From oil-turbine to Model E: early days in retrospect.

This account deals approximately with the period 1940-1960, in which time Svedberg's invention of the first reliable ultracentrifuge was making a great contribution to developments in physical biochemistry. In the 1930's, only at Uppsala could an ultracentrifuge experiment be conducted, but by the 1960's electrically driven machines-the "Model E"-were widely available, particularly in the USA. The instrument was used extensively to follow fractionation methods for purifying proteins and other biological macromolecules and to determine their sedimentation coefficients and molecular weights. Here the author calls upon some of his own experiences, having been able to work with some of the now-famous contributors to the subject. As an analytical instrument, a good example was the proof by a velocity run that a particular sample of DNA was free of protein. This was essential in allowing the viscosity and titration characteristics to be ascribed to the hydrogen bonds formed between purines and pyrimidines-an observation subsequently proving critical. The Model E was easier to use than the oil-turbine, but discrepancies at first appeared between values for the sedimentation coefficients found from the two machines. These were subsequently resolved when attention was directed to rotor temperature uncertainties.