Winding Your Way Through DNA Symposium
San Francisco, California
Friday Evening, September 25, 1992
James D. Watson, PhD
Harold has given me a relatively simple task of essentially repeating my
little book, The Double Helix, and so I prepared some slides from
the book, the pictures in it, and I will race through the story and then
make a few comments at the end, seeing the discovery after 40 years.
The origin of the work goes back to my days as a student at the
University of Chicago where I believe it was the fall of 1945, I read a
little book called What Is Life? by the German physicist Erwin
Schrodinger, who had received the Nobel Prize for his work in wave
mechanics. Schrodinger said the essence of life was essentially
heredity, and the key problem was, "What is the gene?" And in
particular, "How can you copy a gene?" The gene must carry very
specific instructions, and there must be some marvelous trick by which
you can exactly copy the structure of a gene. So that inspired me, the
following term at the University of Chicago, to go to the lectures by a
very famous geneticist named Sewall Wright, who gave a course entitled
Physiological Genetics. From that I came away with the feeling there
were three problems.
- "What is the gene?" Wright in his course had recounted the
experiments of Avery which suggested that the genetic substance was DNA.
- The second problem was one which Schrodinger posed, "How
is the gene copied?"
- The third problem was, "How does the gene function?"
It controls the cell, the essence of cells are proteins. There were
already some experiments which could be summarized under the idea that
every gene somehow provides the information to make protein. As a
result of Wright's course, I decided to go to graduate school and learn
genetics. I went to Indiana University and there had the good luck to
work under Salvador Dali--not Dali, but Luria. (laughter) [Aside by
Watson: There is a reason for this mistake, which is that Dali once
invited me to lunch with Mia Farrow, and I've just seen (her) latest
movie. (The lunch was) A long time ago; I didn't even know who she
I worked on bacterial viruses and did a thesis which Max Delbruck, who
was teaching at Cal Tech, told me that I had the good fortune that it
was a very dull thesis. So I wouldn't make the mistake of continuing it
but could try something better when I went on as a post doc. But I did
go to Copenhagen. I was supposed to learn biochemistry but I was bored
by it. I went to Naples for two months, and there was a scientific
meeting at which I met an English scientist Maurice Wilkins, who I
hadn't heard of before, who had taken x-ray pictures of DNA.
The marvelous thing I heard was that these pictures were very good. So
there was a well-defined structure. I had known before that DNA yielded
x-ray diffraction patterns but it was sort of murky. But this
time there was a marvelous structure that someone could find. And by
this stage I had since realized that the essence of a gene was a
molecule and genetic experiments were never going to say something deep
about the molecule, and I had to study the molecule.
Suddenly from seeing the x-ray picture of Wilkins it was clear that
someone was going to be capable of studying it and I thought, "Why not
After I'd left Naples I went to Geneva and my friend Jean Weigle, who
was also at Cal Tech, told me that Linus Pauling had solved the
structure of proteins. Linus had done it by just guessing it. I got
back to Copenhagen and wrote to Luria and said, "I want to do x-ray
crystallography. Can you help me get in a lab?" By luck Luria had met
John Kendrew, then a young English scientist who was working to solve
the crystal structure of the protein myoglobin. Kendrew was looking for
someone to work with him and he told Luria that I could come there. I
went to Cambridge and discovered I really didn't want to work on
proteins. I really wanted to work on DNA and that was helped by meeting
Francis was doing experiments and said, "Well, why can't we imitate
Pauling and just guess the structure?" That's what Pauling had
done. The rules of chemistry would lead you to the structure and we
said "Why not?" If one considered the essence of it, the essence of DNA
was it was a linear molecule. The exact chemical bonds had been worked
out actually in Cambridge just that year by the laboratory of Alex Todd
and it was essentially a nice linear molecule where it was a polymer.
The monomer was called a nucleotide and there were four types of
nucleotides depending on the four bases: adenine, thymine, guanine and
cytosine. So that was the thing. How would this molecule fold
up? That was the question we wanted to answer.
Francis was a friend of Wilkins. Wilkins came out to lunch and we
learned two things. One is that Wilkins said, "I actually thought
the molecule contained more than one chain." They were sort of
guessing there were three chains. That was a really important new fact.
