Flood Record — June 1921 — Pueblo Probability Archive — Branch Investigation

The Flood Heroes

The Arkansas came up in the night. By morning, certain individuals had been in more places than any single person could have covered. The record does not reconcile this.

Obscura Home City Archive — Pueblo, Colorado — Flood Record Series

On the night of June 3, 1921, the Arkansas River crested at a height the city's engineers had classified as structurally impossible given the levee system then in place. The water came fast enough that the official warning system — a telephone relay originating at the Canon City gauge station — did not reach the lower residential districts before the first inundation. What followed was not an organized rescue. It was a series of independent decisions made in darkness, in moving water, by individuals who had no coordination with one another and no organizational authority to act.

The archive's interest in the flood record is not in the heroism itself — that is documented, acknowledged, and by any measure exceptional. The archive's interest is in the geography of the heroism: in specific individuals whose documented presence, cross-referenced across independent accounts, produces a spatial and temporal distribution that the archive cannot reconcile with a single person's physical capability to move through a flooded city.

This filing documents those individuals. It does not explain what it documents.

The Night — Reconstructed

The flood peaked between 11:40 PM on June 3rd and approximately 2:15 AM on June 4th. During that window, the lower districts of Pueblo — the Bessemer neighborhood adjacent to the CF&I works, the Grove district along the river's south bank, and the railroad yards to the east — sustained catastrophic inundation. Water reached twelve feet in some residential blocks. Current velocity in the street corridors was sufficient to overturn vehicles and demolish frame structures.

Rescue accounts from that night were collected in three separate waves: newspaper interviews conducted in the days immediately following; a formal inquiry by the Pueblo city government in August 1921; and an oral history project undertaken by the Pueblo County Historical Society in 1947 and again in 1962. The archive has cross-referenced all three bodies of testimony.

The accounts agree on the broad facts. They diverge on the specific question of who was where, and when.

The Irreconcilable Accounts

The following individuals appear, by name, in independent accounts placing them at geographically separated locations within timeframes that do not permit conventional transit between those locations. The archive presents these not as evidence of any particular mechanism, but as a documentation of a discrepancy in the record that has not been resolved by any conventional explanation the archive has encountered.

Elias Drummond

Documented in the Grove district pulling a family from a collapsing porch at approximately 11:55 PM by two surviving members of that family (1921 newspaper interview; 1947 oral history). Documented simultaneously — same timeframe, within the same inquiry — at the railroad depot on the east side, three-quarters of a mile distant, by a depot worker who identified him by name and described his clothing. The depot account was submitted independently and was not connected to the Grove account until the 1947 review.

Two Locations · 11:50–12:10 AM · 0.76 mi separation

Constance Aldrich

A woman identified by first and last name in three separate 1921 accounts as having guided flood survivors to elevated ground via a specific alley route in the Bessemer district. The same name appears in a 1921 account from the north bank levee road — over a mile from Bessemer — within the same one-hour window. The accounts are specific enough that misidentification across all three is considered unlikely by the 1947 reviewers. No resolution was offered.

Three Accounts · 12:15–1:10 AM · 1.1 mi separation

Thomas Okafor

A CF&I millworker identified in the company's own post-flood personnel record as having assisted in evacuating the Bessemer pump house at 1:00 AM. His name also appears — submitted by a different party to a different inquiry — in a survivor account from the southern residential district, placing him there between 12:45 and 1:15 AM. The pump house and the residential district are separated by moving flood water of the kind that, by the engineering assessments conducted afterward, was not crossable on foot during that window.

Two Accounts · 12:45–1:15 AM · Water barrier present

The archive notes: the 1921 inquiry did not flag these discrepancies. The reviewers in 1947 noted the Drummond and Aldrich accounts as "requiring clarification" but offered none. The 1962 oral history project collected one additional account referencing Aldrich by name at a fourth location; that account was filed without comment.

The archive further notes: misidentification in conditions of chaos and darkness is expected and well-documented across disaster records generally. The archive does not claim the accounts are accurate. The archive claims they exist, that they were submitted independently, that they have not been reconciled, and that the pattern of irreconcilable placement applies specifically to individuals from a city whose probability record already shows a statistically notable rate of consequential presence.

Reviewer Note — Pueblo County Historical Society — Flood Inquiry, 1947

Re: Drummond and Aldrich Accounts — Flagged for Clarification — No Resolution Filed

We have reviewed the 1921 newspaper accounts and the formal inquiry transcripts alongside the oral history interviews collected this spring. The Drummond account presents a genuine geographic impossibility if both sets of witnesses are accurate. We cannot find a reason to discount either. The individuals who placed him in the Grove district are the family he is said to have rescued; they have no motive to misplace him. The depot worker who placed him at the yards had no knowledge of the Grove account and did not read the newspapers.

The Aldrich accounts are more numerous and therefore more difficult to dismiss by misidentification alone. Three separate individuals, in three separate interviews, used her full name without prompting. Two of them described the same physical detail — a dark coat with a torn left cuff — which was not published in any newspaper account we have located.

We have flagged both for clarification. We have no clarification to offer. We are filing this note so that the absence of an explanation is on record.

— Review Committee, Pueblo County Historical Society Flood Oral History Project, 1947 — Names of reviewers on file with the Society

The Flood in the Probability Archive

The 1921 flood is the single largest cluster of irreconcilable accounts in the Pueblo probability archive. The archive has catalogued eleven named individuals whose flood-night presence cannot be fully reconciled across independent accounts. Three of those — Drummond, Aldrich, and Okafor — present the clearest documentary discrepancies. The remaining eight are flagged but require additional source verification before the archive will include them in the primary record.

The archive does not have an explanation. The archive has a pattern, documented across testimony that was not collected with any awareness that a pattern was being assembled.

Flood Record — Pueblo Probability Archive — Branch Investigation

This filing draws on newspaper accounts from the Pueblo Chieftain, June 4–14, 1921; the City of Pueblo Flood Inquiry Report, August 1921; and oral history transcripts from the Pueblo County Historical Society flood projects of 1947 and 1962. All named individuals are drawn from those sources. The irreconcilable accounts are documented in the sources as described. The archive has not added claims beyond what the sources contain.

Flood Record June 1921 Pueblo County Probability Pattern Active Series