The ratios listed in Table VI.C.5 indicate that, for the three asteroids used in the calibration, the flux at 25 µm is higher than expected on the basis of the stellar calibration and the asteroid models by 4 ± 1% once the fluxes have been normalized at 60 µm. This is well within the expected accuracy of the stellar and asteroid models. The dispersion of the ratio of observed to model fluxes at 100 µm is only 3%, the mean being identical to unity by definition of the asteroid calibration procedure. The good fit at 25 µm and the small dispersion at 100 µm gives confidence that the extrapolation of these asteroids is a valid procedure.
An alterative explanation of the results in Table VI.C.5 is that the difference between the observed and predicted 60 to 25 µm flux ratio could be a manifestation of a short wavelength leak in the 60 µm spectral response such as reported from the preflight component tests (Section II.C.4). An examination of the 60 µm to 12 and 25 µm colors of the calibration stars (Table VI.C.3), however, shows no change in color larger than 0.02 mag for different effective stellar temperatures; whereas the colors should change by 0.04 mag for effective temperatures varying from 4000 to 10,000 K, if the short wavelength leak were real.