QUANTITATIVE MICROSCOPIC DETERMINATIONS 159 



This must be avoided. On the other hand, it is advisable not to shift 

 the mount more than may be necessary to obtain an entirely new field 

 as above explained. The counts should be made in groups of ten 

 selected from five nearly equidistant areas of the mount, avoiding the 

 extreme margins of the area of the mount, and the average of the 

 fifty counts is taken. If the group counts (of ten each) show wide 

 numerical variation, it indicates that the material was not uniformly 

 distributed on the counting chamber. The dilutions from which the 

 counts are to be made should be such that not more than five to ten 

 counting elements will appear in one field of view. 



In making numerical counts of minute particles, such as starch 

 granules, trichomes, epidermal cells, sparingly present and rather 

 small sclerenchyma cells or short bast cells, for which purpose the 

 high power is usually employed, it is advisable to make the dilutions 

 such, that not more than from ten to twenty-five of the counting parti- 

 cles will appear in one field of view. In such cases it is the rule to 

 count one half of the particles which lie across the margin of the field of 

 view. If the tissue elements upon which the percentage count is to 

 be based are very sparingly present, from one to two to even less than 

 one per field of view, it is necessary to take the average of many counts, 

 100 fields or more, in order that an average may be obtained which will 

 yield a reliable percentage determination. In some cases coloring 

 agents such as phloroglucin and hydrochloric acid (lignin test), iodine 

 test solution (for starch), chloriodide of zinc, and others, will prove 

 useful.- 



If it is desired to make careful starch counts of drugs and spices, 

 it is necessary to use a higher degree of fineness. The particles must be 

 small enough to pass through a 200 mesh sieve or through bolting 

 cloth. In the case of vegetable substances having starch aggregates, 

 it is advised to count the aggregates as one, rather than to attempt 

 to give the number of individual granules in each aggregate, as, for 

 example, in making starch counts of oat, rice, aconite and buckwheat, 

 but care must be observed not to omit to state this in the recorded 

 results. 



In the above method it is not necessary to give any attention to 

 differences in specific gravity of the different tissue elements and cell 

 contents. Should it be desired to give the percentage of sand present 

 (by weight) in a given vegetable powder, it would be necessary to 

 ascertain by the trial method that mixture of the substance and sand 

 which would give the same counts as the article under examination. 

 Let us assume that by the trial method we found that the addition of 

 5 grams of sand to 5 grams of the powder we obtained the same sand 



