Search Documents

By T Barto, D Thomson, T Norgate, C Doblin

Collaboration between CSIRO Minerals and SunSalt Pty Ltd, a salt producing company located in rural Victoria, has resulted in a commercial process to recover magnesium sulfate (commonly called Epsom salt) from the saline waters of the Murray Basin. The feed for this process is bittern; the liquor remaining after halite (common salt) has been harvested from the saline water by evaporation.SunSalt produces salt and bitterns from three locations, namely Hattah (Victoria), Mourquong (NSW) and Wakool (NSW). The small size and considerable distances between these operations makes it uneconomic to build a stationary processing plant to extract value-added products, in particular Epsom salt, from the bitterns of the three operations. This fact led SunSalt to build a containerised mobile plant that could be moved to each site to process the bitterns at the end of the salt harvesting period.Epsom salt is extracted from the bittern by further evaporation, refrigeration and recrystallisation. In this process the bittern is diluted with five per cent volume of fresh water to minimise precipitation of residual halite in the bittern. Cooling the diluted bittern to between -5 and -10¦C precipitates up to 70 per cent of the magnesium sulfate, which is recovered as 85 to 96 per cent purity Epsom salt by filtration. This product is marketed as an industrial grade Epsom salt. Higher purity (up to 99.9 per cent) Epsom salt (USP or BP grade refined product) is made from the industrial grade product (raw Epsom salt) by recrystallisation.The containerised mobile plant was commissioned in December 2002. The raw Epsom salt generated in the mobile plant is refined to a higher purity Epsom salt in a separate plant located in Mildura. The Mildura plant was commissioned in December 2003.The CSIRO-SunSalt collaboration has demonstrated that the saline waters can be processed into a useful resource, and in addition could create jobs and a new skill base in regional Australia. The production of Epsom salt locally in Australia, which is worth currently about $A600 per tonne, removes the need to import over 3000 tpa from overseas with a net positive impact of $A2 million on the balance of trade.

Jan 1, 2004

By P W. Wypych, D Curry, A P. Grima

The market has become highly competitive for engineering and equipment vendors, who are challenged to reduce costs while improving efficiency and productivity of mine operations through plant improvements and expansions. Although conveyor transfer systems account for a small percentage of the infrastructure cost of a mine, they form an integral part of the process supply chain. Malfunctions or suboptimal performance can result in substantial financial cost from delayed start-up, unscheduled downtime, reduced throughput and increased maintenance costs.Discrete element modelling (DEM) is a powerful simulation method used to virtually prototype and test the performance of transfer equipment throughout a design cycle. When developed and applied correctly, DEM can provide insight into performance that exceeds the traditional analytical methods, especially when dealing with complex ore properties, behaviour and/or situations.Using case studies sourced from the authors’ recent research and industrial project work, this paper will address the industry goal to increase efficiency and productivity as well as the methods available to the engineers responsible for delivering designs that meet the aims of this strategic approach.The main focus of the paper will be the new calibration technology, which has been developed by the authors to generate DEM material models that are linked to real-world properties and flow behaviours via unique dynamic testing and innovative parameter calibration methods. These calibrated DEM material models can be deployed by engineers into their existing workflows, whether it be for feasibility studies, concept or final design, plant troubleshooting or optimisation.This new technology allows engineers to gain greater insight than that provided by traditional approaches to the design of transfer chutes and other process equipment and also allows informed design decisions to be made that meet industry efficiency and productivity requirements.CITATION:Grima, A P, Curry, D and Wypych, P W, 2015. New discrete element modelling calibration technology to facilitate the increased efficiency of ore handling and processing operations, in Proceedings Iron Ore 2015, pp 449–464 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Jul 13, 2015

