Light is radiant power in a narrow band of the radiation spectrum; it is the basic material of vision. Both quantitative and qualitative human responses vary with wavelength, as do the optical properties of materials, so the spectral luminous efficiency (Vλ) is a fundamental relationship in the measurement of light. Light suffers only two kinds of change between its source and our eyes; it may be re-directed many times (from one straight line to another) or it may be partially absorbed. Vision depends on these changes, always in terms of ratios rather than differences. Photometry, which depends on the measurement of radiant power, is traditionally on a linear scale. The units of light are complex, because light has long been a dominant feature of human experience. They may be unnecessarily confused if non-metric units of length are used. They are: Flux - the total luminous power (measured in lumens) Intensity - the angular concentration of flux (candelas) 2 Illuminance - the surface density of incident flux (lux or lumen/m2 ) Luminance - the intensity emitted per unit area (candela/m2 ) Other terms such as transmittance, absorbance or colour temperature are briefly discussed. The SI standard of light is the candela, defined physically in terms of luminance, but sometime in the future the standard will probably be the lumen,defid in terms of an efficacy of 683 lumens per watt for monochromatic radiation whose frequency is 540 x 1012 hertz (wavelength 555 nanometres).

The international method of assessing light by means of the V(λ) function was established by the CIE in 1924. Some shortcomings have since become apparent; in particular, lights with a high blue content may be underestimated. Linearity is assumed so that lights with extended spectral distributions can be measured (Abney's Law). This is only a rough approximation to the behaviour of the light-adapted eye. Linear photoelectric photometers which obey the CIE convention are readily made, and these will rank lights on a scale proportionate to the radiant power. This may not always correspond with the sensation of brightness. In 1951 the CIE recommended a standard scotopic observer, V'(λ), for very low levels. This is seldom used, probably because appropriate conditions are rarely encountered. The very wide range between the two standard observers is not covered officially at all.

The paper briefly traces the development of illumination photometers from Preece's crude greasespot photometer to the present day selenium and silicon photocell photometers, with cosine and colour correction and digital displays. The accuracy and calibration of modern photometers is discussed. In addition to the variation of illuminance values with time there is also the variation in space and the need to determine the average illuminance. The origin of the I.E.S. recommended method for measuring average illuminance is explained and its limitations discussed. There is great interest in the measurement of Contrast Rendering Factor which is a measure of the visibility of a task in a given lighting installation. Blackwell in the U.S.A. and the Building Research Establishment have developed visibility meters which are intended for use in measuring CRF. The need to consider the lighting of the space within the room envelope has resulted in two alternative parameters being suggested. These are scalar illuminance and cylindrical illuminance. Scalar illuminance is the average illuminance on an infinitesimal sphere and an instrument has been developed to measure scalar illuminance. Cylindrical illuminance is the mean vertical illuminance and this may be measured by averaging a series of readings or by using a cylindrical integrator. Luminance measurements are made only infrequently, but they can be important for research work. Typical lUminance meters are described.

This paper tries to highlight the most important aspects of the problem of ensuring that the user of a calibrated reference lamp faithfully reproduces the electrical input originally defined by the calibrating laboratory and considers the influence of these factors upon the specification of the lamp power supply unit. Tungsten filament lamps are the most widely used as references and discussion has been limited chiefly to PSU's for this type of lamp only. The paper touches upon the relationship of electrical input and light output for such lamps; errors which can occur in electrical measurements at AC and DC; specification of PSU performance and facilities; appropriate types of PSU and notes on their strengths and weaknesses including some ideas of relative costs.

The photometry of retroreflectors involves special techniques and accurate and reproducible results are difficult to achieve. The photometric performance of a retroreflector depends critically on the measurement geometry and the angular apertures of the illuminating source and the photometer. To simulate practical use conditions, very small angles are involved which need precise adjustment and which limit the amount of light available for measurement, so that a sensitive photometer is needed. Precautions are necessary to avoid errors due to stray light. Many retroreflectors are coloured so that the spectral power distribution of the illuminating source must be specified and careful attention paid to the adaptation of the photometer to the CIE standard photometric observer. Work is in progress internationally which it is hoped will lead to recommendations for a measurement technique which will reduce the present discrepancies in results between laboratories to an acceptable level.

