|National Archives and Records Administration|
The 1890 census was the first to use punchcards and an electical tabulation system. (Courtesy Bureau of the Census)
Prologue: Quarterly of the National Archives and Records
Spring 1996, vol. 28, no. 1
"First in the Path of the Firemen"
|Newspaper photographs captured the scene after the devastating fire and pointed out the need for safe storage of national records.|
The morning after was an archivist's nightmare, with ankle-deep
water covering records in many areas. Although the basement
vault was considered fireproof and watertight, water seeped
through a broken wired-glass panel in the door and under the
floor, damaging some earlier and later census schedules on the
lower tiers. The 1890 census, however, was stacked outside the
vault and was, according to one source, "first in the path of the
firemen."(11) That morning, Census Director Sam Rogers reported the
extensive damage to the 1890 schedules, estimating 25 percent
destroyed, with 50 percent of the remainder damaged by water,
smoke, and fire.(12) Salvage of the watersoaked and charred
documents might be possible, reported the bureau, but saving even
a small part would take a month, and it would take two to three
years to copy off and save all the records damaged in the fire.
The preliminary assessment of Census Bureau Clerk T. J.
Fitzgerald was far more sobering. Fitzgerald told reporters that
the priceless 1890 records were "certain to be absolutely ruined.
There is no method of restoring the legibility of a water-soaked
Four days later, Sam Rogers complained they had not and would not be permitted any further work on the schedules until the insurance companies completed their examination. Rogers issued a state-by-state report of the number of volumes damaged by water in the basement vault, including volumes from the 1830, 1840, 1880, 1900, and 1910 censuses. The total number of damaged vault volumes numbered 8,919, of which 7,957 were from the 1910 census. Rogers estimated that 10 percent of these vault schedules would have to be "opened and dried, and some of them recopied." Thankfully, the census schedules of 1790-1820 and 1850-1870 were on the fifth floor of the Commerce Building and reportedly not damaged. The new 1920 census was housed in a temporary building at Sixth and B Streets, SW, except for some of the nonpopulation schedules being used on the fourth floor.(14)
Speculation and rumors about the cause of the blaze ran rampant. Some newspapers claimed, and many suspected, it was caused by a cigarette or a lighted match. Employees were keenly questioned about their smoking habits. Others believed the fire started among shavings in the carpenter shop or was the result of spontaneous combustion. At least one woman from Ohio felt certain the fire was part of a conspiracy to defraud her family of their rightful estate by destroying every vestige of evidence proving heirship.(15) Most seemed to agree that the fire could not have been burning long and had made quick and intense headway; shavings and debris in the carpenter shop, wooden shelving, and the paper records would have made for a fierce blaze. After all, a watchman and engineers had been in the basement as late as 4:35 and not detected any smoke.(16) Others, however, believed the fire had been burning for hours, considering its stubbornness. Although, once the firemen were finished, it was difficult to tell if one spot in the files had burned longer than any other, the fire's point of origin was determined to have been in the northeastern portion of the file room (also known as the storage room) under the stock and mail room.(17) Despite every investigative effort, Chief Census Clerk E. M. Libbey reported, no conclusion as to the cause was reached. He pointed to the strict rules against smoking, intactness of electrical wires, and noted that no rats had been found in the building for two months. He further reasoned that spontaneous combustion in bales of waste paper was unlikely, as they were burned on the outside and not totally consumed.(18) In the end, even experts from the Bureau of Standards brought in to investigate the blaze could not determine the cause.(19)
The disaster spurred renewed cries and support for a National Archives, notably from congressmen, census officials, and longtime archives advocate J. Franklin Jameson.(20) It also gave rise to proposals for better records protection in current storage spaces. Utah's Senator Reed Smoot, convinced a cigarette caused the fire, prepared a bill disallowing smoking in some government buildings. The Washington Post expressed outrage that the Declaration of Independence and Constitution were in danger even at the moment, being stored at the Department of State in wooden cabinets.(21)
Meanwhile, the still soggy, "charred about the edges" original and only copies of the 1890 schedules remained in ruins. At the end of January, the records damaged in the fire were moved for temporary storage. Over the next few months, rumors spread that salvage attempts would not be made and that Census Director Sam Rogers had recommended that Congress authorize destruction of the 1890 census. Prominent historians, attorneys, and genealogical organizations wrote to new Secretary of Commerce Herbert Hoover, the Librarian of Congress, and other government officials in protest. The National Genealogical Society (NGS) and Daughters of the American Revolution formally petitioned Hoover and Congress, and the editor of the NGS Quarterly warned that a nationwide movement would begin among state societies and the press if Congress seriously considered destruction.(22) The content of replies to the groups was invariably the same; denial of any planned destruction and calls for Congress to provide for an archives building. Herbert Hoover wrote "the actual cost of providing a watchman and extra fire service [to protect records] probably amounts to more, if we take the government as a whole, than it would cost to put up a proper fire-proof archive building."(23)
Still no appropriation for an archives was forthcoming. By May of 1921 the records were still piled in a large warehouse where, complained new census director William Steuart, they could not be consulted and would probably gradually deteriorate. Steuart arranged for their transfer back to the census building, to be bound where possible, but at least put in some order for reference.(24)
The extant record is scanty on storage and possible use of the 1890 schedules between 1922 and 1932 and seemingly silent on what precipitated the following chain of events. In December 1932, in accordance with federal records procedures at the time, the Chief Clerk of the Bureau of Census sent the Librarian of Congress a list of papers no longer necessary for current business and scheduled for destruction. He asked the Librarian to report back to him any documents that should be retained for their historical interest. Item 22 on the list for Bureau of the Census read "Schedules, Population . . . 1890, Original." The Librarian identified no records as permanent, the list was sent forward, and Congress authorized destruction on February 21, 1933. At least one report states the 1890 census papers were finally destroyed in 1935, and a small scribbled note found in a Census Bureau file states "remaining schedules destroyed by Department of Commerce in 1934 (not approved by the Geographer)."(25) Further study is necessary to determine, if possible, what happened to the fervent and vigilant voices that championed these schedules in 1921. How were these records overlooked by Library of Congress staff? Who in the Census Bureau determined the schedules were useless, why, and when? Ironically, just one day before Congress authorized destruction of the 1890 census papers, President Herbert Hoover laid the cornerstone for the National Archives Building.
|Even after the outcry in 1921, thirteen years later the Census Bureau destroyed the remaining 1890 schedules.|
|In 1942 the National Archives accessioned a damaged bundle of surviving Illinois schedules as part of a shipment of records found during a Census Bureau move. At the time, they were believed to be the only surviving fragments.(26) In 1953, however, the Archives accessioned an additional set of fragments. These sets of extant fragments are from Alabama, Georgia, Illinois, Minnesota, New Jersey, New York, North Carolina, Ohio, South Dakota, Texas, and the District of Columbia and have been microfilmed as National Archives Microfilm Publication M407 (3 rolls). A corresponding index is available as National Archives Microfilm Publication M496 (2 rolls). Both microfilm series can be viewed at the National Archives, the regional archives, and several other repositories. Before disregarding this census, researchers should always verify that the schedules they seek did not survive. There are no fewer than 6,160 names indexed on the surviving 1890 population schedules. These are someone's ancestors.|
The Fate of the 1890 Census, Part 2
Return to the top