The size and location of several of the largest 19th century earthquakes in the San Francisco Bay area, including the June 1838, 8 October 1865, 21 October 1868, and 24 April 1890 events, are re-evaluated by comparing these historic events to several modern events, in particular the 1989 Loma Prieta earthquake. Similar high intensities in the Monterey Bay area during both the 1989 and 1838 events indicate that the 1838 shock probably ruptured the Loma Prieta segment in addition to the Peninsular segment of the San Andreas fault. If so, the resulting earthquake would have been of moment magnitude MW ≧ 7.2. Thus, like segments of many other plate boundaries, the Loma Prieta segment exhibits heterogeneous rupture in that sometimes it ruptures with the Peninsular segment, as in 1906 and probably in 1838, and at other times it ruptures alone, as in 1989. In contrast, the area of high intensities generated by the 1865 earthquake was centered in the Santa Clara Valley and the eastern foothills of the Santa Cruz Mountains, indicating that the 1865 shock probably was located northeast of the Loma Prieta rupture and not on the San Andreas fault. In addition, the felt areas, taken to be one of the best measures of relative source size, of modified Mercalli intensities (MMI) V and VI for the 1865 shock are considerably smaller than that of the 1989 event. Relationships between felt areas of intensities V and VI and moment magnitude are developed for the San Francisco Bay area. Based on these relationships, the average computed moment magnitudes for the 1865, 1868, and 1890 earthquakes are 6.51 ± 0.18, 6.84 ± 0.20, and 6.32 ± 0.16, respectively.
If these results are correct, then the Loma Prieta segment, taken to include all faults within the San Andreas fault zone, i.e., within a zone up to a few kilometers wide, appears to rupture with or without the Peninsular segment every 76 ± 11 years. These data on repeat times may be the most accurate information now available for ascertaining the date of future rupture of that segment and for testing recurrence models. The smaller size of the 1865 earthquake would be consistent with it being part of the long-term seismic build-up to the 1868 event rather than being paired with it as events of similar size. If the 1865 earthquake did occur northeast of the Loma Prieta segment, then a future source of that type, located closer to San Jose and the Santa Clara Valley, could pose a considerable hazard to the southern San Francisco Bay area. On-going searches of letters and diaries of missionaries and early settlers could yield new information that should help to further constrain the locations and magnitudes of these and other historic earthquakes.