Sanborn fire insurance maps are detailed city plans, usually at scales of 50 or 100 feet to an inch. They show individual building "footprints," complete with construction details, such as building material (brick, adobe, frame, etc.), height (of larger buildings), number of stories, location of doors, windows, chimneys and elevators, use of structure (dwelling, outhouse, hotel, church, etc.), street address, and occasionally the ethnicity of the occupants. Other features shown include lot lines, street widths, water pipes, hydrants and cisterns, and fire-fighting facilities.

       The purpose of this article is to share with the reader some miscellaneous information about Sanborn Company fire insurance maps which I have accumulated over the years, and which might be useful in interpreting the scope and content of these cartographic treasures. This information is certainly not unique, and in many cases represents my own personal conclusions.

       Thematic maps produced for fire insurance rating purposes contained a wealth of detail associated with their original intended use: to assign a premium for the policy commensurate with perceived risk. The risk assessment was a judgement call on the part of the policy writer, based on information presented in the "report," which consisted of detailed urban plan and a short narrative description of water supply, prevailing winds, and fire-fighting facilities. In the United States, the Sanborn Company map dominated the industry which produced these "reports" to the point where "Sanborn" and "fire insurance map" became synonymous.

       Before the turn of the century travel was time-consuming, and agents in one town either had to travel to another town to inspect the property, or were obliged to base the premium on whatever information they were provided by the property owner. Agents in possession of fire insurance reports, however, rarely needed to personally inspect the property, because the maps provided sufficient information, presented in a uniform manner, to assess risk.

       Thus, as the maps became obsolete for insurance purposes, the very detail that was so essential to fire policy-writing turned into an archive of historical detail. As-built town configurations, presented in a uniform manner, allowed comparisons among communities mapped at different times in diverse locations. This context, the "wide-area uniform eye-witness historic primary record," provides the general framework for some of the interpretive and anecdotal information which follows. It is based in some instances on personal experience and on close association with Sanborn Co. map reproductions.

       Generally, scaling accuracy is excellent, except when it is noted that the street is "widened." This was to give the insuring agent room to write when annotating the map directly, giving names, values, policy numbers, etc.

       One spectacular case of gross distortions is noted. Upon comparing Old Albuquerque NM for 1902 (sheet 17) prior to platting, to the same area for 1908 (sheet 30) after the town was platted, I noticed some buildings were "re-configured," rotated, and dis-placed. I suspect that the 1902 map was correct, and that the 1908 map was purposely distorted when the locations of the buildings were found to straddle property lines as determined by the plat survey. Other towns when confronted by similar situations simply showed things the way they were: houses in the middle of a proposed street, property lines running through obviously permanent buildings, etc.

       A Petersburg AK attorney, whose house and office jutted out into a current street, obtained a set of maps from me and in 1986 successfully defended his position in Superior Court that the house was there before the street, and saved his house from condemnation by the city. An adjustment was made to the street.

       The compass rose on Sanborn Company maps is not a reliable indicator of true north. If the base map was a superposition of the fire survey over a plat map (which was the general rule if a plat map was available), then the compass rose seems to be as accurate as that of the plat map. Sometimes a plat map obviously existed, but might not have been easily available to the Sanborn surveyor. In the case of Tus-carora NV for 1890, the plat map, the mining claims diagram, the Sanborn map, and personal sighting of Polaris all disagreed with one another by substantial margins, well past any magnetic declension errors. I attributed these anomalies to human rather than systemic error and did not research it further.

       In the late nineteenth century, spelling consistency of toponyms was not always achieved. In some cases I suspect that popularized variants, misuse, mispronunciation, and misspellings by the population at large caused orthographic shifts that fed place name legends. Examples of variants in spelling sometimes show up in street names: Bonanzo or Bonanza St.; De Visadero later becomes Divisadero. In Mokolumne Hill CA an early map showed both Legger Street and its main hotel, the Legger House. The Hotel changed hands and became the Hexter House. Later on it became the Hotel Leger, an obvious attempt to gallicize its image, while the street continued to be known as Legger.

       However, the names associated with the ownership and/or lease of warehouses as shown in Dakin's California Warehouse Book (1897) are probably faithful renditions of official records or names obtained from on-site interviews, hence less subject to mutilation by the masses. This may be of some comfort to genealogical researchers.