By M Kim, J Park, I-K Suh, E Kim, J Lee

One of the key technologies in developing hydrogen-enriched ironmaking operations is the control of the carburisation, melting, and dripping behaviour of the iron ore burden materials. When a hot compacted iron (HCI) is charged in the blast furnace or the FINEX Melter-Gasifier, iron oxide reduction, the reduced iron’s carburisation, and the melting of iron and slag occur successively. This study investigated the reduction, carburisation, melting and dripping behaviour of HCI under a COH2 gas mixture using the high-temperature X-ray radiographic technique. Under 100 per cent CO gas atmosphere, the dripping temperature was 1410°C. When 5 per cent H2 was introduced, the dripping temperature slightly decreased to 1400°C. Furthermore, the dripping temperature gradually increased as the H2 gas composition increased. As 30 per cent H2 was introduced, the dripping temperature reached 1414°C. The carbon concentration in the drained metal samples gradually increased with increasing H2 composition. It is considered that H2 decreased FeO content in the slag, yielding the increased melting temperature and viscosity of the slag. On the other hand, H2 accelerated the carburisation of the reduced iron by CO gas, enabling the homogenisation of the oxide materials to form a liquid phase. Therefore, it is considered that introducing H2 in the reactor is helpful to increase productivity and energy efficiency by accelerating the liquid phase’s formation.

Aug 21, 2024

By J Harding, S Munro, R A. Bearman

Despite the importance of the primary crushing stage, performance optimisation is not well analysed with sufficient focus to give improvement opportunities. At the Newcrest owned Telfer operation, liner life plays an important part in scheduling planned maintenance. Major consideration is given to monitoring crusher performance, as the highly abrasive nature of the feed significantly impacts the liner wear life. A study was undertaken by the lead author which focused on establishing a strong baseline of accurate data pertaining to the performance of the crusher, in terms of, operating parameters, liner life and key machine design factors. As part of the study, mining feed materials were examined for a particular set of liners and this revealed that improved methods should be pursued. Due to the competitive nature of the mining industry, the business solution to complex problems with long shutdown cycles should be streamlined to incorporate a non-operational modelling approach that is significantly more efficient than traditional online trial and error methods. In order to dramatically reduce the time and business risks, it was decided to engage services to model and simulate the liner life by varying main operational and mechanical factors. The key to all simulation work is gathering high quality baseline data. Such data allowed deployment of the Kinematic Crusher Model and validation of the model against worn 2D wear liner profiles. The proven model was then successfully deployed to predict operational values that defined crusher performance. To enhance the study, further investigation into a set of mechanical and operational simulations generated a sensitivity matrix measuring key performance parameters throughout the wear life of the liners. The study revealed throughput as a highly sensitive parameter and a change in eccentric throw provided the greatest improvement in extending liner wear life.

Aug 24, 2022

By A Giblett, B Putland

This paper provides a simple guiding framework for grinding circuit optimisation, focussing on common themes in past optimisation success stories, the conditions for success and techniques that can be employed to identify and define opportunities. Many significant grinding circuit improvements have been reported for operating plants, based on recurring themes such as the optimisation of mill feed size, mill liner design, process control, critical size, grinding media and classification. The paper condenses these experiences into a set of simple questions for the operator to ask, and provides guidance towards identifying and achieving the optimisation goals.CITATION:Giblett, A and Putland, B, 2018. The basics of grinding circuit optimisation, in Proceedings 14th AusIMM Mill Operators' Conference 2018, pp 129–140 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Aug 29, 2018

By Caelli T, Robey M

Most research into planning of open pit mines has been based on evaluation of volume-element block models in terms of their net present value. The major issues revolve around two interdependent problems the optimum ultimate pit and time-based scheduling. Despite feedback between the two problems, they are usually treated as separate tasks.The ultimate open pit is defined by the removal of whole blocks so that their net present value is maximised. This gives planners an estimate of the total worth of the venture in present terms, and is crucial to obtaining funding. All cost and profit values are assumed to be static throughout the mine life, and the only physical constraint usually taken into account is that of safe, average, wall slopes, which must be approximated by the removal of whole blocks. Modelling of any further constraints (for example, access roads and true wall slopes) is usually achieved by adding them to the optimal pit by hand. Typically this degrades the model by between five and 20 per cent of its value - which then brings into question whether it is still the true optimum.This paper reviews the dominant methods in use for this first task the ultimate open-pit planning problem. Some new optimisations are given, and the algorithms are examined in terms of their efficiency and accuracy. Finally, suggestions are made for improvement of model accuracy and eventual expansion into solutions of the more complete dual problem - that of simultaneously finding the optimum ultimate pits and dynamic schedules for the venture.