Lighthouse optics are usually complex designs, the lens components tailored to suit both the type of light source available at the time of installation and the navigational requirements of the locality. The original light source may have been either a paraffin flame or an incandescent mantle; modern technology now requires that electric lamps are used. It is not economical to make lamps comparable in shape to the original source and it is also not practicable to reconstruct and refocus the existing lens components. The lens designs differ widely, each to suit the needs of its station, and the beam characteristics are affected differently by the choice of electric lamp. These effects are not usually calculable by accepted methods, as for example those in BS 942. A programme of investigation was instituted to select a suitable current production lamp for a range of lens designs, most of the lamps were only available for trial on site. The criteria were adequate luminance, long life, high luminous efficacy, correct colour and a combination of shape and uniformity compatible with the lens. These criteria could only be properly judged by photometric measurement of the actual beams of light. A photometric technique for on-site measurements had to be developed. There were three principal problems; the changing atmospheric attenuation over long distances, the rapid horizontal traverse of a rotating beam, and the mechanical impossibility of tilting the optic to provide a vertical scan through the beam. The solutions are: a high-intensity standard projector (0.2 Mcd) and a sensitive telephotometer for measurements down to the visual threshold, a rapid response amplifier and recorder or a time-integrating photometer to measure the horizontal scan of the beam, and large prismatic panels in front of the lens to depress or elevate the beam through small angles (±2°) in the direction of the photometer. By these means, the fourth problem of interrupting the regularity of the signals given to mariners at night did not arise.

As a result of the 1972 International conference on the 'Prevention of Collisions at Sea', the Department of the Environment was requested to conduct tests on navigation lanterns to enable the Department of Trade (Marine) to certify those which they considered complied with the agreed International Regulations. The development of a test programme and the presentation for the photometric and chromaticity measurements on these lanterns is discussed. The photometry is conducted on a remotely controlled goniometer. The lanterns are mounted on a dividing head table, with the geometric centre of the lens aligned to zero degrees in elevation. Intensity distribution values are recorded by means of a selenium photo voltaic cell/CIE filter combination and direct reading galvanometer. These values are then used to construct unique log-linear isocandela diagrams. The colours of the lights are measured on a modified Lovibond Schoffield Tintometer. Suitable amounts of permanent glass filters in either one or two of the three subtractive primaries, red, yellow or blue are introduced until a subjective colour match is achieved between the sample and reference fields. Readings are then converted to x, y, z co-ordinates and checked for compliance with the International Regulations.

There is an extensive and increasing demand for photometric measurements on low level light sources. To make such measurements low intensity standards are required. The tungsten filament lamps which form the normal standards of luminous intensity have a minimum useable intensity of the order of one candela. For low level photometry a standard lower than this by several orders of magnitude is required. The normal methods of photometric attenuation are either insufficiently powerful or have other serious drawbacks. For these reasons a device using a neutral white diffusing surface and a small calibrated aperture is used. It will give an intensity attenuation of up to the order of 106 times.

Starting with the well-known nomenclature, symbols and units of thermal radiators and referring to the well-known nomenclature, symbols and units for quantities of visible light one can define a new, quite analogous nomenclature for an arbitrary receptor (detector) with a spectral sensitivity R(λ) . It is pointed out that it might be wise to reject the familiar, however historically grown units in the field of visible light (the candela) and to stick to (light-)watts that match better to the International System of Units (SI) . In an analogous way a system of additional SI-units, named (recepto-)watts is presented. Some examples are given.

For monitoring individual persons we use plastic films which discolour to UVR. This change, measured spectrophotometrically, is proportional to exposure. Polysulphone film has been found to have chief biological application as its spectral reactivity matches normal sunburn. PVC films incorporating various photosensitive chemicals have also been under study. Polysulphone, as a personnel dosimeter, has been used to study hospital-bound geriatric patients, office workers, laboratory workers and gardeners. People indoors were exposed to about 5% of ambient during the working week, but at weekends 50% of weekly dose could be attained during outdoor leisure activities. But bedridden geriatric patients received negligible UVR. Film badges have also been used to study anatomical distribution of solar UVR over the body, and the distribution in patients undergoing phototherapy. Film has been employed to study psychiatric patients on tranquilliser drugs where photosensitivity is an adverse side effect. Film badge dosimetry for UVR remains unexploited in industry; here its simplicity and cheapness would seem outstanding features.

