THE METRIC SYSTEM AND THE ENGINEER

By J L. WHITWELL (Visitor)

ABSTRACT

A brief history of the metric system from its inception to the acceptance of the Systeme International. The reasons given by the Metric System Committee of the South African Bureau of Standards for considering a change from the foot pound system of weights and measures. Some engineering problems which will arise with a change of unit and the change of size of raw material, semi-processed material, and manufactured components. How manufacturing machinery is involved. Possible solutions are suggested. 

MR J. L. WHITWELL, C.Eng., A.M. I.E. E., A.M.(S.A.) I.E.E., was born on 10 October 1919, in Ulverston, Lancashire, England. He was educated at Ulverston Grammar School and Liverpool University. He joined the British Post Office as Traffic Superintendent (Telephones) in 1939 and served in the armed forces from 1940 to 1946 as a major in the Royal Signals. He went to East Africa as an executive engineer in the East African Posts and Telegraphs in 1948 and served in Kenya, Tanganyika, and Uganda, becoming Regional Director of Uganda. He left East Africa at the end of 1961 and came to South Africa in 1962, when he joined the South African Bureau of Standards. He was appointed Technical Secretary of the Metric System Committee. Mr Whitwell was elected a Fellow of the Royal Statistical Society in 1951. 

1. INTRODUCTION

On May 20, 1965, the Report of the Metric System Committee, appointed by the Council of the South African Bureau of Standards, was tabled in Parliament by the Honourable the Minister for Economic Affairs. The Report, which recommended the early adoption of the metric system of weights and measures in the Republic, is being considered by the Government.

The implications of a change of units for weighing and measuring are wide and this paper deals only with some of the aspects affecting the engineering profession.

2. THE METRIC SYSTEM

In 1790, the government of the French Revolution instructed the French Academy of Sciences to submit proposals for the replacement of the chaotic multiplicity of units in use in France by a single system of weights and measures. The French scientists decided that such a system should not be based on man-made arbitrary reference standards, which might deteriorate or be lost or damaged, but that it should be based on permanent measures provided by Nature. The unit of length was chosen as one ten-millionth part of the distance from the pole to the equator along the meridian passing through Paris and was called the metre.

The unit of mass was to be the mass of a cube of distilled water of a given length of side at normal atmospheric pressure and at its temperature of maximum density. A cubic metre of water was too large a unit for practical purposes. It was decided that the unit of mass would be a cubic centimetre of water at 4°C and at a pressure of 760 millimetres of mercury and was called the gramme.

About time, it was first proposed to decimalize the day, but eventually the traditional second was adopted and defined as 1/86400 of the mean solar day.

The French scientists proposed that all other units should be derived from the three fundamental units the meter, the gramme, and the second. They also proposed that, except for time, all multiples and subdivisions of basic units should be in accordance with the decimal system, and that the decade multiples would be indicated by prefixes based on Greek words, and the decimal subdivisions by prefixes based on Latin words.

These prefixes  are as follows:

 Since then, the range has been extended upwards by the addition of giga (G) (109) and tera (T) (1012) and downwards by the addition of nano (n) (10-9), pico (p) (10-12), femto (f) (10-15) and atto (a) (10-18).

The proposals of the French Academy of Science were eventually approved by the French Government and the metric system was decreed by law in 1795, and the use of all other systems banned in 1837.

The developments in France aroused great interest in other countries, resulting in an international convention being held. On May 20th, 1875, seventeen countries, including Great Britain and the U.S.A., signed the "Metre Convention." Most of the signatories proceeded to make the metric system the only legal one for their country, but Great Britain and the U.S.A. would not go further than to accord it legal recognition only.

An international committee-the Cornite International des Poids er Mesures (CIPM) was created, under whose aegis was established the Bureau International des Poids et Mesures (BIPM), a permanent body with offices at Sevres, near Paris, but with extraterritorial status. The Bureau is responsible for the laboratory work in connection with, and for the safe keeping of the international reference 'etalons' or standards.

Although the cubic metre of water was too large a unit, for practical purposes the cubic centimetre of water was too small for use as a standard. The CIPM decided to make the standard of mass one cubic decimetre of water at 4°C and normal atmospheric pressure, i.e., one thousand grammes or one kilogram me.

In other words, the first internationally agreed fundamental units were those of the present international M.K.S. system.

2.1 M.T.S. and C.G.S. systems

What should have become the M.K.S. system was unfortunately displaced by two other systems-the metric technical system and the C.C.S. system. The metric technical system used the metre, the kilo-gramme-force, and the second as fundamental units. The disadvantage which this system was found to have was that the definition of the kilogramme-force required an assumption of a value for 'g,' the acceleration due to gravity. The value of 's' assumed was 9.80665 m. s2 which was believed to be the value at sea level and 45° latitude. The metric technical system therefore suffered from the disadvantage that one of its fundamental units could only be used after allowance for variations in 'g.'

The C.G.S. system, proposed by the British Association for the Advancement of Science, uses the centimetre and the gramme as the fundamental units of length and mass, and was adopted at the first International Electrical Congress in 1881. It has been, until recently, firmly established in all branches of physics.