Jan 1, 1996

By Maloszewski J, Kidybinski A, Sawka B

A forecast of mining conditions in Upper Silesia collieries for the beginning of the 21st century has special focus on average depth of mining, energy of seismic tremors induced by mining and energy of dynamic loads on roadway supports. A method of dynamic loads calculation is based on volume of failed rock surrounding an opening and the seismic energy of tremors anticipated. New types of yielding steel arch supports have been designed to resist massive static and dynamic loads and a system of supports has been tested both on 1 : 10 scale models and in natural size. The support sections used are 29, 36 and 44 kg/m.

Jan 1, 1988

By C Hedemann, N Silveste, H Davis, C J. Moran

For nearly two decades water scientists have been debating the need to use appropriate models for the questions being asked and the data available. This paper puts forward a rationale for a hierarchical framework for water balance modelling as it applies to mine sites. The hierarchy is based on the complexity of models and their data requirements. Scientific models are the most complex and require the most data; next is engineering models, then systems models and the simplest representations are given by conceptual models. A position is put forward that systems models have a range of uses that the others do not provide. A systems model, called SiteMiser, is described and its calibration illustrated. The model is applied to seven coal mines from the Northern Bowen Basin in Central Queensland. Currently, there is little relationship between total water use and coal production. This relationship is not any stronger when fresh (raw) water is compared to reused water. Leading practice water use levels are simulated for all sites. The simulation results demonstrate that there is considerable opportunity to reduce water consumption and, in particular, to import less fresh (raw) water to the sites.

Jan 1, 2006

By R P. Varden

This paper looks at the design of the Bentayga raisebore with subsequent back analysis using a LiDAR survey. A search for LiDAR raisebore assessment publications shows little work has been done on this subject. The paper examines the empirical design and RS2 modelling assessment with the bored condition. The high resolution of the LiDAR enabled mapping of weak structures that were located on the geotechnical investigation hole the orientations of which could not be determined. In addition, the scan showed were water ingress was occurring, potential stress overbreak and structural controls on stability. The LiDAR results were able to confirm no additional support work such as a shotcrete liner is necessary. The LiDAR results have been analysed in one metre increments as volume of overbreak against the empirical stability criteria (McCracken & Stacey 1989). The future use of LiDAR mapping over a range of rock mass conditions will provide quantification of the established empirical assessment improving its reliability. It is also a useful and accurate tool to monitor long term stability of a raisebore due to stress change or ore pass due to wear.

Mar 15, 2021

By B Levasseur, C Olsen, E Ghasemian Langeroudi, F Larachi, F Safizadeh, C Gagnon

Gold recovery by the cyanidation process is often accompanied by the leaching of other species. This leads to the depletion of the level of free cyanide and oxygen available for gold cyanidation. The largest proportion of the consumed cyanide is diverted by the cyanicides (eg S, Cu, Zn, Fe, etc) into non-valuable complexes.COREMÆs Extractive Metallurgy team has been conducting a research program on the improvement of the gold cyanidation process. As a part of this program, the current study aims to investigate the effect of certain operating parameters on the reactions involved during preoxidation and cyanidation. Flotation tailings and single mineral samples such as pyrite, sphalerite and chalcopyrite, along with their different combinations, were analysed to highlight the influence of ore mineralogy on the operating conditions. The effects of different kinds of oxidants (oxygen, hydrogen peroxide, ferricyanide and ozone) and chemical additives (ammonium sulfate) were investigated.Experiments conducted in alkaline solutions reveal that mineralogy is a key parameter since pyrite, chalcopyrite and sphalerite react quite differently during preoxidation and cyanidation. A significant increase in sulfur dissolution, especially on pyrite, was observed by applying preoxidation. H2O2 preoxidation showed a positive effect on increasing sulfate formation, a desirable by-product from the cyanide consumption point of view. Preoxidation of synthetic ores with ferricyanide resulted in boosting thiosulfate formation. Tests conducted on flotation tailings from a polymetallic mine revealed that a short preoxidation increases the dissolution of the ore leading to higher gold recovery and lower cyanide consumption. Regarding the formation of thiocyanate, the most undesirable by-product of the cyanidation, ferricyanide and ozone were found particularly efficient to hinder their formation. Finally, the addition of ammonium sulfate during the cyanidation process showed some promising results in terms of both gold dissolution and cyanide consumption.CITATION:Ghasemian Langeroudi, E, Levasseur, B, Safizadeh, F, Olsen, C, Larachi, F and Gagnon, C, 2013. Optimisation of operating parameters for cyanidation of sulfide-bearing gold flotation tailings in order to prevent the impact of cyanicides, in Proceedings World Gold 2013 , pp 111-120 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Sep 26, 2013

