A 5m, scaFle is standardized based on light wave standard by the interference compa rator, a slight modification of the one described in a previous report . The optical standards used are 5 cm. and 50 cm. Fabry-Perot etalons made of iron pipe and a 5 m, etalon constructed on a concrete block. The 5 cm. etalon is determined in terms of the wave length of the cadmium red line. The 5 m. etalon is determined by the successive comparisons of the shorter etalons using normal incident ray by spectroscopic method . To eliminate the effect of the change in the length of the etalons and the systematic error due to timed measurements of respective etalons, the 5m, etalon is determined symmetrically before and after its comparison with the 5m, scale, by either departing from the measurement of the 5 cm, etalon or from the comparison of the 5 m. etalon with the 50cm, etalon. The mean of the two values obtained from the go and back measurements is adopted as the length of the 5 m. etalon when it is compared with the 5 m, scale. The systematic error due to the timed measurements is clearly recognized according to the order of measurements, especially in the 5 m . etalon (Table I). When this error is elimi nated, however, the 5 m, etalon is found to change its length much less than the other etalons do (Table II) and the amount is estimated more or less lμ during the comparison with the 5 m, scale. From 11 sets of measurements, the scale is determined with the standard deviation of ±1.05μ and the mean error of±0.32μ (Table IV).
In our privious reports, to measure a distance of 5m, by interference of light waves, we multiplied lengths by ten times in each step . For a distance shorter than lm., how ever, we have already found the possibility of applying twenty times multiplication at a time. In this report, we examine the applicability of the multiplication of more than ten times to a distance of 5m, and succeed to compare a 5m, etalon directly with a 25cm, etalon and even with a 20cm, one respectivery, using normal incident ray and spectroscopic method. These cases correspond to 20 and 25 times multiplications respectively . Although details are not given as to the latter comparison, a check experiment is described performed as to the former comparison. The direct comparison of 5m . and 25 cm, etalons is compared with the indirect one in which a 50cm, etalon is used as an intermediate distance. The result shows the discrepancy of 7×10-8 for the total length between these two comparisons. This inconsistency, however, may be considered inevitable in view of the instability of the atmospheric conditions in the measuring room and of the high and different temperature coefficients of respective etalons. It is confirmed that a distance of 5m, is measurable through only one step of multiplication from the distance of 20cm, which is measurable directly by the light from a Hg-198 lamp.