2.2 M.K.S system

In 1902 an Italian engineer, Giovanni Giorgi, suggested that the practical units of current, voltage, energy, and power (ampere, volt, joule, and watt) fitted in exactly with the metre, kilogramme (as the unit of mass) and the second. He further pointed out that the metre, kilogramme, and second formed a coherent absolute system of units covering mechanical as well as electrical phenomena. This system is referred to as the M.K.S. system.

In this connection, a coherent system of units is one in which the quotient or product of any two-unit quantities in the system leads to the unit of the resultant quantity. For example, in a coherent system unit velocity results when unit length is divided by unit time, and unit force results when unit mass is multiplied by unit acceleration.

The advantages I am having such a universal practical system was well appreciated, but the deep entrenchment of the metric technical and the C.G.S. systems slowed down the adoption of the M.K.S. system and it was only in 1954 that the tenth conference of the 'Conference Générale des Poids et Measures' (CGPM) adopted the M.K.S. system. In 1960, the eleventh CGPM formalized the system and gave it the name of 'Systeme International d'Unites' (abbreviated as 'S. I'). In addition to the metre, kilogramme and second, three additional basic units were considered necessary-the ampere for electrical and magnetic phenomena, the degree Kelvin for thermodynamics and the candela for visible radiation.

The International Standards Organization (ISO) has adopted the S.I. and has issued a series of recommendations for units to be used for such subjects as radiation, acoustics, electricity and magnetism, nuclear physics, chemistry, heat, mechanics, space and time, and periodic and aperiodic phenomena.

In the S.I, there are thus six basic units which are as bellows:

All other units of the S.I. are derived from the foregoing basic units, some of which are listed in Appendix B.

The derived units are self-explanatory, but I draw attention to the unit of force-the newton-and to the unit of pressure-the newton per square metre. The newton is equivalent to about 102 grams weight at the earth's surface.

3. REASONS FOR CONTEMPLATING A CHANGE

The Metric System Committee gave the following reasons, inter alia, for considering a change to the metric system of weights and measures in South Africa:

(a) The necessity for South Africa to look for future markets for its industrial products. These markets would be metric countries, and therefore South African manufacturers would have to produce metricized products.

(b) The simplicity of the metric system compared to the inch-pound system would simplify calculation, and more efficient use of available manpower would be made.

(c) Time would be saved in education and technical training, and this time could be used to raise the standards.

Since the publication of the Committee's Report, the Government of the United Kingdom, South Africa's major trading partner, has decided to adopt the metric system in industry, and firm steps are being taken to make this decision effective.

4. CHANGING TO THE METRIC SYSTEM

Should the decision be taken for South Africa to adopt the metric system of weight and measures the Systeme International would be used, because this system is rapidly replacing the older C.G.S. and other metric systems for expression of international standards.

A clear distinction must be made between units and sizes. Although the metric units employed have been internationally accepted, sizes have not. As a result of international co-operation in such bodies as the International Standards Organization, the International Electrotechnical Commission, the Common Market, the European Steel and Coal Community, a large amount of agreement and standardization has, however, been attained. The units in which this standardization is expressed are those of the metric system.

British and American standards are mostly expressed in terms of the foot-pound-second system, although there has been for some years a clearly defined trend on the part of both the British Standards Institution and the American Society for Testing Materials to use metric units as well as the foot-pound units. The United Kingdom has already stated its intention to adopt the metric system in industry over a period of from 10 to 15 years and has embarked on a revision of all its standard sizes with a view to bringing them into line with those of the metric-using countries.

As with the basic units, the metric standard sizes have been logically developed instead of having just grown, as did the British and American standards. The use of preferred numbers in conjunction with metric units has provided an excellent combination in deciding the range and magnitude of sizes.

Metric standard sizes are arranged so that the intervals between adjacent sizes increase in a definite proportion as the sizes increase. Although the intervals between sizes in the Birmingham gauge system also increase with size, the increases follow no definite pattern. A comparison between the Birmingham gauge sizes for sheet and corresponding ISO standard sizes is shown in Appendix A.

From the engineering point of view there are two distinct aspects of a change-over to the metric system -a change of units and a change of standard sizes.

Before a change of units can be contemplated there must be a declaration as to which units are to be adopted. In view of the general international acceptance in principle of the Systeme International as the ultimate system of units, it is suggested that South Africa would have no alternative but to declare such system as the one to be used in the Republic. A note of caution is necessary here, however, in that several metric units are extremely well entrenched and the introduction of new units which are not yet used widely by industry in metric countries must be done in step with these countries. An example is the S.I. unit of pressure-newton per square metre. This is not widely used in industry, despite its logical raison d'etre, and the kilogram me per square centimetre persists.

The change-over would take place in two stages:

Stage L-Existing sizes would remain the same but would be expressed, for example, in centimetres or millimetres instead of inches.

Stage 2.-Sizes would be altered to conform to international metric standard sizes or to sizes expressed as whole numbers of centimetres or millimetres.

