Document Z8vwjRm848xJeo24LYJ41X7QV

186 FRANK A. PaTTV extension being provided with a side connection or Y leading to the interferometer,;' The result obtained with the Willson apparatus was 101.6 p.p.m. carbon tetraf chloride. The results obtained with the interferometer, plotted in Figure 4, averj age 123 p.p.m., with a minimum of 6 p.p.m. and a maximum of 1096 p.p.m. durin the V2-hour sampling period. 'q More than one interpretation of this variation in results is possible. It may. b.- that we-have here a true measure of the actual difference between the mean of thi series of grab samples and the average obtained from an integrated continuou sample. At first glance one might say that the difference here of 21 p.p.m. between tl two methods is within the experimental error of the interferometer and assume tl 10 IS 20 TIME, MINUTES Figure 4. Carbon tetrachloride vapor in air as determined by means of a gas mterferosiif combustion result to be correct. However, although there is a probable error'-op p.p.m. for one reading on the interferometer, for a series of readings the result^ tained is much closer to the correct one, and probabilities are that the combust apparatus gave slightly low results,80 especially for the period of time that, the'con .-centration_was-abov.e-500-p.p.m._The-corr.ect-xesult_may._w.elLbe-assumedl.t;oiM^" been 5 to 15 .per cent, above the combustion results, or betweenil07 :and:117'lp1^ The graph is an interesting illustration of another point, that is, of the^ufil of trying to evaluate a variable exposure by a single, or even a few, random g: samples. To be reliable the grabjajrnples must be taken at frequent regularly spji intervals. These particular samples represent an abnormal variation in concent tion in that the spillage was frequent and variable, with constantly good ventilation. Under such conditions, if the chief concern is maximum level oflip ;? 00F. H. Goldman and C. G. Seegmiller, J. Ind. Hyg. Toxicol., 25, 181 (1943). SAMPLING AND ANALYSIS OP CONTAMINANTS 187 Srafeaaislffl1*, the industrial hygienist will choose properly spaced grab samples; if gp^gSffij^anore dependent upon cumulative exposures, he may prefer the inte^riV''^pmpie. B. ADSORPTION FOR EVALUATION BY WEIGHT ffilfe^Jd.'iny^vanations of the adsorption method introduced by Greenburg31 have by different investigators. The principle is the same in all and the i.fieri*`)inings are similar. The usual adsorbent is activated charcoal or silica gel ^f^'5i&,nvenient weighing tubes. These materials are not selective but most vapors and gases, especially those of high molecular weight, ^^ajf^^'ir'nple of vapor in air is drawn through the adsorbent at a constant rate, or ^ess Per m*nutei and after a suitable sample is collected the tube i ^contents is weighed. Since practice has shown that variations in weight of ^^df^grl" tubes, and even more so of charcoal tubes, often amount to more than j'tjs'.mg^vWit bin a 24-hour period as a result of static charges, moisture, and other ^iJtq^aJsiiV-^n though no air has been passed through them, it is obvious that def^e^&fe^sults require samples weighing at least 50 mg. If the sample is collected Jr^ffA'iie-r'atepf,! liter per minute, and it is desired to determine the concentration of S{MTMbi.%.,air of a room where the average concentration amounts to 50 p.p.m. ^ is evident that a sampling time of about 5 hours would be ^^^^m^dependable results. For higher concentrations this method is more ^^^ffi|since less time is required to obtain a sample of suitable weight. Tubes ^^^^^tj|;tfr several hour's in air of the same humidity and temperature as the """fejo be analysed<tnay -be'useH witKout'filters for removing water vapor l^|ai|b|dioxide from th& air sampled. However, it is the usual practice in the ^nfc^t'gaccuracy to employ such filters wherever they do not alter the concen- i^iocJm|the contaminant. Moskowitz and Burke82 found it advantageous to have T^^bcft^and after the adsorbent. The ones ahead of .the adsorbent removed ^tt^^^^carbon dioxide from the air, and the ones following the adsorbent col- gj||iffi||&and carbon dioxide lost by the adsorbent 'as. the air passed over it. fjipn method obviously collects an integrated^sample, while it is fre- ||[Jsii;able to follow the changes in concentration in-order to obtain infor||bhe height and duration of peak values. The-method is applicable to Bl%rs f high molecular weight and moderate,t,qxicity., It.must always ^^^tealsthat-many'vapors-cannot be filtered" through: drying agents and J^s|itfeout partial or even complete loss within those,,agents. Since the an accuracy only to the nearest milligram, weighing usually can ISmSlished at any laboratory or in the field. Vapor sampled adsorbed on ac- ^^^Imor-silicai.gehmay-also be transported-to-the-iabor-ator-yj'released bv jffl||<j.*ir through the adsorbent, and the vapor determined by suitable Mm* i analysis (see also pages 185 and 186). JlllnWg, U. S. Pub. Health Repts., 41, 1516 (1926); W. A. Cook and'A. L. .CplePp. |ff. Toxicol., 18, 194 (1936). 5 -.bwitz and W. J. Burke, N. Y. State Ind. Bull., 17, 168 (1938). , wli.