The Koenig-Martens spectrophotometer as installed and used at the Bureau of Standards has been tested by measuring the transmissions of rotating sectored disks of known angular aperture. Measurements of the spectral transmission of blue filters by different observers have also been compared. The photometric probable error at various wave-lengths has been computed from the observations of three different observers.
From the data, which are presented in considerable detail in tables and graphs, the following conclusions are drawn:
- 1. For heterogeneous illumination, the probable error of a single observation (involving one reading each of θ1 and θ2) is about seven tenths of one percent of the transmission when measurements are made at the field brightness most favorable to accurate matching.
- 2. For homogeneous illumination the probable error of a single observation of transmission 0.37 is nearly constant and equal to about one percent of the transmission in the case of the following wave-lengths of mercury and helium: mercury, 435.8, 546.1 mμ; helium, 447.2, 492.2, 501.6, 667.8, 706.5 mμ.
- 3. For heterogeneous illumination, the actual error2 of the mean of ten observations for transmissions even as low as 0.015 and for wave-lengths between 430 and 700 mμ is usually less than one percent of the transmission, provided the transmission is not changing rapidly with wave-length.
- 4. There is no systematic error of measured transmission with respect to wave-length.
© 1924 Optical Society of AmericaFull Article | PDF Article
More Like This
Irwin G. Priest
J. Opt. Soc. Am. 8(1) 173-200 (1924)
Kasson S. Gibson and Margaret M. Balcom
J. Opt. Soc. Am. 37(8) 593-608 (1947)
K. S. Gibson
J. Opt. Soc. Am. 21(9) 564-587 (1931)