5. CHANGING UNITS

In Stage 1, all that would be required would be straightforward conversion from inches to metric units, using conversion tables and applying the rules regarding accuracy and tolerances. For example, a tube with an outside diameter of 5/8 inch would be designated as 15.8 mm or 15.88 mm (depending on the accuracy of manufacture) and a tube with a diameter of 2 inches as 50.8 mm (l inch=25.4 mm exactly).

Such a change could be started immediately, and the Metric System Committee considered that, with the co-operation of all concerned, a period of three years would be sufficient for all articles to be re-designated in the new units.

6. CHANGING SIZES

For many goods of a simple nature, particularly where precise measurement is not important or where manufacturing tolerances are wide, Stage 1 could be dispensed with and the articles produced in the nearest whole centimetre or millimetre size.

Where greater precision is required or where the product consists of several inter-related components, the problem of changing the sizes is more complex.

Industrial production can be regarded as a chain:

(a) Raw materials (ores, pig-iron, ingots, etc.).

(b) Semi-processed materials (metal sheet, rods, timber, etc.).

(c) Manufactured components (rivets, bolts, nails, pressings, etc.).

(d) Assembled products.

A change in size of a product in one link of the chain would, in general, must be preceded by changes in the earlier links. Such changes would be more significant in the later links. For instance, the size of ingot from which rod of a given size is made is not of importance (link (a)-(b)); the size of rod from which a component is made is of greater importance but not vitally so (link (b)-(c)). The sizes of the components of an assembled article are important in that they must fit in with other components (link (c)-(d)):

It would be of no use for a manufacturer to attempt to change his assembled product (with its components) to accepted metric sizes unless he can assure that metric components are available from his suppliers. Although the change of sizes must therefore start at the beginning of the chain of production, such a change would be triggered off by demand from the end of the chain, i.e., at the factory which uses the components and by the purchasers of the product. As in the case of Stage 1, the active participation of Government and associated organizations would be essential, particularly by revising specifications in terms of standardized metric sizes.

If the changeover is to be affected with the minimum of inconvenience, a co-ordinated plan will be necessary, and the co-operation of industry will be vital. This is where engineers will play a leading role.

It is of interest to note that British industry, having proposed to the United Kingdom Government that the metric system be adopted and having had the proposal accepted, is giving the utmost support to the measures which are being taken.

6.1 Raw materials

The size of raw materials (ore, ingots, etc.) will not be seriously affected by a change to the metric system. Such materials are usually sold by weight, and there should be no difficulty in expressing weight in kilograms or metric tonnes instead of pounds or tons.

Ingot size is not of particular importance as products produced from the ingot usually bear no relation to either the shape or size of the ingot. The size of ingots could therefore be left as they are but expressed in whole centimetre or millimetre units without exceeding existing tolerances.

6.2 Semi-processed materials

6.2.1 Gauges. The use of so many gauge systems at present is often quoted as one of the drawbacks of the non-metric system. The gauge systems in use in this country include:

French Wire Gauge,

Birmingham Gauge (Sheet),

Galvanized Sheet Gauge (USA),

Zinc Gauge,

American Wire (Brown and Sharpe) Gauge,

Steel Wire Gauge (Washburn and Moen) (USA),

Roebling Wire Gauge (USA),

British Standard Wire Gauge (also used for copper, aluminium, and stainless-steel sheets),

Music or Piano Wire Gauge,

Birmingham (Stubbs) Iron Wire Gauge,

Stubbs Steel Wire Gauge,

Whitworth Gauge (wire and sheet), and

Lancashire Gauge (wire and sheet).

In many industries confusion arises between purchaser and manufacturer because unless the gauge system is specified, a quoted gauge number may mean different sizes.

Should the metric system be adopted, the International Metric Series for thicknesses of sheet and diameters of wire would be used. These sizes, based on preferred numbers, are expressed directly in millimetres, and cannot therefore be misinterpreted. The necessity to refer to tables to convert from a gauge number to an actual size, now so often the case, would not arise.

6.2.2 Rolled products. The dimensions (and therefore the strengths) of such rolled steel structural sections as beams, joists, T-bars, channels, and angles, would ultimately be produced to European (Euro norm) standards. Close liaison would be necessary between the building and basic metals industries so that the timing of a changeover could be mutually agreed upon.

Where rolled products such as bars or rods are to be further processed, the sizes of the products will ultimately be made to standard metric sizes. If there is to be an interim period during which the sizes will continue as now, the differences in sizes between the inch system and the metric system are not significant. For instance, a shaft that must have a diameter of 50 mm can be turned equally well from a steel rod with a diameter of 2 1/8 inch or of 54 mm.

In the case of plate, sheet, rod, and wire, the thicknesses and diameters of the metric standards are closely related to those of the inch system (Birmingham Gauge and Standard Wire Gauge) and in most cases are within the manufacturing tolerances already allowed. (Appendix A.)

The adjustment of machinery to allow for changes in thickness of flat stock (plate, strip, and sheet) involves only a change in the spacing between the rollers, an adjustment which is readily available in most cases. Where round or square stock is being rolled such adjustment would not be possible. The life of the rollers is short, however, and they would be used until they are worn and then replaced or re-machined and the grooves of the finishing passes adapted to metric sizes.

