Finding a treasure trove of documents about the family of one of the earliest cases of autism has led this column to offer two observations: Mercury may be associated with the disorder from the beginning, and cutting-edge research near the nation's capital may help explain why it was first discovered at Johns Hopkins University in nearby Baltimore.

There is another possible clue from that early case, and it bears directly on the observation by many parents that restrictive diets seem to improve autism symptoms in affected children.

Specifically, those parents have found that a so-called GF/CF diet -- one free of gluten-containing grains and casein-containing dairy products -- helps clear up both behavioral issues and physical maladies like disrupted intestinal tracks. The grains in question are cereals -- gluten is a protein in wheat, rye, barley and most oats.

What does that have to do with the information this column recently uncovered about Case 2 -- a child known only as Frederick W. in the original medical report on autism published in 1943?

Well, we identified his father as a prominent plant pathologist named Frederick L. Wellman, who at the time his son was born in 1936 was a senior scientist at the U.S. Agriculture Department's main research center in Beltsville, Md., a suburb of Washington and only about 30 miles from Hopkins.

Wellman was trying to stop fungi from killing plants, and his resume shows that he was working on fungicides at that point in his career -- including newly developed organic mercury compounds. We found this especially interesting because Wellman had sales brochures in his archive for fungicides made with ethyl mercury. That is the same species of organic mercury used in a vaccine preservative called thimerosal, which some parents and a minority of researchers blame for triggering the recent rise in reported cases of autism.

In the 1920s Morris Kharasch, an organic chemist at the University of Maryland adjacent to the Beltsville research center, filed a barrage of patents that paved the way for both compounds. His dual focus was evident in his "Who's Who" entry: "awarded patents along pharmaceutical lines, and treatment of fungus diseases of small grains."

Both thimerosal and ethyl mercury fungicide first came on the market about 1930; the first autism case in that original medical report dates from 1931. We also identified Cases 1 and 3 and found plausible connections to fungicide in those cases as well.

Of course, this is speculation, and some readers have understandably challenged it as highly hypothetical. And that is admittedly true -- these are not answers we're offering, but questions. For example: Would a review of those early cases find environmental links overlooked in the earliest reports? Does the idea that mercury might be associated with autism, but not necessarily via vaccines, suggest mercury could be a factor even now, after thimerosal has been removed from most pediatric vaccinations? There is no question our planet -- our air, our streams, our fish -- is increasingly mercury-toxic and, according to the Environmental Protection Agency, putting thousands of American children at risk of developmental disorders.

Those amount only to hypotheses -- but they are testable ones; my suggestion all along has been that reasonable tests have been ignored, perhaps even avoided. A classic example is whether autism is less prevalent in never-vaccinated Americans such as many of the Amish. After I raised that issue and cited anecdotal and unscientific reports that it might in fact be, the critics howled -- but two U.S. representatives, a Democrat and a Republican, introduced a bill ordering the Department of Health and Human Services to find out.

So on to the diet question. Frederick L. Wellman's archive, at North Carolina State University, has brochures for those two ethyl mercury fungicides. The most interesting is New Improved Ceresan, which is for use on "wheat, oats, barley, rye, sorghums, millets and flax."

Wheat, oats, barley, rye ... where have we seen that list before? In the gluten-free autism diet, that's where.

"Gluten is a protein and is contained in foods, such as wheat, barley, rye and oats," according to "At the present time, we do not know why the gluten/casein-free diet helps many autistic individuals." One theory is that they release opioid-like substances in the gut that can migrate to the brain.

Well, here's another hypothesis -- could some of those grains be grown in places where residual toxins -- ethyl mercury, say, but in fact any environmental toxin -- are getting into them and thus into our kids? And if some child's body burden or susceptibility is already at the tipping point, could that aggravate or even induce physical and mental symptoms that go by the name of autism?

Sounds fantastical, but one thing we learned about mercury by researching Case 2 is that this stuff can stick around. A government study in the 1990s at Beltsville, where Frederick W.'s father was experimenting with mercury fungicides in the 1930s, found concentrations of mercury 2,000 times the U.S. average. And that was presumably decades after anyone messed with mercury there -- mercury fungicides were phased out in the 1970s after scientists recognized their toxicity extended well beyond fungi.

This is speculative, yes. But so is the idea touted by serious scientists that autism is triggered by excessive "maleness" in the developing brain; so, for that matter, is the belief that autism is a genetic disorder with no environmental trigger at all.