Document pmbZQwdGNbvN198Zg9vME6EjX

A Correlation between Silicosis and the Dust Exposure of South African Gold Miners Paper--" Aspects of Dust Records, Dust Exposures and Silicosis " At this stage of the proceedings, the President introduced Mr. R. S. J. du Toit, and said that before calling on Mr. du Toit to present his paper, he would ask Mr. Kitto to take the Chair, as he wished to make a technical contribution to the discussion. Mr. Kitto then took the Chair, and Mr. du Toit presented his paper. There were several contributions to the discussion. Office-bearers and Members of Council for the Year 1959/1960 The President thanked the scrutineers for the work they had done, and announced the result of the ballot as follows:-- President: P. H. Kitto. Vice-Presidents: J. H. Quilliam and D. F. H. Grave. Members of Council: P. F. Cilliers, R. S. J. du Toit, J. A. Northcott, N. Paige-Pagani, J. P. Rees, L. A. J. Slabbert, F. C. Startup and P. J. Vermunt. The retiring President congratulated Mr. Kitto, the newly elected President, and Mr. Quilliam and Mr. Grave, the two newly elected Vice-Presidents. He also congratu lated the newly elected Council Members and formally handed over the affairs of the Society to the new President. The President: Mr. Beadle, Ladies and Gentlemen, I would like to thank you very much for the honour you have bestowed upon me. I am fully aware of the many achievements of the past and the leaders who have been Presidents of this Society who have worked hard to accomplish this. 1 would like to pay tribute to our retiring President, Mr. Beadle. He has worked harder than any other member of the Society; not only has he worked willingly, but he does it because he enjoys it. If I did not have the support of the many other Council Members and the Vice-Presidents, I would not be able to carry on, and I am sure that Mr. Quilliam and Mr. Grave will help me and I am very pleased to see them in office. I am also very pleased that we have new blood on the Council, and I am quite sure that they will work hard to carry on the work of the Society. Mr. Quilliam: On behalf of the newly elected Members and Mr. Grave, I would like to thank the members for conferring this honour upon us. I am sure that Mr. Grave and I will give our President our full support to maintain the achieved high standards of the Society. The President said that before closing the meeting, he would like to invite those pre sent to join him in a cup of coffee which would be served in the foyer. He asked the newly elected Council Members to remain with him in the hall for a brief discussion. The meeting terminated at 11.40 a.m. A CORRELATION BETWEEN SILICOSIS* AND THE DUST EXPOSURE OF SOUTH AFRICAN GOLD MINERS By R. S. J. du Toitf Mr. President, ladies and gentlemen, It is indeed an honour to have been asked to read a paper to you today and I can assure you that it is a great pleasure to do * I uss the specific term silicosis rather than the general term pneumoconiosis because the dust dealt with is essentially silica dust. t Technical Research Officer, Silicosis Research Section Unit of S.A. Mines Dept. so. The paper is based on a dissertation which I submitted to the University of the Witwatersrand. Before commencing, I have a list of acknowledgements to make. They are: The Government Mining Engineer for permission to undertake the investigation and to publish this paper; The Transvaal and Orange Free State 162 Journal of the Mine Ventilation Society of South Africa, July, 1959 A Correlation between Silicosis and the Dust Exposure of South African Gold Miners Chamber of Mines for permission to use certain information from records of service of employees; The Pneumoconiosis Compensation Com missioner for permission to study the records of service of mine workers available at the Board; The Director of the Pneumoconiosis Bureau for permission to select suitable cases and to use the medical findings of the Bureau in connection with the cases selected; The Superintendent of the Routine Division of the South African Institute for Medical Research for permission to have access to the post-mortem reports on the lungs of mine workers by the Pathologists of the Pneumoconiosis Unit; Dr. Ian Webster, Senior Pathologist of the Pneumoconiosis Unit, for extracts of post-mortem reports on the lungs of mine workers; The Director of the Pneumoconiosis Research Unit of the Council for Scientific and Industrial Research for permission to have certain computations performed by a member of his staff. I also wish to acknowledge advice received from: Mr. J. P. Rees, formerly Director of the Dust and Ventilation Laboratories of the Transvaal and Orange Free State Chamber of Mines (now retired); Mr. T. L. Gibbs, Deputy Government Mining Engineer; Professor R. A. L. Black, Professor J. E. Kerrich, Mr. T. E. Edwards, Dr. H. S. Sichel, Mr. D. G. Beadle and Mr. J. de V. Lambrechts. The author is also indebted to Mr. R. J. Pieterse for the diagrams. SYNOPSIS Calendar year trends in the dustiness per shift were prepared for about a dozen dif ferent classes of mine workers. The konimeter dust sampling records of the Govern ment Department of Mines over the 24-year period 1931 to 1955, were used in the con struction of these curves. The work histories of 57 deceased Euro pean miners were analysed and by means of the above curves their total dust exposures were estimated. Degree of silicosis was determined post mortem, and statistical analyses show that although there is considerable scatter, there is nevertheless a significant correlation be tween degree of silicosis and total dust exposure. CONTENTS 1. Dust Dosage. 1.1 Introduction. 1.2 Medical aspect. 1.3 The dust results used. 1.4 The dustiness of mine air at strategic points. 1.5 The dustiness of an underground shift. 1.6 The classification of mine workers into main groups. 1.7 Dust exposures. 2. The Sample of Mine Workers. 3. Medical Classifications. 4. Results. 5. Statistical Analyses. 6. Concluding Remarks. 6.1 (a) The instrument. (b) The method of treatment. 6.2 Application. 6.3 Records of service. 6.4 Recorded dust counts. Part I: Dust Dosage 1.1 Introduction First thoughts about the present investiga tion was had by about 1951, as a result of the views expressed by Beadle 1 and others concerning the performance and limitations of the konimeter. Then I studied the work by Patterson,2 Lambrechts,3 4 5 Rabson,6 Davies, Aylward and Leacy,7 Hasenclever,8 and Roeber.9 Further stimulus was had from the articles by Roach,1011 Wynn,12 13 Wright14 and Hatch.16 Finally, the following three factors really made me decide that it would be worth while to undertake the investigation: Firstly, there was uncertainty regarding the useful ness of konimeter results to indicate the silicosis hazard in mines; secondly, there were the konimeter results which were pub lished since as long ago as 1931--something no other country in the world possessed; thirdly, there were the records of service kept of all European mine workers since as Journal of the Mine Ventilation Society of South Africa, July, 1959 163 A Correlation between Silicosis and the Dust Exposure of South African Gold Miners long ago as 1916, another item not to be had anywhere else in the world. The Problem: The problem was to prove, if possible, whether konimeter counts were related to the silicosis hazard due to dust in mine air, and if so, whether the relation was sufficiently close to warrant the continued use of the konimeter as the dust sampling instrument on South African mines. It was argued that the problem would be solved if the dust exposure of a mine worker was a measure of his degree of silicosis, where dust exposure is the product of time and dust concentration as measured by konimeter. Unfortunately, the actual dust exposures of mine workers are not available. All we can do, therefore, is to make the best of a bad job and to estimate, as best we can, their dust exposures. As a result of inaccuracies introduced by these estimates, also, as a result of the vary ing susceptibility of individuals to contract silicosis and furthermore, as a result of the fact that only broad classifications of the degree of silicosis in persons are possible, there will be a tendency to obscure any functional relationship between the two measures. Before dealing with the actual investigation I wish to say a few words about some of the general aspects of the dust problem. 