By G J. Halstead, R C. Steinhaus

The True Pipe® in-line sampler was developed in an attempt to address the need for a mechanically reliable and correctly designed sampler to suit one-dimensional pressurised slurry pipelines, often found in metallurgical process plants but for which no reliable and representative sampling solution as yet exists.The design, whose operation is based on the principle of symmetry, is described by Dominique François-Bongarçon (2005) with the synchronous opening and closing of relevant valves. It representatively isolates portions of the particulate material stream at predetermined intervals for effective extraction of primary increments for subsequent sample compositing.Theory of Sampling applicable solutions have proven elusive in flowing pressurised particulate material streams, with various attempts to date not having formulated a viable, mechanically reliable and acceptable design.Compliance to sampling theory is critical, and a number of tests were conducted where certain operating parameters were altered to determine their influence and sensitivity on sampling representativeness.The results of the initial series of tests performed on a carefully designed rig are presented, with additional work discussed for necessary investigation in the near future.Getting closer to a robust, economic and correct design for this challenging application in the minerals industry may now not be that far away. Plant management staff will see improved metallurgical mineral balances around their process plants and significantly reduced sampling discrepancies, particularly in tailings applications. Improved quality control is now possible where, due to plant head room constraints, classical cross stream sampling methods cannot be employed.CITATION:Steinhaus, R C and Halstead, G J, 2014. True pipe sampler – a correct design utilising the principle of symmetry in pressurised slurry particulate streams, in Proceedings Sampling 2014 , pp 127–134 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Jul 29, 2014

By Haase W. C

The content of this paper will identify recent developments in walking dragline technology including utilisation of planetary swing dnves, high voltage generators, unique bucket configurations and increased machine capacity resulting from changes to front-end geometry. The benefits to mine operators for increased performance, reduced cost, improved availabilities, simplified maintenance, and ease of operation will be identified and discussed. The feasibility and application of the technology as a retrofit on existing equipment will also be covered in the presentation. All the technology being presented has been recently commissioned into operating Australian mining operations and will be presented from a practical, as well as theoretical perspective.When the Marion 8000-Series Dragline was introduced in the early-1970s, it quickly became the industry standard for high productivity and availability. Features like outboard bearing propel, four-chord deep section boom and the over/under fairlead have proven their superiority in thousands of hours of round-the-clock operations in a variety of tough digging conditions worldwide. Marion has currently sold 31 of these 8000-Series Draglines into the Australian mining community.

Jan 1, 1992

By R Pinksen

The Iron Ore Company (IOC) of Canada Mining and Operations Departments expressed an interest in reducing the amount of oversize material (>1 m × 1 m × 1 m) being produced and its subsequent downstream effects including in-pit sorting and reduction, damage to mobile equipment, and significant delays in the loading pockets and the primary crushers. To achieve this goal an oversize reduction project was undertaken by the mine and to date has significantly improved upon baseline measures.One key approach used to reduce the amount of oversize was the implementation of thicker emulsion. The introduction of a thicker emulsion into the blastholes provided more resistance to the forces created by dynamic water and reduced the infi ltration of the emulsion into the cracks and fissures of the rock mass. Blast results showed reduced amounts of oversize and more consistent fragmentation throughout the muck piles. Additionally, a significant reduction in nitrates leached into the mine water discharge has been realised.To determine the effectiveness of the oversize reduction project, IOC used shovel based camera technology to assess fragmentation presented to the shovel at the digging face. A digital vision system (DVS) was installed on one of the shovels to capture images that were stamped with global positioning system (GPS) locations. A large population of digital images was analysed to produce a robust fragmentation distribution baseline that was used as one of the mine key performance indicators (KPI). Changes to blasting practice were subsequently measured against the baseline KPI to assess effectiveness.

