Every Sonoma county schoolkid knows well the story: A suspicious fire in 1913 destroyed the mansion built for Jack London just days before the celebrity author and his wife were to move in. Rumors of arson circulated for decades until a 1995 study concluded it was started by spontaneous combustion of oily rags left by careless workmen. I completely believed that theory until I saw the newspapers from the day and began reading primary sources about the incident. Now I’m convinced it was arson after all – and it seems pretty certain whodunnit.

To begin, we have to look at problems with the 1995 report, written by a team of crack forensic investigators. Unfortunately, we can’t read the full document – neither the county nor the SSU library has a copy – although there is an 800-word summary published at the same time (some additional details can be found here). Should the whole report turn up and contain info that contradicts my assumptions I’ll gladly publish corrections here.

Their main conclusion was spontaneous combustion caused the disaster due to discarded cloths soaked in linseed oil used to finish interior woodwork. Natale Forni, masonry contractor of the project, also blamed spontaneous combustion but thought it was because of the “turpentine-saturated waste with which the woodwork had been rubbed down.”*

Turpentine? Linseed oil? Which was it? The answer is both – mixed with other ingredients. And this reveals a weakness of the 1995 top gun team: They included engineers and criminologists, but apparently no architectural historian.

Let me interrupt here to mention my other preoccupation: Old house restoration. Over the last decade my wife and I have returned Santa Rosa’s 1905 Comstock House to mostly its original condition and along the way I’ve immersed myself in the study of turn-of-the-century wood finishing methods. Just as Sherlock was always hectoring Scotland Yard detectives about different types of tobacco ash, I can bore anyone to tears about those archaic varnishing formulas (unfortunately, I can even tell you how some of them taste). Visit one of my other blogs, “Restora Obscura” for more on this topic.

As the workmen were in the final stages of finishing the woodwork, they would not have used pure linseed oil inside the house. That stuff was only applied to wood as a first coat before it was brought inside and put into place (in order to cover the back and ends of the piece as well). When it came to finish work, they would have used a recipe of linseed oil and turpentine along something else, depending on how they wanted the surface to look and be preserved. Most probably it was the old standby 1:1:1 mix of the two with beeswax, which forms a paste (think shoe polish) that fills any imperfections and polishes up beautifully.

But the 1995 investigation didn’t explore waxy rags, instead considering linseed oil alone. As stated in the summary: “A loosely piled handful of cotton rags dampened with boiled linseed oil has been shown by our experiments to be capable of self-heating to flaming ignition in a few hours.” Well, yeah, that’s a popular middle school science fair project. Visit YouTube and enjoy hours of videos showing little autoignition bonfires with linseed oil rags. (My personal fave is the one where the guy got bored because it took too long and left the smoldering pile, only to conclude the video by showing a lump of charcoal. “The wife came home and said ‘your rags are on fire.’ So. There you go.”)

Those home videos always show test rags piled on concrete driveways or in Weber grills, places where heat will collect on sunny days. That’s the other important factor in getting rags to burst into flames – the higher the ambient temperature, the more likely it is a fire will start. And the investigators emphasized that night was extremely warm: “The one striking observation of London’s neighbor was that the night of the fire was the hottest night in memory,” supposedly still over 100° after dark.

That is absolutely false. According to the weather reports in the Press Democrat and other Bay Area newspapers it was a typical August day, with highs in the low 80s. There was a light westerly wind in the afternoon and marine layer fog overnight.

The mistaken neighbor of London’s was thinking about the record-breaking heat that happened six weeks earlier with the hottest day in Santa Rosa history – 113 degrees at one o’clock, with 126 recorded in direct sun (see “HUGE WILDFIRES, DROUGHT, RECORD HEAT: THE AWFUL SUMMER OF 1913).”

The provable temperature gaffe and likelihood the investigators tested the wrong material are reasons enough to question spontaneous combustion as the cause. But what about arson? They concluded it was a “low probability” because there was apparently just a single point of origin – the high-ceiling dining room, where there was no certainty the fire would spread and destroy the whole building. Nor would a dedicated arsonist be content with setting a single blaze, they concluded.

It’s now understood there are six motives for arson, as defined by The National Center for the Analysis of Violent Crime (part of the FBI). One of the motives is revenge. According to the Center’s textbook, a subtype involves personal retaliation, where the fire is “intentional, pre-meditated and targeted. Although he may sometimes set a fire impulsively…He blames his actions on conflict with others that leads to feelings of anger while setting the fires.” The fire will usually be “set using available materials found at the scene” inside buildings were there is open access.

Ladies and gentlemen, let me introduce you in the following article to Jack London’s troubled brother-in-law Captain James Shepard, who quarreled with Jack a few hours before the fire.

1 Stone, Irving; “Sailor on Horseback,” 1938, pg. 300


Photo courtesy Sonoma County Library

Leave a Reply

Your email address will not be published. Required fields are marked *