1.2 Medical Aspects Dust sampling is intimately associated with the medical aspects of the dust problem. Of special interest is the anatomy of the respiratory system,16 but space does not permit me to expand on this aspect. The range of harmful particles: Of im portance also is the range of harmful particles. It is generally accepted that particles larger than about 5 micron do not represent a hazard because they would settle out before reaching the alveoli. The relative importance of ultra micro scopic particles, i.e., of less than about i micron, is not such a simple matter and we have to admit that to date there is no evidence that such particles are harmless. On the other hand there is the question of: What is the best single descriptive para meter to measure ? One of the recommendations of the recent Pneumoconiosis Conference 17 covered this point and the recommendation reads: "that, in the light of the present available evidence, the best single descriptive para meter to measure, be considered the surface area of the respirable dust in the case of quartz dust." Although coming after I had completed the investigation it is of interest to consider how the konimeter count compares with the surface area of airborne dust. Lambrechts,3 Rabson,0 and Beadle 1 showed that the koni meter count was high for coarse dust and low for fine dust. Therefore, the konimeter tends to allow for the larger surface area of a coarse dust as compared with a fine dust for the same number of particles. It is of course doubtful whether the allowance is the correct amount. Be that as it may, the essence of the matter is whether the dust results can be correlated with medical evidence. 1.3 The dust results used This brings me to the results I used. There was the question of whether to use the results recorded by the Department of Mines, or whether to use those recorded by the Chamber of Mines, or whether to com bine the two sets. I decided to use the results of the Department of Mines 18 only (as published in the Annual Reports of the Government Mining Engineer) in spite of the fact that the results of the Chamber are based on a much greater number of dust samples than those of the Department of Mines. The main reasons for this decision are as follows: (1) all the mines did not always submit returns to the Chamber; e.g., in 1936, returns were submitted by only 24out of the 35 working mines 19; (2) the Chamber of Mines stopped pub lishing results of dust determinations by 1950; (3) the different mines did not always use the same treatment technique 19; (4) only by 1950 did the Chamber change over from immersion-ignition to ignition-immersion-ignition treat ment whereas the Department of Mines changed to the latter method as early as 1942; (5) the konimeter results published by 164 Journal of the Mine Ventilation Society of South Africa, July, 1959 A Correlation between Silicosis and the Dust Exposure of South African Gold Miners the Chamber were not cast in the same classifications as those pub lished by the Department of Mines which would make a combination of the two sets of data very difficult; (6) the classification made by the Cham ber was not as suitable as the one made by the Department for the in vestigation undertaken; (7) finally, it was doubtful whether any materially improved accuracy would result from a combination of the two sets of results. Justification for using the dust counts recorded during routine inspections: A question that arose was whether the obser vations made by Inspectors of Mines could justifiably be used to represent conditions over the whole shift. The reason for such a question is the fact that the majority of dust samples taken by Inspectors of Mines are taken between about 9 a.m. and 12 noon, whereas the underground shift of most occupations extends for several hours on either side of this period and for some occu pations over the full 24 hours of the day. The answer to the question lies in the fact that all the mines concerned employ Venti lation Officers20 who keep an eye on the dust conditions in mine workings and in the fact that inspections by Government In spectors of Mines may be made, and dust samples taken, whenever there is reason to believe that satisfactory standards are not maintained. (Complaints21 to this effect may be made by any employee.) Another point to consider was:-- The degree of bias of the recorded dust counts: The results as published by the De partment of Mines are not completely free from bias22 in the statistical sense.* The reason for the bias is that one of the main objects of taking dust samples is to " smell " out places with poor conditions, and having found them, to keep on taking samples there until conditions are satisfactory. * This criticism, incidentally, applies also to the dust sampling results of most other concerns. Hence, if there are two places to choose from, an Inspector, like any other person who has to see that dust production is kept down to the minimum, will be inclined to go to the place where poorer conditions are suspected or known to exist. On the other hand, the Inspector of Mines has not always got such a choice, but mostly has to take dust samples in places where his routine inspections lead him to. On the whole, therefore, it was felt that the vast majority of dust spots could be considered as random samples so that the results recorded at different classes of underground places, over the same period, could be com pared one with the other and similarly, the results recorded over different periods, at the same class of place, could be compared one with the other provided the dust assess ment technique did not alter, or if altered, was allowed for. Hence, I was satisfied that there was justification for using the results recorded by Government Inspectors of Mines. 1.4 The dustiness of mine air at a number of strategic points To simplify matters, it was decided to use the results recorded at only six of the under ground points for which the Government Mining Engineer publish results. The six strategic points chosen are: (1) Air entering downcast shafts at sur face. (2) Air leaving downcast shafts before passing tipping stations. (3) Air in main intake airways away from downcast shafts. (4) Air in stopes. (5) Air at development end faces. (6) Air at the bottom of sinking shafts. The variation in the dust count of mine air with calendar years: It was necessary to estimate the variation of the dust count with calendar years, and to do this the effects of the various techniques introduced over the years had to be allowed for. Estimates of the effects were obtained by comparing the results for successive periods during which the dust determination technique remained unaltered. A significant difference in the count between the end of one period and the beginning of another was assumed to have been brought about by the change in the technique. The three periods of steady conditions: There are three periods during each of which the dust assessment technique remained practically unchanged. They are: (1) 1931-1941.* (2) 1942 24 to 1948. (3) 1949 25 and onward. Journal of the Mine Ventilation Society of South Africa, July, J959 165 A Correlation between Silicosis and the Dust Exposure of South African Gold Miners * An altered classification was introduced in 1940 23 which consisted of a distinction being made between dust samples taken in intake, working point, and return air, of stopes, development ends, and sinking shafts, and only two years later, viz., in 1942 was the ignition-immersion-ignition technique in troduced, so that in the case of stopes, development ends, and sinking shafts, a difference between the first and the second period was due to the combined effect of these two changes. An example of the method of dealing with one of the strategic points is depicted in Fig. 1, and a summary of the effects established of the various changes made in dust sampling technique, is given in Table I. (1) Top of downcast shafts; (3) Bottom of down cast shafts; (6) Main intake airways; (7) Stope faces; (8) Development end faces; (9) Shaft bottoms. Fig. 1. Dust Count in Stopes. () Annual mean counts. (a) Best estimate (present technique). (h) Best fit to 1931-39 mean counts. The trends in dust concentration, from 1931 to 1955, for each of the six strategic points are depicted in Fig. 2. 