Jul 11, 2011

By A Marton

Concerns with crest losses have seen changes to a number of design and operating parameters to retain the integrity of the wall and the berms of the Kidston Gold MinesÆ Eldridge Pit. Until trials began in 1999 most final wall designs utilised large diameter blastholes up to the presplit row to achieve the wall angle. The berm crests were being lost, wall support requirements were increasing and affecting mining schedules and several fault areas were putting pit access at risk. Standard designs had been determined during the early mining years and while periodic modifications had been made there was a need for major blast pattern revision. This paper traces the development of wall control blasting programs over the years, relating designs to the geology and the evidence in the walls with the theories of the day and includes the recent trials that were conducted in what is a most unforgiving geology. Trials and improvements to ensure the walls were safe were conducted even as the mine neared its economic life. Instrumented monitoring of trials blasts has greatly increased the understanding of the damage criteria and enabled rapid revision of designs. Recent changes included reduction in presplit energy, charge redistribution and use of smaller blast holes for the trim patterns. Determination of a minimum proximity-to-final wall limit for production blastholes, establishment of the correct standoff to berm crests and walls and selective use of presplitting techniques for protecting crests have resulted in cost effectiveness, improved bench crest retention and an overall reduction in visible wall damage.

Jan 1, 2001

By Y K. Ji, B J. Chung

In recent years, the utilisation of ultra-fine ore in ironmaking has gradually increased due to the depletion of high-quality iron ore. The use of ultra-fine ore in the ironmaking process is typically energy intensive and leads to high CO2 emissions. The improvement of sinter reducibility is one of the methods used to reduce the environmental impact of using fine ore.In this study, a high-strength and high-reducibility sinter manufacturing process was investigated by controlling the sintering reaction. The key technology used in this method is the manufacture of mini-pellets by selective granulation equipment. The mini-pellets consist of iron ore, limestone and burnt lime and are designed to melt at temperatures below 1200°C. The molar ratio of CaO to Fe2O3 of the mini-pellets is 0.25~0.75. The 2~5 mm diameter mini-pellets are manufactured with burnt lime as the binding material.During the sintering of mini-pellets, goethite, which has relatively high reducibility, should be retained as residual ore in the sinter cake. The strength of the sinter cake was kept by the bonding reaction of the mini-pellets, which have a low melting temperature. This is the main principle of the reaction-controlled sintering method.The reaction-controlled sintering method manufactures a high-strength and high-reducibility sinter cake and increases sinter productivity by around 15 per cent using ultra-fine iron ore.CITATION:Chung, B J and Ji, Y K, 2015. Production of high-quality sinter by controlling the sintering reaction, in Proceedings Iron Ore 2015, pp 21–28 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Jul 13, 2015

By E McLean

The decision making check-list and hierarchy of information for process and plant design is generally wide-ranging and complex. The nature of grades, mineralisation and ore types in an ore body, its mine development sequence, and the body of results from metallurgical and physical test work present complex and competing influences on flowsheet development, values for design, equipment selection, plant efficiencies and operating strategy. The challenge is how to manage all these variables, to understand their impacts on performance, and how best to provide a cost-effective plant design and robust process that meets daily/weekly/monthly operating objectives. Throughput, production and grade or product specifications need to be met. The ore body is not average; its range of metallurgical and physical characteristics may be moderately to highly variable. This paper assesses some of the key competing influences and conflicting issues in design and equipment selection, and their effect on production and performance expectations.

Aug 8, 2011

By H Wright

Black Rock Cave (BRC) is a sublevel cave which was developed in 2018–2020 and has been in production since November 2020. BRC is part of the Mount Isa Mines complex and extracts a leached orebody comprising of high-grade chalcocite ore, situated below the historic Black Rock Open Cut workings. The mine is owned by Mount Isa Mines and operated under contract with PYBAR Mining Services. In 2021 a number of rockfalls occurred throughout the mine and were increasing in frequency. This culminated in a major rockfall on the decline in June 2021, after which the decision was made to withdraw all persons from the mine, initiate rehabilitation and conduct further investigations to deduce a safe method to provide long-term support. The rockfall investigations have shown that the corrosive nature of the highly leached rock mass can consume a split set in less than 18 months. A review of potential corrosion resistant ground support methods concluded that steel split sets have a safe lifespan of six months at BRC, and that all workings must have corrosion resistant support installed prior to exceeding the safe life of the friction bolts. The review recommended the use of 4 m cable bolts as the most suitable method of corrosion resistant support. This new support system, and the requirement to rehabilitate all mine workings within the leached orebody, had an impact to the schedule and an effect on both equipment and manning levels at BRC. Accelerated jumbo rehab allowed access to the production levels within two months of the rockfall. Cable bolting then followed the jumbo bolting in all leached areas. The direct impact has been ten months of reduced production, due to night shift closures of the decline to enable cable bolting. The impact to the mine has been minimised by the small flexible team that enabled the mine plan to change rapidly and often to accommodate the challenging conditions. In less than two months from the rockfall event, production had resumed. Within ten months of concluding that the mine required; full rehabilitation, a new support system and a full re-schedule, normal mining operations had restarted, with pre-event productions rates being achieved.

