and after the 

 installation of adequate cleaning equipment and methods. 



Test No. 



Per cent of 



fields 

 with molds. 



Bacteria; 

 million per 

 cubic centi- 

 meter. 



Yeasts and 

 spores per 

 1/60 cubic 



millimeter. 



1 



45 

 66 

 64 

 54 



40 

 60 

 52 

 43 



65 



2. . 



57 



3 



52 



4... 



56 









57.2 



48.7 



57.5 







5 



30 

 16 

 21 

 15 

 20 



9 



10 

 10 

 8 

 7 



21 



6 



15 



7 



10 



8 



7 



9 



13 









18 



8.7 



11.2 







A study of Table 1 emphasizes the change in the counts between 

 those made before and those made after cleaning, in that the counts 

 have been reduced in the case of mold to about one-third and of 

 bacteria and of yeasts and spores to about one-fifth. 



The question may arise in the minds of some readers as to whether 

 the marked change shown in the counts might not have been the 

 result of other influences than the cleaning. The only other change 

 made was a minor one in the system of sorting. In Table 2 are 

 given the results of efficiency-of-sorting tests. The method of making 

 these tests is described in detail on page 18. The tests were made 

 on the sorted and washed stock ready for the crusher and represent 

 the approximate percentage of decay that was going into the finished 

 product. The data in Table 2 show that the average percentage of 

 decay during the first 5 days was 1 per cent, while during the last 

 3 days (the after-cleaning period) it was 0.94 per cent. Obviously 

 these figures are negligible in accounting for the marked changes 

 observed in the counts. 



