Table of contents

Adsorption errors, caused mainly by water vapour, in the observed values in the calibration of the highest-quality mass standards have become significant. Adsorption corrections for various surface conditions of weights have not yet been established.

The propagation of adsorption errors to each calibrated value may be decreased by the improvement of weight series or weighing designs. But there is no measure to estimate the propagated errors.

There are two kinds of Equivalent Surface Area for adsorption (ESA) defined in this paper. One applies at constant humidity; the second, at fluctuating humidity. The first one depends upon the geometry of the weights used, the second one also depends upon the calibration scheme employed. Consequently the second ESA is proposed as a practical measure for conditions of fluctuating humidity.

The electron storage ring BESSY was investigated for use as a primary standard of spectral irradiance at wavelengths between 400 nm and 1000 nm. A method was devised to transfer the spectral irradiance scale realized on to tungsten halogen lamps by means of a radiometer that measured either the spectral irradiance of BESSY or that of the lamps. The relative systematic uncertainty of the spectral irradiance produced by the storage ring was determined to be 0.3%. Presently, the calibration of tungsten halogen lamps as radiometric transfer standards of spectral irradiance from synchrotron radiation yields a relative systematic uncertainty of 0.8%, which is below the uncertainty that is currently achieved using black-body radiators. Spectral irradiances of tungsten halogen lamps measured by means of synchrotron radiation and by means of black-body radiation agree within the combined uncertainties of the two methods. We conclude from the investigations that a reduction of the uncertainty of the spectral irradiance of the storage ring to approximately 0.1% can be achieved if it seems worthwhile. Also, the uncertainty in the calibration of the spectral irradiance of tungsten halogen lamps could be reduced below 0.5% if desired.

Advances in crystal growth technology together with advances in preferential etching techniques may make it possible to produce a nearly perfect single crystal of silicon in the form of a 111 tetrahedron. The density homogeneity of such an artifact together with the high precision to which its centre of mass could be located with respect to its faces would suggest its use as an attracting mass in experiments to determine the gravitational constant G. In this paper a closed-form expression for the gravitational field of a 111 tetrahedron is derived.

The theory of the thermal boundary layer at the walls of a spherical acoustic resonator is discussed in detail. For gases at low pressures, the temperature-jump effect is found to make a significant contribution to the resonance frequencies of the radial modes but not to their acoustic losses. Experimental results are reported for argon at 273.16 K and pressures between 15 and 248 kPa, and compared with the theory. These were obtained using the four radial modes with lowest frequency of a spherical resonator with a radius of 60 mm. The thermal accommodation coefficient between argon and the aluminium wall of the resonator was found to be (0.84 ± 0.05). The results suggest that a determination of the gas constant with a fractional imprecision of 1 × 10^-5 or better should be possible using a spherical acoustic resonator.

Measurements of the quantized Hall resistance R_H (i) (i = 2 or 4) in 7 different heterostructures (six GaAs based, one InP based) are reported. R_H (i) is measured in terms of Ω_LCIE by means of a resistance-ratio measurement bridge using a cryogenic current comparator.

The peak-to-peak scatter of the results is 8.5 × 10^-8. An estimation of R_H (i = 2) in terms of Ω_LCIE is given with a 1 σ (one standard deviation) total uncertainty of 2.2 × 10^-8.

A new interferometric measurement of the caesium D₂ resonance line wavenumber has been performed. The source is a diode laser stabilized on the fluorescence light of a caesium beam. Its wave number is measured with an improved interferometric wavemeter (lambdameter). The σ_D2 wavenumber is determined with an accuracy of 3.5 × 10^-8.