The second thing he said was that there was trouble, and when he had
been in America his problem had been given to a woman named Rosalind
Franklin, and they didn't get along. But he said he didn't know what
Rosalind was doing but she was going to give a seminar in several weeks
and I could come and hear it.
I went and heard it. And I won't say, but I heard it wrong. I came
home... and we said, "Well, we will make a helix." I had heard
her talk wrong and thought there was very little water in it and
therefore it should be a rather compact structure. We tried to make a
structure. We said, "Well, there are four types of bases. There is no
way we can stick in a regular sequence of bases in the center in a
regular fashion." So we tried to build molecules where the main chain
was on the inside.
I should say that what Francis and I generally did was we talked most of
the time, or we walked. I think we walked for an hour after lunch.
We ended up with a structure in which we thought that conceivably the
phosphatidyl groups were held together by magnesium ion, and we built a
structure with three chains and it wasn't perfect but we were very
happy. We called up the people at King's College [at the University of
London], particularly to get
in touch with Miss Franklin and she said she would come out [to
Cambridge] and look at it. She looked at it and said it was all wrong.
The molecule was filled with water and we sort of retreated and we were
told in fact the problem belonged to the people at King's.
I really had nothing to do with it except grow my hair for about six
months without cutting it. This was a response to Crick's wife's
comment: she said I was bald because I had a crewcut. She said I
looked like an American serviceman so I decided to grow my hair.
We got slightly worried when we knew Linus Pauling was going to come,
because we thought he might see Miss Franklin's data. But then he
couldn't get a passport because the State Department thought he was too
friendly with the Communists. So he never came. The only visitor we
did see was an Austrian chemist by the name of Erwin Chargaff who showed
up in the middle of the summer and talked to Francis and me, and talked
about doing an analysis of the bases in DNA. He said that the amount of
adenine was approximately equal to thymine, and guanine to cytosine. In
different DNAs of ratio of adenine to guanine would vary, but what was
always constant was that adenine approached that of thymine, and guanine
Francis got quite excited about it. I didn't like it at all because
Chargaff was probably the most unpleasant person I'd ever met. I just
wanted to ignore anything the wretched man said.
Now, Francis actually then knew a chemist named John Griffith. They got
talking and Francis saw that maybe A and T could attract each other and
that might explain [Chargaff's ratios]. Unfortunately, instead of doing
it the way the base pairs are now, they put the adenine on top of the
thymine. Griffith did some calculations and with time Francis lost
interest. Interesting, Griffith was a good chemist. He was the nephew
of the Griffith who discovered transformation. But nothing really
I was doing bacterial genetics to pass the time, but then we heard about
Linus Pauling and we were really frightened that Linus would get on to
the problem and just guess it. We heard that he had a structure. We
said, "Well, he doesn't know the data at King's." So we kept telling
ourselves he doesn't have it. Then his son Peter had become a research
student at Cambridge and a manuscript [from his father] was sent to
Peter late in January. We read it and the structure was just lousy. We
couldn't believe it. It had to be wrong.
So I rushed into London and told Wilkins that Pauling proposed this. He
proposed something crazy. Maurice told me a couple of things. One,
the good news was that Miss Franklin was going to leave and he could get
back working on it. Then he showed me a picture which he had taken.
This was a perfect picture of a helix. So I saw this picture and then
actually saw Rosalind Franklin briefly and told her that Pauling had
published a structure which was wrong. She said--well, you know, she
was annoyed with me. We didn't get on.
Anyway, I went to supper with Wilkins and said, "This picture is such a
perfect helix." We kept saying it was a helix. This picture means that
we should get back and build models. Maurice was more or less saying,
"I'll do it when Rosalind leaves the lab." But I went back to Cambridge
and told Francis and told Professor Bragg that there was this
extraordinary picture, and there was no doubt about it being a helix and
we should go back to trying to build models.
So Crick and I did, and Francis said, "Why not put the bases
inside?" When he said that there were actually facts which really
said the bases had to be inside. There were two sorts of experiments.