1.5 The dustiness of an underground shift Having got estimates for the dustiness of mine air at strategic points, the next problem was to arrive at estimates for the dustiness of the shift of underground workers. The dustiness of an underground shift depends essentially on the position or posi tions in the mine where the shift is spent, and the shift dustiness of underground workers can be estimated from the propor tions of the shift spent at or between these strategic dust levels. If the proportions of the shift were known, the mean dust level per shift could be computed. The actual proportions are, unfortunately, not known and estimates thereof have to be relied on. Effect of errors in the proportions used: To estimate the mean dust level per shift of an occupation from the proportions of the shift spent at various dust levels, is the same as using loading factors to arrive at weighted mean values. In such cases, the accuracy of the factors are not of critical importance because relatively large errors in the loading factors affect the weighted mean value to a relatively small extent. E.g., a 50 per cent, error in the loading factors of two quantities which differ as 2 is to 3, results in an error of only 10 per cent, in the weighted mean value. To simplify the computation, mine workers were classified into a few main groups for each of which the shift dustiness could be estimated for any particular calendar period. 1.6 The classification of mine workers into main groups Officials and non-officials were considered separately. The non-officials were classified according to the position in the mine where the main portion of the shift is spent and on this basis, the following main classifica tions were made: (1) Those employed essentially in and around shafts, referred to as SHAFT WORKERS; (2) those employed essentially along travellingways, referred to as TRA VELLINGWAY WORKERS; (3) those employed essentially along entrances to and exits from working 166 Journal of the Mine Ventilation Society of South Africa, July, 1959 A Correlation between Silicosis and the Dust Exposure of South African Gold Miners TABLE I Summary of Allowances made for the Changes in Technique Classification Separation of intake, working point and return counts--(1940) Changes Made Acid Vapour to ig.-im.-ig. (1943) Intake air at the top of downcast shafts (1) .. Intake air at the bottom of downcast shafts (3) Air in main intake airways away from downcast shafts (6) .. .. .. .. .. Stopes (7) .. .. .. .. .. Development ends (8) .. .. .. .. Shafts (9)J ................................................... Not applicable Not applicable Not applicable + 27 %t + 19 %t Nil Nil + 44 %f Nil Nil Nil Nil 1 % Vaseline in Xylol (1949) * + 11 p.p.c.c. Nil Nil Nil Nil Nil *The question of whether the effect of the change would be proportional to the count was considered but rejected in favour of a constant value on account of the fact that the altered technique resulted in an improved field which would be independent of the dust content of the air sampled. tEffect of 1940 and 1942 changes combined. JThe effects of the changes in technique could not be proved to have a significant effect on the annual means recorded, but allowance was made for an apparent disproportion in the number of spots taken during blowing over operations. places where rock is excavated, referred to as MISCELLANEOUS MINERS; (4) those employed largely along en trances to and exits from, but to a certain extent also in working places where rock is excavated, referred to as the " S.D.M." group, standing for stoping, development and miscel laneous mining; (5) those employed essentially on the three main methods of rock excava tion, viz., stoping, developing and shaft sinking. 1. The following occupations were in cluded under SHAFT WORKERS: Hoist Drivers, Onsetters, Shaft Timbermen, Underground Banksmen, Underground Storemen, Underground Pump Attendants, Underground Riggers, etc. 2. The following occupations resort under TRAVELLINGWAY WORKERS: Loco Drivers, Haulage Timbermen, Underground Artisans (e.g., Fitters and Electricians), etc. 3. The following occupations were included under MISCELLANEOUS MINERS: Drive Timbermen, Trammers, Pipe Fit ters, Track Layers, Reclaimers, Sand Fillers, Waste Packers, Haulagemen, etc. The following occupations were included in the " S.D.M." class: Early Examiners, Learner Miners, G.M.T.S. -- Apprentices, Assistant Miners, etc. 5. The fifth class constitute the virgin-- rock breakers and this group was sub divided into: Stopers, Developers and Shaft Sinkers. Some occupations were regarded as com binations of the above such as developmentcleaners, stoping-developers, stope-timbermen, etc. The relative weightings of the strategic dust levels used for these groups of workers are given in Table II. Mine Officials were treated differently. In their case it was argued that the dustiness of their underground shift would approxi mate the mean dustiness of the main groups of non-official workers whom they supervise. Journal of the Mine Ventilation Society of South Africa, July, 1959 167 A Correlation between Silicosis and the Dust Exposure of South African Gold Miners TABLE II Estimated Proportions of Shift at the Strategic Dust Levels Type of Mine Worker Shaft Workers ....................................... Travellingway Workers .. .. .. Miscellaneous Miners .. .. .. Developers .. .. .. .. .. Stopers .. . . .. .. .. Shaft Sinkers .. .. .. .. 0-1 0-50 0-20 0-10 005 0-05 0-05 0-3 0-50 0-50 0-25 0-10 0-10 0-10 Dust Level Proportions 0-6 0-7 (See fi gure 2) _ 0-30 0-50 0-50 0-30 -- _ -- 0-10 -- 0-55 -- 0-8 __ -- 0-05 0-35 -- --- 0-9 Total ___ _ -- -- 0-85 1-0 1-0 1-0 1-0 1-0 1-0 TABLE m Method of Dealing with Officials Type of Official Combination Underground Foremen and Engineers .. .. Travellingway and Shaft Workers Shift Bosses .. .. .. .. .. .. Stopers, Developers, Miscellaneous Miners and Travellingway Workers Mine Overseers, and Miscellaneous Officials .. Those for Shift Bosses with shaft workers added thereto The combinations made for this purpose are given in Table III. * Samplers were given an extra 30 per cent, onto that of Miscellaneous Officials to allow for the extra dust to which they are considered to be subjected during the chiselling of samples. The trends in shift dustiness thus obtained for the various main groups of workers are depicted in Fig. 3. Skipmen presented a special problem because although their shift is essentially spent in the shaft area, the mean dust level to which they are subjected is above that for shafts because of the puffs of dust caused by the filling of skips, to portions of which skipmen are subjected. To allow for this, it was arbitrarily assumed that skipmen would on an average be subjected to a dust count 50 per cent, higher than that of " Shaft Workers." Of interest in connection with shift dusti ness are some comparisons of " estimates " with " actual " observations. Full shift dust studies were done in 1949 and 1950,26 and also in 1926.27 A comparison between estimates and actual observations for 1949 50 are given in Table IV. To do a comparison for the 1926 results, it was necessary to extrapolate the dust curves from 1931 back to 1926 and, secondly, to allow for the changes in tech nique. Due to changes in technique, it was estimated that the 1926 results would have been about 30 per cent, higher had the pre sent technique been followed. A com parison for the classes of workers on which actual observations were made in 1926 are given in Table V. The agreement is good and shows that the method followed has some merit. 