6.2.3 Drawn products. The change from standard inch sizes to standard metric sizes of drawn products would be of the same order of magnitude as that of rolled products.

Machinery producing wire or other drawn products can be easily adapted to metric sizes because only the dies need to be replaced. Dies have a comparatively short life and the timing of the change would be decided in consultation with the users of the products.

6.3 Manufactured components

Components may be regarded as either specialized and peculiar to one assembled product or as standardized and available for use in any assembled product

Changes of size of specialized components would be initiated by the manufacturer assembling them. In many cases, however, the sizes are fixed by utilitarian considerations and would not require alteration. For example, the component parts of a chair, the sizes of which correspond to those of the human body, would be the same whether the chair is to be used by a Frenchman or a South African, and there would be no need to change those sizes, although the units in which they are expressed would be changed from inches to centimetres.

The problem of standardized components is more complex, particularly about fasteners such as bolts, nuts, screws, and rivets.

Much plant, machinery, and many motor vehicles in use in this country are already made to metric standards and although bolts, screws, etc., are available in metric sizes, most fasteners are non-metric. However, the demand for metric standard fasteners is increasing rapidly. To illustrate this, the ratio of the sales of metric to non-metric spanners in South Africa is approximately 4: 3 at present.

The Chairman of the General Council of the British Standards Institution stated in 1965 that Whitworth, BSF, and BA threads were no longer acceptable by the world in general and that British trade was suffering as a result. It is estimated that although stocks of these threads will have to be maintained for the servicing of existing equipment, they will become obsolete in about ten years. One of the first moves in a change to the metric system would be the adoption of ISO metric screw threads and sizes, as has already been initiated in the United Kingdom.

In general terms, the modern machine tool allows production in either the metric or the inch system no matter whether it is made according to metric or inch design. There are few manufactured components the size of which cannot therefore be increased or decreased to the nearest metric size without major adjustment of machinery.

Where machines are calibrated in inch units-for instance, lathes, surface grinders, milling machines, universal grinders-it has been suggested that the choice between whether to alter the calibration of lead screws and use the direct centimetre-millimetre measurement or to continue using the inch calibrations and employing conversion tables should be left to the individual factories concerned.

In this connection it is of interest to note that many machines in use in South Africa have been manufactured in Europe and have been modified from metric to inch calibrations. Conversely, American, and British-made machines are usually capable of easy conversion from inch to metric calibrations.

7. ENGINEERING EDUCATION

Engineers of professional level should find little difficulty in understanding the new units. Many are already familiar with them. It will, however, be necessary for some, if not the majority, to readjust their standards, based perhaps on many years of experience in the foot-pound system. For instance, a structural engineer knows from experience what load a 12-inch x 10-inch joist can take, but he would have to think a little harder when confronted with a choice of sizes based on centimetres. It would not be long before the engineer's judgment would be as certain in the new system as it is in the old.

The training of artisans is another problem which will have to be faced. From discussions with firms employing technicians and artisans from Europe, it was found that their employees rapidly assimilated the inch-pound system with its complexities. There is every reason to believe that South African artisans will easily change from the inch-pound system to the simpler metric system.

In schools and technical colleges, teaching of the metric system will take preference over the inch pound system and will, after a few years, replace the latter system entirely.

8. CONCLUSION

In making a change of units of weights and measures there are many aspects to be considered such as education, commerce, public reaction, public services, surveying, aviation, legal and several others which it is not possible to discuss in a paper of this nature. Some of the aspects concentrated on are those which directly concern engineers. Everyone will be involved in varying degrees in such a change-perhaps only in their private lives where they will buy petrol by the litre instead of by the gallon, but most of them will face engineering problems of one kind or another. The engineering profession will solve these problems and will thereby reap the benefits of a simpler, more logical, and labour-saving system of design and calculation, and will be able to provide engineering services, vital to the country, with even greater efficiency.

BIBLIOGRAPHY

1. Report of the Metric System Committee - S.A. Bureau of Standards. Pretoria, May 20, 1965.

2. Engineering Units-H. S. Hvistendahl-Macmillan and Co. Ltd., London.

3. The International System CSI) Units - British Standards Institution. London CBS 3763:1964).

DISCUSSION

Professor S. Smoleniec (Visitor): The Systeme International of units of measurement has so many indisputable advantages that its universal adoption would be of great benefit to all concerned.

In the Systeme International all the units are systematic and by this virtue they facilitate even the most complex computations by complete freedom from conversion factors. Furthermore, they are based on the internationally accepted standards, which are stable and reproducible to a high degree of accuracy. This system, prior to 1960 known as the M.K.S. system, has already rationalized practical electrical units and replaced entirely the three previously used systems, namely, the C.G.S. electrostatic, electromagnetic, and Gaussian systems. However, as Mr Whitwell rightly stated, the S.I. units are universally acceptable, but their sizes are not.

