Key points are not available for this paper at this time.
An observational selection bias for Cepheid variables that flattens an observed P-L relation from its true slope is identified. Its existence is simulated by truncating an unbiased (complete) list of Cepheids in the LMC at progressively brighter magnitudes and showing thereby that the slope of the P-L relation of the remaining variables is too shallow. This slope effect is shown from literature data for Cepheids in the LMC, NGC 6822, IC 1613, M 31, M 33, WLM, Sextans A, Sextans B, NGC 3109, and NGC 300. Apparent distance modulii in the B band relative to the LMC are obtained as DELTA (m - M) = 5. 65 for NGC 6822, 5. 65 for IC 1613, 6. 0 for M 31, 6. 2 for WLM, 6. 3 for M 33, 7. 3 for Sextans A and B, 7. 4 for NGC 3109, and 7. 6 for NGC 300. If (m - M) AB_ = 18. 95 for the LMC, obtained by using the adopted Cepheid zero point discussed in the text, the apparent blue modulus of M 33 (about which there has been previous debate) is (m - M) AB_ = 25. 25. If AB_ = 0. 6 mag for the M 33 Cepheids (Freedman 1985), the true modulus of M 33 is (m - M) = 24. 65 from these data.
Allan Sandage (Mon,) studied this question.