1.7 Dust exposures The next step was to compute the dust exposures of underground workers from their records of service, but as the duration of the underground shift varies from one occupation to another, it was decided to allow for this. The following figures were considered reasonable estimates of the durations of underground shifts: 168 Journal of the Mine Ventilation Society of South Africa, July, 1959 A Correlation between Silicosis and the Dust Exposure of South African Gold Miners Non-Officials: Artisans -- 7 hours. All others -- 8 hours. Officials: A. Mining: (i) Shift Bosses -- 6 hours. (ii) Mine Overseers -- 3^ hours. (iii) Managers -- 3 hours (this occupa tion did not " come up "). B. Survey and similar officials: (i) Juniors -- 4 hours. (ii) Seniors -- 3\ hours. This occupa tion " came up " very seldom). C. Engineering officials: (i) Underground Foreman -- 6 hours. (ii) Underground Engineers -- 3 hours. (This occupation " came up " very seldom.) These figures were used to compute the dust exposures as the product of time TABLE IV " Actual " versus " Estimates " for 1949-1950 Shift Dustiness Class of Worker Estimates for 1950 Actual Observations (1949-1950) during full-shift dust studies Number Observed 1. Travellingway Workers .. .. .. .. .. 65 p.p.c.c. 2. Skipmen .. .. .. . . .. .. .. 75 p.p.c.c. 3. Miscellaneous Miners .. . . .. .. .. S3 p.p.c.c. 4. Early Examiners .. .. .. .. .. .. 101 p.p.c.c.* 5. Developers . . . . . . . . . . . . . . 103 p.p.c.c. 6. Stopers .. .. .. .. .. .. .. 116 p.p.c.c. 7. Shaft Sinkers .. .. .. .. .. .. 161 p.p.c.c. 66 p.p.c.c. 78 p.p.c.c. 75 p.p.c.c. 103 p.p.c.c. 90 p.p.c.c. 120 p.p.c.c. 180 p.p.c.c. 6 13 16 27 43 67 8 *As the mean of stopers, developers and miscellaneous miners. Journal of the Mine Ventilation Society of South Africa, July, 1959 169 A Correlation between Silicosis and the Dust Exposure of South African Gold Miners TABLE V " Actual " versus " Estimates " for 1926 Class of Worker Miscellaneous Miners .. .. Early Examiners .. .. .. Stopers .. .. .. .. Developers .. .. .. .. Shaft Sinkers .. .. .. Samplers .. .. .. .. Observed 1926 185 206 216 316 284 240 Mean Counts p.p.c.c. Obs. x 1 3 (" Actual ") Extro-polated from 1931 (" Estimated ") 240 241 300 268 320 281 366 410 378 382 325 313 " Actual " " Estimated " 1-00 M3 1-14 0-89 0-99 1-04 (expressed in hours) and particles (as ob tained from konimeter sampling). From the records of service could be ob tained the mine worker's occupations, the calendar period of each occupation and the number of shifts employed on each occupa tion. By means of the shift dustiness curves against calendar years (see figure 3), the mean dustiness per shift for each period of occupation could be computed. Hence, for each period of occupation, the dust exposure as the product of time and dust concentration could be estimated and hence the total dust exposure during a mine worker's whole underground career could be estimated. An example of such an estimation is given in table VI. 2. The Sample of Mine Workers The next item I wish to deal with is the method of obtaining the sample of mine workers. It was decided to base the investigation on post-mortem, rather than on X-ray findings because of the former being the more reliable finding. It means that deceased mine workers only qualified for selection. The cause of death, whether by accident, from sickness or from any other cause did not matter. Only cases whose lungs were examined under the microscope were included. Additional conditions laid down for the selection of cases were that they had to be European mine workers who: (1) were found free from tuberculosis; (2) commenced underground service after 1925; (3) worked in controlled gold mines of the Witwatersrand and Extensions and the Orange Free State only; (4) did not have breaks from under ground service totalling more than five years; (5) continued to work underground up to within five years of death; (6) became deceased after 1955, and (7) had more than 10 years' service. During the period at my disposal, a total of 57 cases qualified for inclusion. 3. Medical Classification From post-mortem examinations, the con dition of the lungs with regard to silicosis, could be cast into the following broad classifications: (0) No dust reticulinisation. (1) Dust reticulinisation. (2) Dust reticulinisation with collagen- isation. (3) Occasional islet. (4) Few islets. (5) Moderate number of islets. (6) Large number of islets. (7) Conglomeration. (None of the cases were in this state.) The Composition of the Sample Selected: The composition of the sample selected with respect to class of mine worker is most probably not the same as that of the under ground European labour force on account of the limitations that had to be laid down for cases to qualify for inclusion. However, a difference in composition does not matter because the investigation does not concern 170 Journal of the Mine Ventilation Society of South Africa, July, 1959 A Correlation between Silicosis and the Dust Exposure of South African Gold Miners TABLE VI Computation of Dust Exposure in Particle-Hours 1 23 Oc.n No. 1 2 3-1 3-2 3-3 4 5-1 5-2 5-3 5-4 6-1 6-2 6-3 6-4 6-5 7-1 7-2 8-1 8-2 9-1 9-2 9-3 9-4 101 10-2 11-1 11-2 12-1 12-2 12-3 12-4 12-5 Shifts 526 169 201 29 331 268 114 94 95 13 330 294 49 24 69 178 134 833 30 282 23 29 1 271 6 9 46 131 109 46 282 1,963 H* P S 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 6 8 8 6 8 8 8 8 8 8 8 *Hours per shift. fHours per occupation. jParticle-hours per occupation. the silicosis incidence of the present Euro pean labour force, but the relationship be tween the dust exposure of a person in general and his degree of silicosis. Never theless, the factors and the direction in which they could cause the composition to deviate from that of the existing European labour force are of interest, and they will be considered here. (a) Death was the prime selector! The cause of death did not matter. It could be anything and none died from silicosis. A bias with respect to condition of the lung is, therefore, excluded, but the workers employed on the more dangerous occupa tions probably stood a better chance of 4 H.p.t Oc.n (2 x 3 t- 100) 42-6 13-5 16-1 2-3 26-5 21-4 9-1 7-5 7-6 1-0 26-4 23-4 3-9 1-9 5-5 14-2 10-7 66-6 2-4 22-5 1-8 1-7 0-1 21-7 0-4 0-7 3-7 10-5 8-7 3-7 22-6 157-0 5 P.P. C.C. (Mean Shift Dustiness) 231 234 159 211 147 133 125 121 118 159 131 100 126 95 94 111 119 87 116 84 64 91 116 83 90 83 83 83 83 83 83 50 Total: 6 P.H.p.f Oc.n (4 x 5 v 10,000) 98-4 31-6 25-6 4-9 39-0 28-5 11-4 9-1 9-0 1-6 34-6 23-4 4-9 1-8 5-2 15-8 12-7 57-9 2-8 18-9 1-2 1-5 0-1 18-0 0-4 0-6 3-1 8-7 7-2 3-1 18-8 78-5 578-3 being included than other underground workers. (b) The second factor was the fact that the next-of-kin might receive increased (but not decreased) compensation as a result of post mortem examinations where the deceased had in life not yet been certified to be in the second stage (and the consent of the nextof-kin was necessary for such examina tions). Decreased compensation is not possible, because even if the deceased is at post-mortem found in a stage less advanced than he was in life certified to be, the com pensation rate remains at that for the more advanced stage. (c) The compensation factor also acted Journal of the Mine Ventilation Society of South Africa, July, 1959 171 A Correlation between Silicosis and the Dust Exposure of South African Gold Miners in another way, viz., on account of the pos sibility of increased compensation to the next-of-kin, preference is given to post mortem examinations on cases who were in life not in the second* stage of pneumo coniosis. This is done to cause the least possible delay in the payment of increased compensation where due. Factors (b) and (c) were probably responsible for the stage reached during the investigation that, when 51 cases qualified for inclusion, only two cases were beyond the second stage. It was then decided to specially select from there onwards, only cases beyond this stage. Only six cases could be found within the time at my disposal. 