Two -years ago I was requested by an overseas publishing house to suggest the most appropriate metric units for their forthcoming edition of Thermo-dynamic Tables. It has been apparent from the very start that the S.I. units are either too small or too large for tabulation purposes. For example, the value of specific enthalpy of saturated steam at atmospheric pressure presented a formidable figure at 2 675 800 J/kg, whereas the dynamic viscosity of air at 3 000K had to be expressed as a minute fraction, namely 0.00001846 Ns/m2. When a derived unit is given a name, such as joule, the addition of an acceptable prefix, e.g., kilojoule, is both logical and convenient. It is not customary, however, to add a prefix to a derived unit of a composite dimension such as Ns/m2 because it may suggest that the prefix applies to the first term, i.e., to newton, and not to the unit of dynamic viscosity. For this reason, the French proposal to introduce a new unit of dynamic viscosity, the Poiseuille (PI = 1 Ns/m2), appears attractive, because it would justify the use of the conventional prefixes such as micro for tabulation purposes.

Although increasing, or decreasing, the size of a unit by means of a decimal multiple should not present any difficulty to an engineer familiar with slide rule calculations, its use should not be restricted only to those powers of ten which are included in the conventional prefixes. Thus, it is more convenient to name N/m2 the pascal (1 Pa = 1 Ns/m2), and to use a suitable prefix for a unit of pressure, stress, or tensile strength, such as Megapascal, than to adopt the unit called bar equal to 105 1 Ns/m2. However closely the size of bar may approximate one standard atmosphere the power index of five is not included in the list of conventional prefixes. Furthermore, this unit has been traditionally restricted to the meteorological field and there is an inherent resistance to expressing stress and tensile strength in bars for engineering purposes.

It is not surprising, therefore, that despite unquestionable advantages of the Systeme International even the most ardent supporters of metric units, such as U.S.S.R., are reluctant to legalise this system without reservation.

Yet it must be admitted that systemization of units on the international level is the only solution to this difficult problem. The Systeme International units of measurements is the right move in this direction. New derived units, based on the six fundamentals S.I. units, and whose sizes would be generally acceptable, can be found. The benefit by far outweighs all objections of a justifiable nature. After all, we have no better choice. The system which we are using at present was described by Lord Kelvin as the "absurd, ridiculous, time-wasting, brain destroying British system of weights and measures."

Mr R. von Vivenot (Visitor): The formation of larger areas of national economies and the continually closer interlacing of engineering science makes existing obstructions in the interchange of commercial goods, caused by different basic units appear more conspicuously than heretofore. Therefore, it is welcome that this problem is being taken up to arrive at a solution as speedily as possible.

As representant of the German industry, which in 1875 decided in favour of the metric system within the framework of the treaty of the international metric convention, we naturally favour an early change by South Africa to the metric system. The question which South Africa must now decide is, whether its future chances on the world markets are to be encumbered by a double-track system of weights and measures and whether one would agree to burden the national economy with two systems, together with the increased expense coupled therewith. In our opinion there is only one solution to this question, which is to give preference to the more modern and better system.

We share the opinion of the author, who shows the way to introduce the metric system. Based on experiences which the German industry has gathered during discussions regarding the introduction of the metric system in other countries, may we point out also the following items:

1. When changing to the metric system the problem is not solved by expressing a certain numerical value by another one, e.g., 10 inches by 254 millimetres. A change of this sort, which will certainly be necessary during the first phase, does not render any help whatsoever for the export of industrial products manufactured according to the metric system. It is only during the second phase that the change becomes interesting economically, i.e., when the figures of the internationally preferred series are adopted, upon which also are based the ISO and IEC recommendations. Consequently, and indeed for this reason, the attempt, prepared by extensive preliminary studies, should be made to start the second phase as soon as possible.

2. The introduction of the metric system in South Africa should not constitute a juridical problem since she belongs to the signatory States of the Metric Convention whereby the metric system is optionally valid alongside other units. It is therefore not a question of introducing or permitting the metric system in South Africa, but of instituting the metric system as obligatory by law for the whole of the national economy.

3. The introduction of the metric system in standardization has never been a question of timing but rather one of mental attitude. Should the question of such a change be seriously considered industry or other competent authorities should prepare the ground for a satisfactory acceptance of the metric system without delay. This could be done by a working committee who would co-ordinate all transactions to be undertaken.

Normally, establishing a new standard requires three years; the life span of a standard until its next change is five years. Considering these figures, and purely as a statement, a fluid change to the metric system is possible, more so since at the most the life span of one generation suffices as a changeover period.

4. Naturally, the cost of such a change-over is enormous. The main portion of the cost of the changeover, to be borne by the State, would be required for propaganda and the change-over in public spheres. We consider the preparatory means to be provided are: Tables of exchange for measures of length, area, and volume, for weights and measures. These tables could be issued as booklets or, for handy use in workshops, in the shape of slide rules. Change-over costs [or industry, who will eventually also profit by the change, would not be so considerable since factory and similar installations undergo a normal wear and tear and must be renewed sooner or later.

