The light curve of the Type Ia supernova SN 1895B (in NGC 5253) is imp
ortant because Sandage et al. have measured a distance to the host gal
axy by means of Cepheid variables and thus have derived the Hubble con
stant. However, the last time the supernova plates were measured was i
n 1923, and the resultant photometry is inconsistent at the 0.8 mag le
vel. In this paper, I report on the results of my remeasurement of the
old supernova plates with a scanning microdensitometer. I find B-magn
itudes of >11.3 +/- 0.3, 8.36 +/- 0.07, 8.51 +/- 0.06, 8.54 +/- 0.05,
and 9.34 +/- 0.08 on JD 2,413,359, 2,413,383, 2,413,384, 2,413,385, an
d 2,413,393, respectively. The last plate was a prism plate, for which
I present a flux-calibrated spectrum, taken similar to 15 days after
maximum. Even though only loose constraints can be placed on the light
curve shape, the firm limit on the peak magnitude (<8.49 +/- 0.03) wi
ll provide a firm limit on the Hubble constant. The most likely shape
is that of the average Type Ia template of Leibundgut et al., for whic
h the best-fit peak B-magnitude is 8.26 +/- 0.11. The deduced probabil
ity distribution for H-0 is peaked at 51 +/- 7 km s(-1) Mpc(-1), with
extreme limits of 61 +/- 12 and 26 +/- 5 km s(-1) Mpc(-1).