4. The Results Obtained The results obtained are given in Table VII and depicted in Fig. 4. As can be seen there is a very wide scatter and in order to test the association between silicosis and dust exposure, the results were analysed statistically. obtaining the sample will be restated briefly: I set up a " screen " for a certain period. All the cases which passed the screen were included. This gave cases No. 1 to 51. Only 2 out of these 51 cases were beyond classification " 4 " on the Pathologists' scale. The screen was then altered so as to include only cases beyond this stage from there onwards. These cases naturally had to satisfy all the other conditions previously laid down. This produced the final six cases, viz., Nos. 52 to 57. As explained before, the sample selected is most probably not representative of the composition of the underground European labour force, neither with respect to class of worker nor with respect to the prevalence of silicosis. However, the sample may be subdivided into classes with respect to silicosis, and as far as I can judge, the dust exposures in each class is a random sample of the popu lation of dust exposures of that class. Furthermore, the dust exposures appear to be normally distributed because the cumu lative frequency distributions of the dust exposures plot on fairly straight lines on probability paper. Random estimation theory was, therefore, assumed to hold and the association between silicosis and dust exposure, was assessed by using the variance ratio test28 (on the advice of Prof. Kerrich). The calculations are given in Table VIII. The variance ratio test compares the variance between classes with the variance within classes. As there are 57 observations that can be cast into 7 classes, there are 50 degrees of freedom within classes, and 6 degrees of freedom between classes. Fio. 4. Silicosis Vi. Konimeter particle hours. 5. Statistical Analysis Random Theory: A question that arises in connection with the sample selected is whether random theory holds. To provide the answer to this question, the method of * Full compensation is paid to the next-of-kin from the second stage onwards. The variance between classes comes to 9 32, and the variance within classes comes to 1-71. The variance ratio is, therefore, 5-45. For 6 and 50 degrees of freedom the variance ratio is 4-93 for the 99 95 per cent, level. The variance ratio, therefore, reaches beyond the 99-95 per cent, level which is very, very significant. For purposes of comparison, the associa tion between silicosis and years of service, between silicosis and underground shifts and between silicosis and the estimated under ground hours, were also assessed in the same manner as described for particle-hours, 172 Journal of the Mine Ventilation Society of South Africa, July, 1959 between Silicosis and the Dust Exposure of South African Gold Miners TABLE VII The Results Obtained Years Service UGD Shifts (1,000's) UGD Hours (10,000's) Particle Hours (Millions) P.M. Finding Bureau Certification 1 11-8 2-84 2-27 2-44 0 0 2 21-0 5-65 4-52 6-52 6 III 3 21-1 5-97 4-77 3-22 0 0 4 15- 7 3-97 3-17 3-24 4 I 5 12- 3 3-02 2-17 2-22 4 I 6 28- 9 8-24 6-59 6-82 1 I 7 17-2 4-63 3-71 2-98 1 0 8 20-6 3-18 2-55 1-32 0 0 9 14- 9 4-76 3-81 3-55 2 I 10 20-8 5-55 3-42 2-36 0 0 11 19-9 4-56 3-66 3-43 1 0 12 31-8 7-34 4-17 3-95 0 I 13 23-3 6-61 5-29 3-82 2 I 14 26- 4 6-99 5-58 5-78 2 I 15 29- 0 7-97 5-41 5-24 0 0 16 27- 3 7-24 5-79 4-97 1 0 17 23-1 5-00 3-26 2-28 0 0 18 12-6 3-20 2-55 2-86 0 0 19 16- 9 4-45 3-56 4-01 4 I 20 28- 8 8-20 6-56 5-39 3 I 21 17- 9 4-21 2-95 1-49 0 0 22 13- 8 3-42 2-73 2-41 0 0 23 15- 0 4-11 3-28 2-95 0 0 24 19- 6 5-77 4-60 4-54 3 I 25 21-0 4-56 3-65 4-03 3 I 26 25-1 6-67 5-33 6-68 4 I 27 15- 4 3-36 2-69 2-79 4 I 28 20- 4 5-71 4-55 4-03 4 1 29 16- 9 4-54 3-59 4-22 2 I 30 17- 5 4-94 3-93 3-78 3 I 31 14- 7 3-48 2-78 2-86 0 0 32 27-5 7-88 5-52 3-45 1 I 33 14-9 3-90 3-11 3-25 0 0 34 16-5 4-47 3-21 3-34 0 0 35 21 -7 7-23 5-04 4-95 0 0 36 27-6 6-84 2-74 3-73 4 I 37 19-8 5-52 4-32 4-09 4 I 38 19-0 4-32 3-46 3-82 0 0 39 18- 3 4-32 3-45 3-91 4 I 40 14- 1 3-59 2-87 2-76 4 I 41 13-8 3-13 2-51 2-64 3 I 42 19- 5 4-23 2-67 2-67 1 0 43 15- 0 3-39 2-00 1-99 1 0 44 25- 7 6-80 5-44 5-28 5 II 45 15-2 5-45 4-36 2-34 2 I 46 26- 2 8-71 6-97 5-50 2 I 47 21- 0 6-21 4-50 3-83 1 0 48 21- 5 4-65 3-33 4-29 2 I 49 25-7 7-00 4-90 3-69 0 0 50 22- 8 5-40 2-58 2-77 0 II 51 25-8 5-68 4-09 5-23 0 0 52 19- 7 4-45 3-56 3-76 5 II 53 27-0 7-57 6-06 7-39 6 IV 54 27-2 6-97 5-58 7-76 5 II 55 30- 3 7-83 6-15 8-76 5 II 56 20- 3 5-43 3-60 3-60 5 II 57 23-1 6-04 4-85 6-38 5 III Mine Ventilation Society of South Africa, July, 1959 173 A Correlation between Silicosis and the Dust Exposure of South African Gold Miners but the levels of significance reached, were much lower than that for particle-hours. For silicosis versus underground hours, the 96 per cent, level was reached which is, therefore, just significant; for silicosis versus underground shifts, only the 80 per cent, level was reached, whilst for silicosis versus underground years, only the 60 per cent, level was reached. (Compare Figs. 5, 6, 7 and 8.) degree of silicosis and dust exposure, as based on the results actually recorded or else very likely to be recorded during routine rust sampling with the other instru- 1------1----- r k'"' --->r KEY X - Individual cases = Mean .95/. C L- T------ 1--l--------r 1(8 Cases atYandt") 31 2 [22 Cases at*2;'3V4') -- YU........... " _KEE - / X Indicates individual ca5e / means / n h 95% CL. / (8 Cases at *i^ Ml tin |Ci9 Cases at V / t-xx--12* ----- xt-5----- 1 I 2 ^3 4- 5 6 ]__ J_ 7 Particle-hours in millions. Fig. 5. Particle Hours. cco- a^s: . Cl, V J- 2 '33.... 4 r5 4-6 7 8 Number of shifts in thousands. Fig. 7. Shifts. 6 _ KEY X ^Individual cases 5- e = Mean d * 95/o C L. 5 4 o to 3 - 2- I 0 / / / / / I 1< 1 * XX KEY A = Individual cases = Mean *SS% C.L Number of hours underground in ten-thousands. Fig. 6. Hours. Part VI: Concluding Remarks 1. (a) The instrument In view of the strong association between degree of silicosis and particle-hours, there is no need to replace the konimeter with another instrument before a significantly closer association can be established between .5 4 c> C~o 3 2 I 0 ` v k xjLX ---- x 5 10 f5" ` '20""2f" "30' Years of service. Fig. 8. Years. ment. I wish to make it clear, however, that I do not mean to say that there is no need for experimentation to try and find a better instrument. (b) The method of treatment What applies to the instrument also applies to the method of treatment of slides. 2. Application What is really needed is the association between given dust exposure and degree of silicosis, but this is not readily obtainable in practice. However, as a given degree of silicosis has been found to be strongly asso- 174 Journal of the Mine Ventilation Society of South Africa, July, 1959 A Correlation between Silicosis and the Dust Exposure of South African Gold Miners TABLE Vm Variance Ratio Test for Particle-Hours P.M. Particle Classifi hours in n Ex cation millions 0 2-44 5-24 19 60-43 2-95 3-82 3-22 2-28 2-86 3-69 1-32 2-86 3-25 2-77 2-36 1-49 3-34 5-23 3-95 2-41 4-95 1 6-82 2-67 8 30-10 2-98 1-99 3-43 3-83 4-97 3-45 2 3-55 5-50 7 29-50 3-82 4-29 5-78 4-22 2-34 3 5-39 2-64 5 20-38 4-54 4-03 3-78 4 3-24 4-03 10 37-46 2-22 3-73 4-01 4-09 6-68 3-91 2-79 2-76 5 5-28 3-60 6 35-54 3-76 6-38 7-76 8-76 6 7-39 6-52 2 13-91 57 227-38 X 3-181 Ex2 214-4373 nx2 192-4186 E(x--x)2 22-0187 s2 1-223 3-790 129-5190 114-0790 15-4400 2-2057 4-204 132-5414 124-3214 8-2200 1-370 4-076 87-1626 83-0689 4-0937 1-0234 3-746 153-7002 140-3252 13-3751 1-496 5-923 232-6356 210-5034 22-1322 4-426 6-955 3-989 97-1225 1,048-6052 96-7441 907-0188 0-3794 85-6591 141-5864 0-379 (S.S.D. within) (S.S.D. between) df S.S.D. 6 Between classes 55-9273 50 Within classes 85-6591 1-3332 56 Total 141-5864 v2 = 5-445 (6,50 d.f.) (compare with 4-93 at 99-95% level) (29) dated with the particle-hours, the opposite must follow, viz., that a given number of particle-hours must be strongly associated with the degree of silicosis. In view of the wide scatter of particlehours for a given degree of silicosis, it will not be of much practical value to attempt the prediction of the degree of silicosis in a person from his particle-hour exposure as estimated in this investigation. Never theless, the mean particle-hours obtained for a given degree of silicosis may serve as Journal of the Mine Ventilation Society of South Africa, July, 1959 175 A