John Edsall: The most dramatic example of that was Svedberg developing the ultracentrifuge. I
remember very well a seminar that Svedberg gave us, and it was just about 1930. He
had been developing, he started developing the ultracentrifuge around 1920 or maybe
a little earlier. It was, of course, a formidable project to develop a centrifuge
that would give these very high angular velocities, and also have an optical system
in which you could observe what was going on in the machine. But it was quite a few
years that it took him to develop this. He was in the rather early stages of development
when Dr. Cohn came to visit him about 1920 or so. It was just after the first world
war, Cohn had gotten into work on proteins and had then got a fellowship from the
National Research Council that allowed him to go abroad and study with Sørensen at
Copenhagen, who was then the director of the Carlsberg Laboratory.
And before he came back to the United States, Cohn, having heard about Svedberg working
on the development of the ultracentrifuge, he went and talked with Svedberg about
this. And in this seminar in 1930, that Svedberg gave us when he visited Cohn’s lab,
he began by recalling this visit from Cohn ten years earlier. He said, "We had a discussion
then and we disagreed. I said that I expected - I said I was very eager to study proteins
in my ultracentrifuge as soon as it was working properly." He said, "I expected to
find that a preparation of almost any protein would give a very diffuse boundary corresponding
to molecules of various sizes. Cohn disagreed with me, and said, ‘I predict that
you will find that they are really homogenous in size if you've prepared your proteins
properly.'"
And Svedberg, who first took up hemoglobin after the first one he studied, said, "Of
course, I very soon found that Cohn was right and I was wrong." And then he became
one of the most ardent advocates of the view that proteins were real genuine molecules.
In fact, I remember there was a conference on protein structure involving Wrinch among
other people held by the Royal Society in London in 1938, in which Svedberg said emphatically,
"I believe that proteins are true molecules with a definite structure and I think
if you alter a single atom you may be able to detect the difference."