German industry is convinced that the introduction of the metric system in almost all countries having a customs system based on inches and pounds is only a matter of time. According to evidence supplied by the International Bureau for weights and measures, the situation is as follows:

The units of metric system have been introduced by law or legal ordinance in 100 independent states. In 86 of these states the metric system is obligatory, i.e., legally binding, and in the other 14 states it is optionally valid alongside of the other units. Of these 100 states 40 are at present signatory states of the metric convention. Ten of the Commonwealth states also figure in these tables. Of these, Australia, Great Britain, India, and Canada are signatory states of the metric convention. While in India and Malta metric units are obligatory, the remaining eight states (Australia, Great Britain, Canada, Kenya, Malawi, New Zealand, and Uganda) have introduced it on a facultative basis. Of the further 11 Commonwealth nations not figuring in the list, the four states of Ceylon, Malaysia, Nigeria, and Pakistan are considering joining the metric convention. Of other states, Afghanistan, Saudi Arabia, and Nepal are also entering into negotiations to join the metric convention.

Mr T. O'D. Duggan (Visitor): I am, I suppose, generally in favour of the introduction of the metric system, if only because it is inevitable and there is no point in opposing what is inevitable.

That there are advantages to be gained from the use of a single, world-wide system of units can certainly not be denied. Of the available systems the metric system is the most systematic and therefore the best choice.

Mr Whitwell has dealt at some length with the question "which metric system?" and there seems no doubt that the Systeme International is the one to accept.

I think that, in considering this matter, we should be clear exactly what it is ware after. Too often the proponents of the metric system overstate their case. We should, for example, be careful to distinguish between standardization, rationalization, decimalization, and metrification.

The systems of wire and sheet gauges currently in use in those countries using the inch-lb system are all standardized-but there are too many of them!

What is required here is rationalization, i.e., the reduction of the number of systems to only those that are necessary. Better still "gauges" as such should be done away with and sheet or wire sizes given in actual dimensions.

Decimalization, or the expression of quantities, particularly fractional quantities, in decimal notation is not confined to the metric system. In precise manufacture, for instance, the decimal inch has been used at least since the time of Whitworth, so much so in fact that the "thou" is practically a unit.

In all human progress one can note the same trends: simple beginnings leading to greater and greater diversification and complexity, and then, ultimately, a rather rapid simplification. One has only to compare the clumsy and complex numerical systems of the Babylonians, Egyptians, Greeks, or Romans with our present system to see the truth of this.

So, having developed many useful systems along individual and often independent lines, the time has arrived for us to simplify, rationalize, standardize, and decimalize and settle on one system.

It has always been man's desire to have natural standards for all fundamental units. The natural standards in vogue at any time have been many and various. As standards of length, for example, we find that men have used the length of a barleycorn, of a foot, a pace, a thumb joint, (he arms from armpit to fingertip, the seconds pendulum. We even find that at one time it was intended to use the ten millionth part of the earth's quadrant through Paris!

None of these, though often satisfactory for the unsophisticated society of their day, have met the fundamental requirements of a standard, viz.:

(i) that it be immutable,

(ii) that it be convenient in use,

(iii) that it be universally available.

None of the early standards was immutable and the length of the meridian through Paris was unattainable. The metre is, therefore, not a natural standard it is entirely arbitrary as is the Imperial standard yard.

It is only recently that we have standardized on what appears to be an acceptable natural standard, a particular wavelength of Krypton 86, and the length of the metre and the yard expressed in terms of this wavelength, as Mr Whitwell has told us. As you will remember, the metre is now defined at 1 650763.73 wavelengths of a particular transition of the Krypton 86 atom.

In passing it would seem much more rational to scrap both the metre and the yard in favour of a new unit of length equal to, say, one million wavelengths of Krypton 86!

The unit of mass is also quite arbitrary while nothing could be more unavailable than the unit of time a fantastic fraction of the tropical year at the beginning of the twentieth century.

Because of the unsatisfactory state of the standard of time there are, as you will know, various proposals being examined to replace our present standard with one based on naturally occurring frequencies within the atom or nucleus.

It is often claimed that the metric system will eliminate the need for all sorts of constants and such peculiarities as 12 to the foot, but 16 to the pound and 112 to the hundredweight, and permit calculation merely by shifting the decimal point. This is to a certain extent true, but, unfortunately, the universe was not built to a metric module so that physical constants will remain with us - g for instance.

To offset any simplification that may result we shall now have to memorize a new hierarchy starting with atta, femto and pico, through deca, hekto and kilo and ending up with giga and tera. If the Greeks can do it so can we, I suppose!

Be this as it may, it is certainly time that we had a unified system of units and standardized sizes of commonly used articles and materials. The longer we delay in reaching agreement on this the more it is going to cost.

The conversion is, in any case, I think, going to cost much more than many imagine even if the introduction of the new system is gradual. The two systems will have to live side by side for many a year. It will in most cases be economical for only new manufacture to be in terms of the metric system. The manufacture of spares and replacement parts for existing machinery and plant will almost invariably have to remain in the original units because of the high cost of converting all the drawings.

Apart from the cost factor, the only serious difficulty would appear to arise from the necessity of becoming accustomed to new standard units, names, and sizes.

Apart from this difficulty, becoming used to new units is a real problem. I see no reason why we engineer, handbook in one hand and slide rule in the other, cannot weather the storm and turn it to advantage.