Correlation between Silicosis and the Dust Exposure of South African Gold Miners a useful guide to aim at in order to keep the mean degree of silicosis in a population of workers below a certain level. The mean relationship obtained (see figure 5) reveals that if it is requred to prevent the mean degree of silicosis in a body of men from getting beyond that at which the mean certification is likely to be for pneumoco niosis in the first stage, the mean dust ex posure should not get beyond about 4 million particle-hours. Example: An example calculated from the figure of 4 million particle-hours to contract pneumo coniosis in the first stage is of interest. Take developers. The estimated mean dustiness per shift of a developer (by 1955) is about 100 p.p.c.c. At 8 hours per day and at 290 shifts per year, it will, therefore, take a developer under present-day conditions about 17 years to accumulate 4 million particle-hours (if he always remains on development). A person cannot start as a developer right from the first shift, because experience in the " miscellaneous " class of work is essen tial, where the dust exposure is less than that as a developer. This will lengthen the period to accumulate 4 million particlehours of dust exposure towards, say, 20 and more years, which figure compares favour ably with that found in practice. 3. Records of Service In view of the significant association obtained between degree of silicosis and underground hours, it is recommended that the underground hours per shift be added to records of service in future and, further more, that the feasibility of arriving at the mean dustiness per shift via routine dust sampling by mine ventilation officers and of also recording this figure on the records of service be investigated. (This, I believe, is done for coal miners in West Germany.) With such a procedure, it would be pos sible to detect persons who accumulate dust exposures undesirably rapidly. Steps could then be taken to retard their dust exposure accumulation rate. This could be achieved in a variety of ways such as: reducing the dust levels of their working places, moving them to working places with lower dust levels, getting them to use dust protective equipment, shortening their underground shift, reducing the number of their under ground shifts per week and so on. 4. Recorded annual means The value of the mean dust counts recorded in the Government Mining Engin eer's Annual Reports would be greatly en hanced if the observations taken where conditions were suspected or known to be bad, and those taken where conditions were suspected or known to be good, such as those taken to check that improvements called for had been effected, were kept seperate from those taken at random. The value of the records would be further enhanced if the confidence limits to the means could be given. In conclusion, Mr. President, I wish to express the hope that this investigation will help to reduce the number of persons con tracting silicosis in years to come. BIBLIOGRAPHY 1. BEADLE, D. G., M.Sc.: An investigation into the performance and limitation of the Konimeter, Journal of the Chemical, Metal lurgical and Mining Society of South Africa, July, 1951, Vol. 51, No. 9. 2. PATTERSON, H. S., on: Comparison of thermal precipitator with konimeter figures, in: The prevention of silicosis on the mines of the Witwatersrand, Govt. Printer, Pretoria, 1937, p. 88. 3. LAMBRECHTS, J. de V., M.Sc. Eng. (Rand): A critical review of dust sampling instru ments; Chem., Met., and Min. Soc. of S.A. December, 1940. 4. LAMBRECHTS, J. de V.: In his reply to discussions on " 3," The J. Chem. Met. and Min. Soc. of S.A. (March-April, 1943), Vol. XLIII, No. 9 and 10, p. 157. 5. LAMBRECHTS, J. de V., M.Sc., Eng. (Rand): Report on an investigation into dust and ventilation conditions in the Copper Mines and the Broken Hill Mine in Northern Rhodesia with particular reference to silicosis, Govt. Printer, Lusaka, 1945. 6. RABSON, S. R.: The performance of the circular konimeter, The Mine Ventilation Society of South Africa, May, 1946. 7. DAVIES, AYLWARD and LEACY: Im pingement of dust from air jets, A.M.A. Archives of Ind. Hyg. and Occ. Med., October, 1951, p. 369. 8. HASENCLEVER, D.: Untersuchung ueber die Eignung verschiedener Staubmaeszgeraete 176 Journal of the Mine Ventilation Society of South Africa, July, 1959 A Correlation between Silicosis and the Dust Exposure of South African Gold Miners zur betrieblicher Messung von mineralischen Staueben, Staub, Heft 41, (September, 1951), p.p. 388--435, Bonn, Germany. 9. ROEBER, Dr. R.: Untersuchungen zur konimetrischen Staubmessungen., Staub Hefte 48, 49 and 50 (1957), p. 445. 10. ROACH, S. A.: Pneumoconiosis Research Unit of the Medical Research Council, Llandough Hospital, Nr. Cardiff on: Measur ing the worker's environment; British Journal of Industrial Medicine Vol. II, No. 1 (1954) p. 51. 11. ROACH, S. A.: A method of relating the incidence of pneumoconiosis to airborne dust exposure, Brit. J. Industr. Med., October, 1953. Vol. 10: p. 220. 12. WYNN, A. H. A. (M.A.) and DAWES, J. G. (B.Sc.): The size classification of airborne dusts in mines; S.M.R.E. Research Report No. 28 (1952), p. 34. 13. WYNN, A. H. A. (M.A.): The assessment of airborne dust in mines. Research Report No. 52 (1952), p. 6, of the Safety in Mines Re search establishment. 14. WRIGHT, B. M.: The importance of the time factor in the measurement of dust exposure; Br. J. Ind. Med., October, 1953, p.p. 237-- 238. 15. THEODORE HATCH: Permissible dustiness, A. Ind. Hyg. Association quarterly 16: 1 March. 1955. 16. SABOTTA, Dr. Johannes: Atlas of Human Anatomy, Vol. II (1939). 17. PNEUMOCONIOSIS CONFERENCE held in Johannesburg, February, 1959, Recommen dation III. 3, The Journal of the Mine Venti lation Society of S.A., June, 1959. 18. THE GOVERNMENT MINING ENGI NEER'S ANNUAL REPORTS for 1931-- 1955, The Govt. Printer, Pretoria. 19. TRANSVAAL CHAMBER OF MINES. Forty-seventh Report. Year 1936, p. 115. 20; MINES AND WORKS REGULATION No. 161 (10), Hortors Ltd., Leader Buildings, Jo hannesburg. 21. MINES AND WORKS REGULATION No. 143(2)(a). 22. THE GOVERNMENT MINING ENGI NEER'S ANNUAL REPORT 1936, p. 87. 23. THE GOVERNMENT MINING ENGI NEER'S ANNUAL REPORT 1940, p. 89. 24. THE GOVERNMENT MINING ENGI NEER'S ANNUAL REPORT 1942, p. 96. 25. THE GOVERNMENT MINING ENGI NEER'S ANNUAL REPORT 1949, p. 102. 26. THE GOVERNMENT MINING ENGI NEER'S ANNUAL REPORT 1950, p. 110. 27. MINERS' PHTHISIS PREVENTION COM MITTEE on: "Dust to which miners are ex posed during a shift." (Sept., 1926 and March, 1927), M.P.P.C. 6/2, (Unpublished). 28. HALD, A.: Professor of Statistics, University of Copenhagen, on: The V2 Test of Signi ficance in: Statistical Theory with Engineer ing Applications (1952), p. 379. 29. HALD, A.: Statistical tables and formulas (1952), Table VII: Percent fractiles of the V2 Distribution. Contribution to Mr. du Toit's Paper by D. G. Beadle For many years I have been convinced that one of the major problems to be solved in connection with silicosis is to establish the relationship between the amount of dust breathed and the effect of this dust on the human lung. This was a view widely held at the recent Pneumoconiosis Conference in Johannesburg, which recommended further work on the subject. It forms the basis of the elaborate Pneumoconiosis Field Re search in Great Britain and the " Occupa tional Dust Sampling " project in which I am assisting the South African Pneumo coniosis Research Unit. Only when we have this information established on a scientific basis will we be able to set scientific standards of dust levels to be aimed at as, at present we do not know, even if we do achieve the arbitrarily set standards to which we try to work, whether pneumoconiosis will be prevented. In selecting a subject for his thesis--on the attainment of which I want now to for mally congratulate him--Mr. du Toit very courageously selected this field. I say courageously because it is a very com plicated field, full of frustrating pitfalls and often complete blanks in our knowledge; it also requires a tremendous amount of work to handle the data one accumulates. I imagine Mr. du Toit must have spent many hundreds