       Town names were not exempt. Hayward CA was know as Haywards; Centerville CA was Centreville. A relatively common shift was to first drop the apostrophe, then the "s" (Hayward's to Haywards to Hayward). There are also many instances of towns changing names; it was common practice to indicate the former name at the first instance of a map with the new place name.

       In a related matter, some Sanborn Company maps were issued for a group of communities, under the name of a dominant one. Sometimes the first sheet would not list all of the places mapped, especially if they were small. Sometimes the dominant town for a multi-town map would change, and the same geographical area in later years would be indexed under a different town name. To confuse matters further, name changes might take place within the subordinate towns. The lesser towns were invariably relegated to the last sheets of a map set.

       Further fuzziness abounds in place names which started out as wineries, factories, lumber yards, mines, train stations, mills, post offices, landings, junctions, etc. Because they contained insurable structures, they were often depicted at the tail end of map sets without detailed attribution except as "specials" on the index sheet. The Sanborn maps consistently indicate the location of these orphans. It is unfortunate that the coordinates of many entries into Dakin's California Warehouse Book are not given, because Dakin provides a wealth of other attributes not found on Sanborn maps: date of construction, capacity, and nature of commodities stored, together with the name of the owner and/or operator.

       The wealth of historical detail found on Sanborn Company maps is sometimes clouded by the abbreviations used. A list was developed over several years by trying to find either legends unique to some maps, or finding unabbreviated notations adjacent to similar buildings with the abbreviations. Minor variation due to regional usage of some terms have been encountered. An example is the use of "Chine" and "Chinese" interchangeably. There is also some uncertainty about abbreviations such as "S" and "Sto"--there is some evidence that they had multiple uses, but only among maps made by different survey teams in divergent locations and years. The list below was developed to help unravel the occupancy and usage associated with the structures. This list is neither complete nor applicable in all cases; it serves only as a guide.

       Other conventions include a large "X" on the structure, from corner to corner. That is a barn or stable. A small "x" and the numeral 1 depicts a shingle roof of a one-story structure. A dwelling with an attic or gabled roof is designated as 1S. A one story structure with a basement is 1B. A dashed line on the outside of a structure represents an overhang with open sides; typically along streets these were balconies with wooden sidewalks underneath. A large black dot in the street is a fire hydrant. Four hash marks on the side of a building indicates that the top of the wall is 24 inches (6 inches for each hash mark) above the roof-wall joint.

       One may ask, why was that important? Initially, these fire-breaks were meant to stop sparks from an adjoining fire from igniting the roof; later is was found that debris collecting at this juncture could either contribute to the hazard, or when wetted down, retain moisture to retard the spread of fire, especially in business districts. Movies show western towns with "false fronts"--squared-off front walls of wooden buildings. These false fronts were spark arresters. Main streets of many western towns were 80 feet wide for two reasons: to prevent fire from jumping the street, and to allow wagons drawn by teams of horses to make a U-turn. This included horse-drawn fire apparatus and pumpers. The notation "No Exposure" on the map border meant that there was no danger of fire spreading from that direction. In some cases, the distance to an unmapped structure would be given, e.g., "Hall 65 feet beyond."

       Sanborn maps amplified on these attributes, often indicating relative elevations, annotation as to passableness, and the nature of the material of which the roofs were made. Sometimes even the fire-breaks within buildings (e.g., "Sand between floors," or "Brick inside walls") attested to a superior resistance to fire.

       The Sanborn surveyors doubled as fire-inspectors of their time, examining suspect structure on-site. It was prudent to allow that, especially for business owners whose buildings were sound and who did not wish to be red-lined for insurance purposes. Those who did not cooperate paid the price. "Entry denied," proclaimed the legend. "Said to have try kettle (an fat-rendering vessel, quite flammable)." The Angels Camp CA 1898 Sanborn map singles out two groups of structures. One is labeled "Never Insure," the other "Do Not Insure." Although Sonora CA and several other nearby towns were mapped extensively, nearby Columbia CA was not, allegedly because each time the surveyors came through, they were run out of town, because the whole place was a firetrap, and the town's mayor wanted no record made of this condition. And finally, the readiness of the water supply to put out fires in Virginia City NV (1890) was confirmed for a particular location by indicating that the pipes were insulated by "four inches of charcoal to prevent freezing."

