DIRECT EXAMINATION 



45 



means of the low power and make counts under the suitable high 

 powers. From ten to twenty of the ruled areas should be counted 

 and from these countings figure the average. It is desirable to 

 make two or three mounts of each sample, thus giving the average 

 of from twenty to thirty areas counted. The countings are to 

 be made in areas free from pulp 

 fragments and including all organ- 

 isms lying within the ruled bound- 

 ing lines and inclusive of half 

 averages of those organisms which 

 lie across the rulings. All count- 

 ings which present characters of 

 doubt are omitted from the final 

 estimates. 



Those organisms which occur 

 within the cell-lumen of the vege- 

 table tissues are not to be counted. 

 To do so is practicably impossible 

 and such countings, even if pos- 

 sible, would add nothing to the 

 value of the findings. In case the 



cells contain numerous bacteria this should be noted in the report, 

 as it certainly indicates decomposition of the material. The prin- 

 cipal decomposition changes due to the invasion of bacteria and 

 other organisms are, however, largely limited to the exterior of 

 cells, especially by those organisms which develop during or after 

 the factory processing. The numerical determinations are there- 

 fore limited to organisms which occur in the matrix and those which 

 have been washed from the exterior of cells by the thorough mixing. 

 The thorough mixing of the samples is a very important part of 

 the procedure. In the case of Kquids and semiliquids, mixing is 

 done by thorough shaking, and in the case of pastes and similar 

 materials, by means of a spatula or a small spoon. 



In making counts of very small or comparatively short bacilli. 



Fig. 9. — Biirker ruling, useful in 

 making counts of milk fat globules, 

 spores, and yeast cells. The average 

 of many counts is taken. — {Carl Zeiss.) 



