Document DDwQK1VVLkw3RmDebegYrQ1zQ

200 FRANZ A. PATTY 2. Counting Konimeter Samples Konimeter samples may be examined and counted by use of the special mic|| scope attached to some konimeters. More satisfactory counts are obtainedlT using a regular miscroscope with light-field illumination and fitted with a sp$cl holder so that the konimeter dust spots can be rotated and any one can be broui into view readily. A satisfactory magnification has been found to be 200a: obtamef by using a 10a; objective with a compensated 20x ocular. The ocular is fitted !TM a squared micrometer scale, similar to that shown in Figure 7, with two croj Figure 7. Micrometer scale for counting konimeter samples! lines that outline sectors of 18 degrees. The parallel lines ai and id are for. ty estimating the size of particles. In counting, the scale is centered over thef spot and all the particles in the two 18-degree sectors are counted, giving. aH! count of'all the dust in 36 degrees or 10 per cent of the entire spot. Fro count must be subtracted a control count of an equal area of the unused plate] adhesive film. The net count of 10 per cent of the total area of the dust.!* multiplied by 10 and divided by the volume of air sampled (in milliliteri obtain the number of particles per milliliter. If the results are desired in . ml particles per cubic foot, the number of particles per milliliter may be rau|| bemultiplied by,6.057.' a 5-m.l. sample, the 36-degree sector couni^ S. Counting Owen's Jet Samples The cover slip with dust deposit is mounted dry with dust side downl microscope slide and examined under a microscope with an oil immersio'ntffi tive, dark-field illumination, and about 1000s magnification. Ah ocular||p| crossbar ruled micrometer with 1-mm. squares is used; a representative fca jHDf 4 -ft SAMPLING AND ANALYSIS OF CONTAMINANTS 201 fuarlllfidth across the ribbon of dust is selected, and all the particles in this Sited. Two or more bands may be counted and averaged if- desired, oiffrhardly justifiable because the ribbon is rarely uniform in distribu^Sj^iilthroughoutits length, especially at each end. The number of bands ffg$f|of sample ribbon must then be determined if this factor has not ousrofl^ehrestablished. Lower magnification must be used to measure the Mlof the ruled ocular disk and the relation between objectives, if not <sdefestablished by the use of a stage micrometer. Since the volume of ^^Eas 50 ml;, the number of particles per milliliter is computed by qJ^M^number found in one band by the number of bands and dividing may easily be computed to million particles per cubic foot by 0.Q28, but no correlation with standard impinger results should such computation. ` SjtetB. DETERMINING PARTICLE SIZE DISTRIBUTION mW' _ ^^Sl|microprojectorB4'Be is perhaps the most satisfactory method of de- m^ra||particle-size distribution of dust samples collected either by the S^^o^gbr more drops evaporated to dryness on a slide), the konimeter, the |||j||the electric precipitator. While a magnification of 1000 diameters Sl!Ox''objective) is ordinarily used for counting, a magnification of n|ejers (90s oil-immersion objective) is recommended for determining ^Wf'^a change in tube length being necessary to adjust the magnifica- yze-distribution-is-determined-by-estimating-the-size-of each of the |||pFesentative fields of the sample using the 0.5-jt squares on the gmfarison. The particles are classified according to diameter in inter- lip;. fi, counts for each class being registered on a Marbel blood |6ther suitable means. The number of particles of a particular size |given count determines that size frequency. These frequencies are Hr percentages of the total count. ' a particle-size determination may be presented graphically, as p|pp#ing the counts as ordinates against the sizes as abscissas at cftsize intervals. This graph clearly illustrates the size variation Show's two typical frequency distributions. ig^i|e*datar-a:r6tthbulatedf'-as:ih--TableI45-which-has-size-g-roups-at !|nas'ffetjuencies in percentage, it is again possible to compare the ||of%ifferent dust dispersions. The median is the cumulative 50 per p^ailepresentative measure of average dust size. ItJskdfildjbe noted, ~ ^ttfiaggASfjmmejiusts .with..wide,yariatipns in size the mode. may i&teSn'KferrfTB pipe' with the median, and would be a better measure of central JgjQin dust, soapstone dust, and slate milling, in all of which the `mbst frequent occurrence, is 1.25 p, although the medians range 8U^mwn'and W. P. Yant, V. S. Bur. Mines Rept. Invest. No. 3289, 1935. i ' * , f(. 1