Mr A. H. Waites (Visitor): At Crown Mines, Johannesburg, there is a light to medium general engineering machine shop where equipment having non-friction bearings as components is manufactured. In the main, these bearings are of metric dimension, but due to the peculiarity of the present system of measuring, designs and drawings must be converted to the inch system. As a result, we have for quite an appreciable time considered changing over to the metric system. To summarize, the main points to be taken into consideration are:

1. Education of personnel.

2. Conversion of existing equipment:

(i) Machines,

(ii) Gauges.

3. Purchase of new measuring equipment.

About (1) we foresee no difficulties worth mentioning. The South African artisan will have no difficulty in adapting himself to the new system.

However, about (2) and (3) the cost could be appreciable, and it will be interesting to know whether this cost will be subsidized in some way or other.

Mr E. M. P. Evans (Member): Whether they agree with his paper, everybody will be grateful to Mr Whitwell for defining those multiples and sub-multiples of metric units which are not mentioned in schoolbooks or engineering pocketbooks yet keep appearing in engineering and scientific publications, such as giga, atto, tera, nano, etc., and have often puzzled readers. It might be a good idea to print this list at the end of the Journal with the Iist of standard units.

The argument for the Metric System really covers two different things-the decimal system and the metre. The only reason we use the decimal scale of notation at all is that God has given us five fingers on each hand. If we had had four or six fingers, we should be using the occasional or duodecimal system of numbers. The Maya of Central America, who were advanced in astronomy though lacking metals, counted by twenties, which suggests that they did not wear boots. But ten is an awkward number because its only factors are 2 and 5, whereas if we use 12 as a basis, we have the factors 2, 3, 4, and 6. Ordinary people long ago discovered that it is more convenient to pack eggs or bottles in three rows of four than in two rows of five, and so invented the dozen, which is now to be forced out of use. It would have been scientifically better if the metric system had been coupled with a duodecimal scale of notation.

It is true that it is easier to multiply and divide by ten than by other numbers, but a person who has learned to do only this is at a disadvantage when meeting other arithmetical problems. The French Revolutionists were so infected with a passion for decimals that they put ten days to a week. This gave workers a rest of only one day in ten instead of one in seven. It was never popular and was abolished by Napoleon. "These things are not settled by philosophers, but by the man in the dirty shirt." They also decimalized the right angle and the degree, and these horrors have infested trigonometrical textbooks until recently.

The base of the Metric System is the metre, which was intended to be related to the size of the earth and so be an unalterable fixed dimension. But there were errors in the original measurement. A recent survey textbook devoted a chapter to its various values, and it remained an arbitrary length until defined by the spectrum.

Much of the elaborate nomenclature for the multiples and divisions of the basic units is of no practical use. Only a schoolboy uses decametres or decimetres; it is much easier to say ten metres or ten centimetres. It is not true that the metric system prevents mistakes in reading dimensions. The celebrated English journal, Engineering, in reproducing drawings of prefabricated concrete flats in Warsaw, misread millimetres as centimetres and showed rooms 80 ft high. One argument for the adoption of the metric system is that the rest of the world uses it. But they have been slow to change over completely. The Turkish Army in Palestine in 1917 was using pipe fittings marked in inches, and recent German machine drawings showed Whitworth bolts in machines dimensioned in metres.

An ideal table of standard bolts, plates, or steel sections should have the sizes arranged in a geometrical not an arithmetical progression. The author claims that metric standard sizes are arranged so that the intervals between adjacent sizes increase in a definite proportion as the sizes increase, but the only such standard, scientifically designed, is that for British Association screws, which is to be abolished. The table of metric rolled steel joists is far less practical than the British. In the former table there are 23 sections with depths differing by only one centimetre in the range between three and twelve inches, whereas in the British list there are in the same range only ten different depths, which is sufficient for all practical purposes. On the other hand, many British rolled steel joists are made in more than one width, while the metric rolled steel joists have only one width. The one table was compiled by practical engineers, the other by theoreticians.

The one undoubted advantage of the metric system is that density is the same as specific gravity.

The main difficulty in changing to the new system will be by the man, whether engineer, draughtsman, or turner, who has been brought up on inches and "thous" and finds that he cannot think in any other medium. How many people today think of the rainfall in millimetres instead of inches and "points," although the official change was made years ago? The only way is to retire or kill off all the present workers and to raise a new generation from the cradle on metres!

Mr I. R. G. Stephen (Vice-President): The history of the science of measurement is in fact the history of mankind, and the adoption of standards acceptable to all nations is a logical outcome of development in this field.

It has been said that a man's material success may be measured by the amount of secret information he possesses. This is very true, but there must come a time when he is forced to divulge at least a part of his knowledge for the sake of standardization. It is only now, with the advent of the second industrial revolution, that the nations of the world are facing the inevitability of the adoption of a common system of weights and measures.

It is interesting to note that one unit of measure, the degree of arc, has remained unchanged through the centuries. The early Babylonians had divided their year into 360 days, and so decided to divide the circle into 360 parts as well. Being skilled in the art of geometry, they knew that a chord equal in length to the radius subtends an angle of 60°. Thus, the number 60 became the basis of their sexagesimal number system, and so the degree was divided into 60 minutes, and the minute into 60 seconds.