of hours of his spare time making the numerous calculations involved in ob taining the results which he had, of necessity, had to quote so briefly today. Mr. du Toit has obviously had to face up to many difficult decisions while carrying out this work. For example, who would care to have to predict what the dust con centrations were in development ends in say 1928, as they would have been if measured by present-day konimeters and present-day techniques ? I have made a detailed study of Mr. du Toit's full thesis, and I would like to place on record that, in matters such as these, Mr. du Toit has made decisions which cannot be seriously chal lenged. He may not have always obtained Journal of the Mine Ventilation Society of South Africa, July, J959 111 A Correlation between Silicosis and the Dust Exposure of South African Gold Miners the correct figures, but I do not think any one could have made better estimates. He has described the methods he used, and I have no constructive comments to offer on them. I must confess, however, that I remain extremely sceptical of his ulti mate claim that his work shows " there is a strong association between degree of silicosis and konimeter particle-hours." I do not think this claim is warranted by the graph he has displayed. I do agree that when he plots " particle-hour " against post-mortem findings he obtains a better correlation than when he plots shifts or hours, but I know that several of our local statisticians do not agree it is a strong correlation. His data un doubtedly shows that putting in the koni meter count does improve the relationship but nevertheless there is such a tremendously wide scatter in the data that one must re serve judgement on the actual value of the correlation. For example, 5 million particle hours can lead to anything from no silicosis to stage 5. mining careers are identified. One sees at a glance that this is one factor which has a tremendous influence on the data. All those men who contracted silicosis with very short dust exposures are those who joined the industry very recently; those who tolerated a much greater dust dose before developing a given stage of the disease are the " oldtimers " who started before 1930. This shows that konimeter particle-hours is not by any means the only factor involved. There can be various explanations for this effect. To suggest a few only-- (i) Men entering the industry in the period 1940 onwards were far more susceptible to the disease--did the demands of wartime lead to a re laxation of the standards set by the Bureau for new admissions to the industry ? (ii) Mr. du Toit's dust data for the various years might be wrong. Per haps there has been much more dust since 1940 than he allowed for, and perhaps there was less in the pre1930 days than the figures he used. Adjustments in this direction would improve the " fit " of his points very greatly. Particle-hours in millions. (iii) The konimeter may not be measur ing the dust correctly. Mr. du Toit's time graphs show that the konimeter dust counts decreased significantly with time, but the fit of this data shows it does not fully account for the variations in disease. The koni meter is very sensitive to coarse dust, and possibly this is the only type of dust which has been reduced over this period; possibly the amount of fine dust--which is more dangerous --has not been decreased; it may even have increased. One can easily explain by this hypothesis this " date-of entry " effect I have demonstrated. Mr. du Toit has been kind enough to let me have access to his detailed calculations and to copies of some of the basic data. I have examined this in some detail, and can take the results a little further. In this figure I have replotted his data in a slightly different form, the important change being that the years in which the men started their (v) Possibly men vary greatly in their susceptibility to equal dust dosage. Perhaps the post-1940 entrants in Mr. du Toit's data represent highly susceptible men and he has been left with only the most resistant men of the pre-1930 entrants. I do not know if one of these, or some other explanation, is the true reason for the 178 Journal of the Mine Ventilation Society of South Africa, July, 1959 A Correlation between Silicosis and the Dust Exposure of South African Gold Miners effect I have demonstrated. All I can say at this stage is that factors other than konimeter count must have a big influence to account for the scatter in the data. Mr. du Toit quite rightly says that any decision to replace the konimeter should depend on whether improved apparatus will show a significantly better relationship. I have applied the data we have obtained to date in our occupational dust sampling pro ject to Mr. du Toit's data, and we can claim that the lit of our thermal precipitator data is better than the fit of our konimeter figures, but we have insufficient data to warrant any decision being made at this stage. Which ever parameter we use to plot against the post-mortem finding we still get a very large scatter in the points. I might add that our konimeter values obtained by actual measurement of the average daily dust exposures of different occupational groups do not agree very well with some of Mr. du Toit's assumed figures for the present day. For example, he assumes that underground foremen artisans have a present-day average dust exposure of 60 p.p.c.c. I agree that this class of men normally works in or near intake air just beyond shaft bottom, but Mr. du Toit's figure assumes that they do not create much dust themselves. In practice these men, particularly fitters, create quite a lot of coarse dust, released from the material they work on, and the konimeter gives high readings when sampling such dust. Our actual observed figure for this class of men is that their average daily dust exposure, by konimeter count, is over 200 p.p.c.c. Mr. du Toit has said that men should not be exposed to more than approximately 4 mil lion particle hours to have a reasonable chance of awaiting pneumoconiosis. On this basis most artisans should contract a similarly advanced stage of pneumoconiosis after about 12 years--I think this is far from the truth ! Incidentally, I strongly support his sug gestion that individual dust exposure records should be kept for all underground personnel. I think that before we can draw any valid conclusions from the data presented, or the data I myself have obtained to date, we must go very much more deeply into the whole problem. There are many uncertainties about the present figures-- (i) Are Mr. du Toit's 57 cases repre sentative of the whole mining popu lation ? (ii) Is it correct to plot the six stages of pneumoconiosis on a linear scale, i.e., is the step from stage 1 to stage 2 equal to the step from say stage 5 to stage 6 ? (iii) Can we draw reliable information from the cases who contracted pneu moconiosis within a very narrow time period ? (iv) What about the many men who have already had long periods in dusty but have not yet contracted pneumo coniosis--will they not alter the whole picture when they become cases for study ? (v) Has the best statistical technique been used to analyse this very com plicated problem ? We have a very wide field still open ahead of us in tackling the problem and much more work is needed before we can draw any valid conclusions. Although I am frankly doubtful whether Mr. du Toit's main conclusion can be substantiated at this stage I do sincerely congratulate him on his efforts. With the time and resources avail able to him he has done a magnificent job, but I hope he will agree that his effort is only a first " bite " at a very big apple. Mr. du Toit and I are working on parallel lines on this problem, but, unlike parallel lines, we do meet sometimes. We have had many stimulating discussions in the past, and I hope we shall continue to have them in the future. Mr. du Toit, I congratulate you on a paper and a thesis, which I have thoroughly enjoyed studying and which I consider a most useful contribution in a very important field. Professor J. E. Kerrich Mr. Chairman and Gentlemen, I should like to make the following comments on Mr. du Toit's