One is that DNA can be denatured. If you vary the pH to break the
hydrogen bonds, the molecule collapsed. You could see that this is
viscous and then as you raise the pH it falls apart. Then there was
another experiment in which you essentially tried to titrate the
bases. This had been done by Gullen and published before I got to
Cambridge. If you look at it carefully it says all the bases are
hydrogen-bonded. They can't just be sitting out interacting with water.
So the bases must be in the center and hydrogen bonded so all the groups
can't titrate as if they weren't hydrogen bonded. We said, "Why not
two chains?" When you looked at the data it didn't have to be three
chains, but it certainly wasn't one. Maybe we could form pairs between
adenine and adenine, and guanine and guanine, and you could form a
structure by which each of the bases attracted itself.
I was extremely excited because I thought maybe this was right. But
then there was a real chemist in the lab, Jerry Donohue, who had been
with Linus Pauling. Now this was the structure that I was manic about
for a week. But Donohue dismissed it as being totally wrong because he
said that all the textbooks have written the thymine and guanine in the
wrong tautomeric form. They're ketos, not enols.
Now deep down he was slightly bluffing, because he just knew one
structure which was keto, but he at least more or less raised doubt that
maybe the textbooks were all wrong. So we wrote it the wrong way and I
had made paper models of the bases and changed around where the hydrogen
atoms were. Then one Saturday morning which I think was March 1, I
found you could make two base pairs where adenine and thymine, and
guanine and cytosine had exactly the same shape. Francis found you
could flip-flop them over in a nice way of symmetry so that you could
have all four bases on a different chain.
And there it was that you could build a structure. This was sort of the
way we wanted to build it after the base pairs were found. Francis
stopped working on proteins and actually built the model. This was the
reason we were really excited, because you could imagine if you
separated the two chains that you could form a complementary one. So if
this were the structure of DNA, the way DNA was copied was probably
So we were, to say the least, happy.
Then the next thing I did was to write Cal Tech and Delbruck and tell
them the news that we had found the molecule. The last sentence was,
"Don't tell Pauling." I said this because we hadn't written the
manuscript and I wanted to write the manuscript and then tell Linus.
But in fact the first thing Delbruck did when he got the manuscript was
to tell Pauling. But Pauling was very gracious and said, "It's a lot
better than my structure." So really when everybody saw it everyone
liked it. There was no reason why it wasn't right and if you really
looked at it the x-ray data confirmed it. And the answer was yes.
Now there is actually after the manuscript had come out we were locally
famous and the local Cambridge paper sent a photographer and we were
photographed, and I'm changed there slightly [alluding to the picture
being shown to the crowd]. Crick's wife said I had to actually look all
right, so she went to the store with me and I got a blazer, tried to
look okay, and Francis is sort of tweedy. There was a photograph of us
drinking coffee in the room where we had talked for 18 months.
So this is really the encapsulated story. Now one thinks back about it.
We could have solved the structure in about a month, if we had taken the
chemical data which really said the bases were hydrogen bonded. It
wasn't published obscurely, it was right there. So why didn't we do it?
I think the answer is that neither Francis nor I knew any chemistry.
And so we tried to avoid chemistry. And you could say, "Well, why
did you succeed?" And I think the answer is there were two of us.
And we were interested.
Then you go on, "Well, why didn't Rosalind solve it?"
Interestingly, the answer probably comes down to the same thing. She
wasn't a chemist. She wasn't an x-ray crystallographer. She was also
an amateur. Because her data in the crystalline form showed a dyad, and
she didn't understand what it meant. It really said one chain is going
up and one is going down. So she (had) worked on Co-A and (had) never
solved a structure before in her life, and she really wasn't really
familiar with space groups. I think the other thing is she had no one
to talk to. She had a research student who didn't know
crystallography, a friend name Luzati who was slightly bizarre,
and bright, very bright, but he didn't know crystallography.
But Francis, when he saw that dyad--Francis actually had learned
crystallography and he went to Cambridge and we didn't need it but
Francis knew that dyad had proved the structure. So, he knew the space
group. Rosalind never knew what it meant. So she could have got it, if
she had been trained in the subject.
The thing you can't answer is, "Why didn't Pauling get it?" And
I'm still as much at a loss now 40 years afterwards as I was then. But
it was good luck for me.