Measurements of mass in Sweden are based on the mass of the Swedish platinum-iridium prototype no 40. The prototype is traceable to the mass of the international prototype and is certified by BIPM in 1889 and 1948. In order to control the traceability of the prototype in accordance to the metre convention it has been decided to subject the prototype to a comparison at BIPM involving a cleaning procedure and a subsequent calibration against the international prototype. The transport of the damageable prototype is always adventurous irrespective of how high the safety precautions are. Therefore a comparison between two of the mass standards of the Statens provningsanstalt (H and MJV1) and the prototype P have been carried out in order to ensure a good secondary standard during the absence of the prototype. The balance which was used at the comparison was a newly acquired mass comparator of two knives type, manufactured by Voland Co. As the aim of the comparison was, if possible, to determine 1 kg with an inaccuracy up to a few ug, such a weighing requires the temperature of the air, the weights and the whole balance to be kept within some hundredths of a degree C. Other error sources which were kept under strict control were air pressure, air humidity and CO2 concentration. The quantitative evaluation of the air density in the laboratory was done by a computer, based on a new formula recommended by BIPM. Because of the high accuracy, which is obtained by doing the comparison in "triangle", the weights H (brass) and MJV1 (stainless steel) were examined by a complete comparison with the prototype P (1 kg - 0.039 mg) in three combinations according to the following scheme: H - P, MJV1 - P, MJV1 - H. To get a sufficient statistical data each "triangle" was repeated 9 times. Accordingly inalles 27 measurements were done. The following mean values and inaccuracies were found for H and MJV1: H = 1 kg + 6.80 mg ± 0.02 mg MJV1 = 1 kg + 0.15 mg ± 0.02 mg