In one of his books Lancelot Hogben mentions the fact that the Babylonians, who lived in the land known as Mesopotamia, and those civilizations which came after them, had considerable foreign trade. Standards were of particular importance in their business transactions, and it is recorded that one of the earliest measures of farm produce was the donkey load. For more valuable goods, the larger unit of weight was the talent (approximately 55 pounds) and the smaller unit, used for precious wares, was the shekel (about one-third of an ounce).

In Egypt, a thousand miles to the west, flourished another ancient civilization, under the influence of the all-powerful priests, who ordered the building of great temples, and the pyramids. Egypt had no foreign trade and so the first essential standards to be chosen were 'those of length rather than of mass. The earliest standards of length were based on the proportions of one man's body, possibly the kings. Thus, we had four "digits" (fingers) equal to one "palm," and seven palms equal to one "cubit" (the distance from elbow to outstretched middle fingertip). The "foot" was, naturally enough, the distance from the heel to the tip of the big toe. The standards thus decided upon were fixed by calibrated rulers made of wood or metal. It is indeed a far cry from the foot of an Egyptian king to the wavelength of the radiation of the krypton atom.

In the changeover from one system to another, time is never on one's side. The longer the change is postponed, the more difficult and expensive does the change become. Mr Whitwell mentions the educational aspects of the change. I am pleased to report that the Department of Education, Arts and Science adopted the M.K.S. system of units in 1965 for all its electrical engineering courses. Electrical engineering has, of course, always had fewer problems in standardisation than has the older-established branches of engineering, but people today are conditioned to the fact that "the old order changed, yielding place to new ... " and I for one do not foresee any less successful a transition in South Africa than has already occurred with the monetary system.

AUTHOR'S REPLY TO DISCUSSION

Mr J. L. Whitwell:

In reply to Mr R. von Vivenot:

The problem of changing to internationally preferred sizes is a complex one and will require careful and detailed planning. The preliminary stage of expressing existing inch sizes in terms of centimetres may well be dispensed with in the case of many product where their sizes can be changed quickly.

he possible use of Stage I was suggested because of the necessity to re-orientate mental attitudes and to inculcate metric thinking in industry as quickly as possible.

It is of interest to note that, taking the population figures for December 1962, issued by the United Nations, 82.7 per cent of the world's population uses the metric system of weights and measures exclusively.

Tn reply to Mr T. O'D. Duggan:

The decimalized inch has been in use for a long time and it has been suggested as an alternative to metric units-particularly by interests in the U.S.A. The world trend towards metric units would appear to make the possibility of international acceptance of the inch increasingly remote.

A change in the sizes of the basic units of any system, whether inch or metric, would involve serious consequences. It is desirable therefore that the sizes of the basic units remain the same, even though the coefficients relating them to natural phenomena are not nicely rounded up. Scientists admit that they are not infallible, and that future research may show that their calculations were wrong. One of the advantages of the metric system is that it can, and does, keep up to date with scientific knowledge, an advantage appreciated by the non-metric countries who define their inch in terms of the centimetre.

In India, the change-over to the metric system was started in 1958 with a 10-year programme. By the end of 1965-after seven years-the change-over was almost complete. The Metric System Committee suggested a period of 10-15 years for South Africa, but it may well be that the change could be achieved' in a much shorter time.

About changing working drawings, in many cases there would be no need to make any alterations. The internal working of a manufacturing plant could remain in inches until the management decided otherwise. The main requirement would be that any handbook or description for external consumption must be expressed in metric terms. To take a particular instance, the sizes of component parts of a motor-car engine could continue to be expressed in inches for internal use in the factory should the manufacturer so wish. Such parts are not standardized as they are peculiar to a particular engine and are not, in general, exact, or rounded-up inch sizes. There are already examples in South Africa of the reverse process, e.g., the production of metric dimensioned Continental cars.

In reply to Mr Waites:

The question of the cost of changing the calibrations of machines and of providing new measuring equipment involves the timing of the change-over. The recommendation of the Metric Committee was that the choice between whether to change the calibrations on lathes, grinders, etc., to metric units or to continue the use of inch calibrations in combination with conversion tables should rest with the individual factory concerned. Such a choice would be governed by such factors as age of the machine, cost of conversion, obsolescence, usage of the machine, etc., and the factory management would apply normal economic criteria to decide when to alter its machines.

In reply to Mr E. M. P. Evans:

There is nothing sacrosanct about the decimal system. It has, as Mr Evans has pointed out, arisen because of counting in tens. Other numerical systems have the advantage of additional factors-and there is no question that the "dozen" package (3x4) is the most practical arrangement for boxes, etc. Another system is the binary system, which counts in two's and which is used in computers. The use of 10's in arithmetic is so deeply entrenched, however, that the use of any other base would mean an almost impossible, if not completely impracticable change involving every inhabitant of the earth.

The International Standards Organization has published standard thicknesses of sheet and diameters of wires based on the system of preferred numbers i.e., arranged in geometric progression. The same organization is engaged in preparing a single list of metric standard rolled sections. At present there are several different metric standard sections ­ French, German (DIN), Euro norm (European Steel and Coal Community), etc., and the reduction to one standard set must lead to more economic production.