paper. Everybody knows that the dust in our gold mines causes silicosis among the miners. Yet no one really knows how much exposure to how much dust produces a given severity of the disease. Mr. du Toit set himself the task of trying to bridge this gap to some extent. From the mining point of view he had Journal of the Mine Ventilation Society of South Africa, July, 1959 179 A Correlation between Silicosis and the Dust Exposure of South African Gold Miners the individual miner's record of where he had worked and for how long and what type of work he had done. In addition he had government reports extending over many years on average dustiness at certain selected points in the mines, such as at the top of downcast shafts, at the bottom of downcast shafts, at stope faces, and so on, where dustiness was assessed by particle counts with a konimeter. From the medical point of view Mr. du Toit had the actual state of the lungs for 57 deceased miners as assessed by post mortem findings. Informed opinion made him doubt the value of the X-ray assess ments of the state of the lungs. He pre ferred to wait until he had data on actual post-mortem findings and it took him three years to obtain as many observations as he did. A distressingly small sample ? Ad mittedly, but nothing can be done about it, for a long time to come. From the post-mortem findings each man was classified as belonging to one of seven arbitrary but carefully defined groups, repre senting medical opinion as to the progressive severity of the disease. From the miner's record it was a com paratively simple matter to obtain (a) years of service underground, (b) thousands of shifts worked and even (c) ten thousands of hours underground. From the individual miner's record, plus government reports on average dustiness it was possible to estimate (d) his total exposure to dust in millions of particle-hours. A laborious business in each case. Next the post mortem rating was plotted against (a), (b), (c) and (d) in turn. In the sample there is practically no association between rating and years of service, or be tween rating and shifts worked; there is slight association between rating and hours underground, and well marked association between rating and particle-hours. What generalisations can be made from these results. What population does this sample come from and what does the sample tell us about the population from which it is drawn ? Mr. du Toit has laid down a method of obtaining observations from a population. In principle, he can obtain more and more of these observations. They come from a somewhat macabre " population " of dead miners whose dependants hope to obtain additional compensation as a result of the post-mortem findings. I shall term it the post-mortem population. Two fundamental questions which must be sharply distinguished are: (A) Is this sample a random sample from the post-mortem population ? (B) How closely does the post-mortem population resemble of the mining popula tion as a whole? Under A: In the sample Mr. du Toit could, and to some extent did preassign the post mortem ratings. He could and did ask the doctors to pick out and examine lungs which were obviously badly affected. The result is that in the sample, the observed percentages of miners with ratings 0, 1, 2, etc., are not unbiased estimates of the corresponding percentages in the post-mortem population. The results of this fact can be very im portant. Take an imaginary example, crude and over-emphasised to drive my point home:-- For miners with number of particle-hours lying between, say, 3 or 4 millions, the situation might be as shown at the top of the next page. Thus in the sample the average rating for those with the given num ber of particle hours would be 4-15; in the population it would only be 1-85. Now, the basic question we want answered is, when we know from his record a miner's particle hours, what is the state of his lungs ? And the estimate obtained from Mr. du Toit's sample is biased, to an unknown amount. I do not believe that this bias is anywhere near as startling as in the distorted example I have given, but the blunt fact is that it is there and it is un known. On the other hand, in the sub-group from the sample consisting of those who have a particular post-mortem rating, say, 2, Mr. du Toit had no control over each individual's measures (a), (b), (c) and (d). Within each sub-group I can see no reason to doubt that such measures are distributed at random and accept his estimates of the average value of, say (d) within each sub-group of his post mortem population. If his confidence intervals for these results are wide it is because his sample was unavoidably small: but he has enough information to show that 180 Journal of the Mine Ventilation Society of South Africa, July, 1959 Rating Anemometer Rods 0 12 3 4 5 6 Per cent, in sample .. . . 5 5 10 10 20 20 30 : Total 100 per cent. Per cent, in population . . 30 20 20 10 10 5 5 : Total 100 per cent. that his measure of particle-hours for the individual is associated with lung condition. I for one doubted originally if this could be done, and congratulate him heartily on his result. Under B: Here I speak as a layman, but I think we are agreed that the post-mortem population is not quite representative of the mining population as a whole and do not know quite how this should affect our judge ments. In conclusion, in spite of minor criticisms, I think this is a valuable paper. I suggest that the next step should be to compare particle hours with a random sample of radiological estimates of lung condition, and would be interested to hear other people's opinions on this suggestion. ANEMOMETER RODS By B. J. R. Botha* and P. J. Finn* Introduction It is an accepted fact that comparatively large errors can be caused in some cases by holding an anemometer by hand (without a rod) in an airstream. These errors arise from the following causes:-- 1. Interference to the air flow through the anemometer caused by the ob server's hand an arm. It should be remembered that when anemometer calibrations are carried out the anemometer is held in position by means of suspension wires and not held by hand. 2. Interference to the general pattern of air flow at the measuring station in small airways due to the presence of the observer's body. 3. Difficulty of achieving the principle of traversing equal areas in equal times. This difficulty is at its greatest in large airways and in awkwardly shaped airways. 4. Difficulty of maintaining the anemo meter at right angles to the direction of the airstream (angle of yaw). There are few cases where the provision of a well designed anemometer rod attach ment will not enable the observer to do the job more easily, accurately and safely. * Consulting Engineer's Ventilation Department, New Consolidated Gold Fields Limited. Anemometer Rod for General Mine Ventilation Work A rod for normal mine work should be light, portable and reasonably robust. It should provide variable rod extension lengths and be provided with a " swivel head " in order to enable the anemometer to be set at an angle to the rod. It should also be so designed that the rod attachment point to the anemometer casing is at the same position as the anemometer control lever. Figure 1 illustrates such a rod. " A " shows the component parts of the rod. At the top left hand side can be seen the " stick end plug." This piece is 4 in long. Then follows a 3 ft rod (or rods) manu factured from f in diam. aluminium tubing with brass fittings at each end for screwing rods together. Lastly, the " swivel head " fitting is seen. Note that this fitting has been designed so that the threaded spigot end (for attachment to anemometer) can be tucked away safely inside the " swivel head " body for protec tion when in transit underground. When the three pieces as shown at " A " are screwed together the rod is ready for transport and has no screw threads exposed to possible damage. On arrival at site the spigot end of the " swivel head " is swung out (see " A ") and the rod is then ready for attachment to the anemometer. This is achieved by screwing Journal of the Mine Ventilation Society of South Africa, July, 1959 181