I was invited to build a five-foot model of tobacco mosaic virus for Expo 58, showing the public the structure my group had found, made from table tennis balls and bicycle handlebar grips.
When an asymmetrical A-DNA image in May 1952 contradicted the accepted helix, I followed the data, and marked it with a mock funeral notice for the "crystalline" helix.
In 1947 I joined Jacques Mering's laboratory in Paris, where I learned to apply X-ray diffraction to amorphous, disordered substances such as carbon — the skill my later research relied on.
I won £10,000 over three years from the US National Institutes of Health — the largest fund Birkbeck had ever received — and began determining the crystal structure of the polio virus.
At Birkbeck I used X-ray crystallography on tobacco mosaic virus and showed in Nature (1955) that all its particles are the same length — against the eminent virologist Norman Pirie — then led the research on its structure.
In a November 1951 seminar I presented the A and B forms of DNA, placed the phosphates on the outside and specified the water content — facts that grounded every later model.
At the British Coal Utilisation Research Association I studied coal porosity using helium to measure density, linking pore constrictions to permeability. The research became my 1945 Cambridge PhD.
My method for getting usable X-ray diffraction images of DNA fibres: a refined fine-focus tube and microcamera, a humidity-controlled chamber, then Patterson analysis of the pictures.
I held that model building in structural science should start only once enough of the structure is known from experiment. Two failed triple-helix DNA models bore that out.
My rule for structural research: wait for the diffraction evidence instead of speculating, and let the X-ray photograph say what the DNA structure is.