Mar 29, 2023

By A L. Randriamanjatosoa, G Small, C Greet

In flotation simulation models, the effects of pulp chemistry other than pH are often ignored and reduced to particles hydrophobicity or contact angle, neither of which can be measured readily in real time in an operating plant. The flotation response of a given sulphide mineral is known to be influenced and can be improved by changing the chemical characteristics of the pulp during grinding and flotation. The on-line measurement of pulp pH, potential, dissolved oxygen, temperature and oxygen demand of a process stream was completed with the PCM® (Pulp Chemistry Monitor) developed by Magotteaux in this study. The data generated together with relevant plant operating parameters were used to construct an empirical model and simulate the metallurgical performance of the ore. The same approach was also applied to laboratory flotation tests data using various grinding media types. The flotation models better fit the actual data when the chemistry of the pulp is taken into account.CITATION:Randriamanjatosoa, A L, Greet, C and Small, G, 2018. Effects of pulp chemistry on plant and laboratory flotation modelling, in Proceedings 14th AusIMM Mill Operators' Conference 2018, pp 423–432 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Aug 29, 2018

By S Gillies, H W. Wu

A study has been undertaken into mine ventilation systems currently in use within Australian modern coal longwall (LW) extraction mines. It reviews systems and discusses evolving changes being adopted to address the more complex challenges. There is a strong move to longer panels, wider faces, greater extraction heights, increased production rates, more efficient ventilation and decreased personnel. Each of these changes is motivated by improvement in productivity and generally leads to cost reduction. In addition, mine workings are moving deeper, which results in increased ventilation control issues such as higher total and respirable dust levels, greater seam gas contents in parallel with lower in situ permeabilities, spontaneous combustion and heat management issues. A comprehensive database of Australian LW ventilation practice is analysed. Currently there are a variety of LW panel ventilation circuits used in Australian underground coal mines due to various combinations of seam characteristics, gas emission rates, spontaneous combustion, geological features and surface constraints. The main issues usually addressed in the designing and planning of ventilation circuits for LW panels are airway velocity, gas concentrations, LW cutting methods (eg Bi-di, Uni-di or half web), ventilation of control devices, pressure differentials and leakage paths and understanding gas concentrations across the length and width of the goaf. If the ventilation circuit can manage the applied contaminant load (gases, heat and dust) at an acceptable cost and circuit duty, then supplementary controls, such as gas drainage, refrigeration and dust sprays and scrubbers, may not be required. CITATION:Gillies, S and Wu, H W, 2013. Review of Australian longwall panel ventilation practices, in Proceedings The Australian Mine Ventilation Conference , pp 87-94 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Jul 1, 2013

FOR some reason there is not any extensive literature in print on the electrolytic refining of copper, although the method is in general use in America and Europe. In Australia there are but three plants in existence one small one at Wallaroo, in SouthAustralia; one at Lithgow and one at Port Kembla the last two being in New South Wales. The largest of these is the Port Kembla plant. It is also the most modern, and embodies all the latest appliances and improvements of electrolytic practice.The plant and methods about to be described is that at Lithgow, with which the writer was associated for some years, and the experience gained there may be of interest. Electrolytic refining of copper is one of the simplest operations of metallurgy, but, like all metallurgical operations, it requires constant attention and a large amount of common sense. The man in charge should be endowed with an insatiable curiosity as to why certain things are occurring, and an instinct of cleanliness, whichmy first mentor in chemistry, Mr W A Dixon, once impressed on me was better than godliness.

Jan 1, 1911