observed savings %: {'20m': 58.2, '1h': 44.2, '8.8h': 35.8, '1d': 33.7, '2d': 27.8, '6d': 25.4, '31d': 21.1}
model                          RMSE    b(7 d)   b(30 d)  params
log, his k=1.84 c=1.25         1.73      24.5      21.3  k=1.84 c=1.25
log, re-fitted                 1.64      25.2      22.0  k=1.818 c=1.214
power a*t^-d                   1.91      24.6      20.4  a=82.45 d=0.1311
exponential a*e^(-t/s)         9.10      32.4      16.3  a=39.96 s=48334 min

his Calculated column reproduced to within 0.1 point(s): ours [57.0, 46.8, 34.5, 30.4, 28.1, 24.9, 21.2] | his [57.0, 46.7, 34.5, 30.4, 28.1, 24.9, 21.2]  (t=64 gives 46.77; he printed 46.7 — a hand calculation from log tables)

'forgotten' in Ebbinghaus's own column IV is 100 - Q: at 1 d 66.3, 6 d 74.6, 31 d 78.9 (never 90)