Unusual light measurements are defined as those requiring a special design of instrument or a radical modification of an existing commercial instrument. The term "Industry" as used in the title has been interpreted in a liberal manner. This allows the topics dealt with to include physiological and medical applications. Topics dealt with are (i) Agriculture, horticulture and forestry - these studies involve a unit called the Einstein which is used in the measurement of photosynthetic action in plants. The spectral form of Photosynthetically Active Radiation is discussed. Other agricultural measurements discussed are the measurement of the efficiency of greenhouses and their materials and also the automatic selection of tomatoes on conveyor belts according to size and colour. (ii) Resin curing by light, both UV and visible is now widely, used in industry for bonding metals and glass. Another application is in dentistry where visible light cured resins are becoming available. (iii) Bioluminescence has been applied for some years in the study of drug action. More recently its application in the important field of anaesthesiology has shown it to be a useful indicator of the potency of anaesthetics. (iv) Phototherapy of bilirubinemia, or infant jaundice, in new born babies is now widely applied and studies are still going on to establish optimum dosage and spectral content of the irradiating lamps.

The requirements of lamps used for the different types of photometric standard are stated and are related to the problems for the photometrist caused by the changes in the commercially available range of lamp types in the past 20 years. The need for suitable lamps to fill the gaps caused by discontinuation of the types used earlier in applied photometry led to NPL developing new heavy-current lamps for the purpose, in conjunction with GEC Ltd. The limited range of existing types of photometric standard lamps is described and the properties of the new lamps are contrasted with those of the existing ones. Principal advantages are robustness and the greatly improved range of possible combinations of operat-ing temperature and life. Although very reproducible in output over extended periods, the new lamps exhibit some convective gas fluctuation noise. While the effect of this can be reduced adequately by using a long time constant or integrating period in the photometry, steps are being taken to find ways of reducing the noise in the light output. Cases are discussed where commercial lamp types can still be used.

Graphic arts is an industry developed over the years to describe a process enabling the printing press to reproduce in quantity a facsimile of the original. It is not intended in this paper to explain the various processes in detail but to generalise and to show how light plays an important feature. Originals, which is the term used for the 'object to be copied' can be in many forms; black/white or colour, and of varying sizes. They can be split into two distinct areas: flat copy and photographic originals.

Street lighting works by illuminating the road surface so that objects such as pedestrians may be seen in silhouette. There is thus an interest in being able to calculate the luminance pattern on the road so that it may be designed to give adequate visual performance to the driver. To do this one needs to know the intensity distributions and positions of the lanterns used and also the luminance factors of the road surface. The luminance factor Q (luminance/illuminance) varies greatly with angle of incidence so that a table of luminance factors is needed for each surface. Following the calculation through one obtains the simple basic formula for luminance of a point L = (1/H)2(I)(r), where H is the mounting height of the lanterns, I is the intensity given in the relevant direction and r is the reduced luminance coefficient (Q cos3y), y being the vertical angle of incidence.

The choice of camera and tube depended upon the consideration of criteria such as the dynamic range, linearity, uniformity of response, resolution, flare, persistence, spectral response and number of blemishes. A method of asessing the resolution of a television camera system which gave the modulation transter function (MTF) directly on an oscilloscope is described. The camera employed in this particular application was of the low light level type, the pictures being recorded on a video tape recorder (VTR) whose characteristics also had to be considered. A calibration grey scale was included in the picture by means of a split viewing arrangement. The fundamental feature of the video analysis system for deriving the luminance values was an analogue processor which gave the average video signal level over any selected area of the television picture. The accuracy of the system has been assessed for road luminance values measured from a mobile laboratory by comparing them with independent measurements using a Pritchard photometer.