       One sometimes wonders why some places were mapped, when it is obvious that there was very little left of the town, and that it was only mapped once. The answer is, to discourage arson for insurance purposes. If a place was out of the way, it was unlikely that an agent would travel there, and if the town was on the decline, there was economic incentive for insurance-related arson. Two vivid examples of this are Bodie CA and Aurora NV, both mapped once in 1890. The towns were on the decline, buildings were being abandoned, but the towns were far from empty. About a third of the buildings in both Bodie and Aurora were shown as vacant. Agents in possession of these maps would be reluctant to insure such buildings; owners would not try to claim that they had thriving businesses and then the building burned down, mysterious-like. After the cyanide leaching process was introduced in about 1893 the towns experienced a minor comeback, but were not mapped again for insurance purposes.

       I have spent many hours explaining to users of Sanborn maps various features and trivia that sometimes are not obvious, unless you known where to look. For collectors of antique bottles who often excavate outdoor toilet pits and privies, I point out that Ione CA at the turn of the century had over 90 mapped outdoor toilets, that are scaleable on the map to within two feet. For western gunfight buffs, I show the 1886 Tombstone, Arizona Territory map and pinpoint the location of the start of the infamous shootout "at the O.K. Corral" (not in the O.K. Corral), which actually started (according to the inquest records) in a side yard between a boarding house and Fly's Photo Studio, and a few seconds later turned into a running gun battle up and down the streets of Tombstone, lasting about half an hour. (There was an eight foot wall between the driveway to the O.K. Corral and Fly's Photo Studio, preventing access to the corral from the side yard where the shooting started.) For San Francisco history buffs, I show the land bridge to the Seal Rocks (long gone), the Opera House in the Mission District, a kerosene furnace for inflating sightseeing hot air balloons in the Haight-Ashbury, the "Incurables" ward at the Children's Hospital on California Street, and the old shoreline in the Marina district. This is from the approximately 700 sheets of the 6-volume 1899-1900 Sanborn atlases for San Francisco.

       I receive requests from people doing environmental impact studies about potential toxic waste locations, which, perhaps by coincidence, seem to favor industrial sites abandoned between WW I and WW II. The massive collection of updated (via paste-over) maps associated with the 1950 census often provides clues in the form of paste-overs showing obliterated industrial sites. On several occasions I have successfully printed, at high contrast and long exposure, copies from the microfilm of such maps where the outline and legends of the site can be made out through the pasteover. This is obviously not necessary if copies for all previous editions are available. This is just an aside to indicate that sometimes faint images may be extracted from existing microfilm of paste-over corrections.

I find it satisfying to discover sometimes surprising fragments of the demographic signatures imbedded in the depiction of a bygone as-built town configuration, recorded to the exacting standards of a by-gone era, and muse about the potential of current GIS data bases enriching the research of future historians. I hope that some thought will be given to archiving GIS interpretive material, so they won't have to guess about how the data bases were contructed nor about what the abbreviations meant.

BIBLIOGRAPHY

Hayward, Robert J.

Fire Insurance Plans in the National Map Collection. Ottawa: National Map Collection, Public Archives of Canada, 1977.

 

Hoehn, R. Philip.

"The Dakin Publishing Company and its Fire Insurance Maps," Western Association of Map Libraries Information Bulletin 7:3 (1976) 47-59

 

----------, and Wm. S. Peterson-Hunt and Evelyn L. Woodruff.

Union List of Sanborn Fire Insurance Maps Held by Institutions in the United States and Canada. Occasional Paper No. 2-3. Santa Cruz: Western Association of Map Libraries, 1976-77. 2 vols.

 

Library of Congress. Geography and Map Division.

Fire Insurance Maps in the Library of Congress: Plans of North American Cities and Towns Produced by the Sanborn Map Company. Washington: 1981

 

Ristow, Walter W.

"United States Fire Insurance and Underwriters Maps, 1852-1968," Library of Congress Quarterly Journal 25 (1968) 194-218.

 

Shkurkin, Vlad.

"Fire Insurance Maps as Primary Historic Records." A paper presented to the 1984 Annual Conference of the HISTORICAL SOCIETY OF NEW MEXICO, Sagebrush Inn, Taos, New Mexico, 27 APR 1984. 4pp. ISBN 0-932732-41-0.

 

Univ of Calif at Berkeley.

UC Earth Sciences & Map Libraries Home Page . Sanborn and Other Fire Insurance Maps.

File v930104d.doc 07 JUN 97