Context.— Grant Medical College and Sir J. J. Hospital, Mumbai (India), have a long tradition of clinical autopsies, wherein autopsy records have existed since 1884. The old autopsy records from 1884 to 1966 were discovered during a refurbishing drive of the pathology department in 2007. Objective.— To know the relative causes of deaths in Sir J. J. Hospital Mumbai from 1884 through 1966. To determine the rate of antemortem and postmortem discrepancies in the prehistology and posthistology era. Design.— The reports of clinical autopsies in the archives of the Department of Pathology from January 1884 to December 1966 were reviewed. Data such as sex, age, and clinical and autopsy diagnoses were collected, based on the autopsy records of 13 024 patients. Results.— Of 13 024 patients, 10 197 (78%) were male and 2827 (22%) were female. Most deaths (55%) involved individuals ages 21 to 40 years. Infections were the cause of death in 7281 of 13 024 patients (56%), followed by cardiovascular diseases in 2138 (16%) and neoplasms in 963 (7%). The overall disagreement between clinical diagnoses and postmortem findings was 4105 of 13 024 patients (31%). Discrepancy between antemortem and postmortem diagnoses in the prehistology era (before 1950) was found in 3053 of 8951 patients (34%), whereas in the posthistology era (after 1950), discrepancy was found in 1019 of 4073 patients (25%). Conclusions.— Historical collections provide baseline data against which modern observations can be compared. Such collections also offer a window on the past and often provide a less biased account of events. Historical collections of museum materials such as ours are valuable because they can serve as potential biorepository materials to facilitate future research. Apart from providing reliable data on the occurrence of diseases, the autopsy has an important role in providing information about conditions that are no longer prevalent. Medical institutes possessing old autopsy records, wet specimens, paraffin blocks, and slides should preserve such historical collections.

The authors state that the data in these registers ''were never published.'' They are correct in the sense that the entire set of autopsy records for that period has not been published. Yet, it is pertinent to note that the data have been used in considerable part for many other published studies. The authors have entirely overlooked the pioneering work of P. V.