Well, I don't know if it could have been avoided. I carried the study
of conjugation in bacteria to a very considerable point, precisely because
I was the main exponent, I felt I had to take a very conservative view,
not to invoke hypotheses or processes for which there was not firm evidence,
and I think partly for that reason, and partly because there's just so much
any one person can do, the later steps of understanding this process fell to
Francois Jacob and Elie Wollman, and I would have enjoyed being able
to do that too, and they were the ones who did it. And, there was a certain
rigidity, I was building a system, I had to defend many aspects of it
against a lot of critics who didn't understand it at all, and in the process
of doing that after a few years, you became locked into one specific view,
and I've learned to be a little more flexible, I hope I have, dealing with
problems a little later on. So, I think that's the place
where I might have added one more brick to the edifice, but why should
one person do it all?
Well...but they were very brilliant experiments
on Jacobe's part, but I had had the idea that conjugation literally meant
that two cells -- I was visualizing something like paramecia -- you had two
intact cells, they exchange nuclei, they might be symmetrically equivalent,
but at least it would be like fertilization, where a gamete delivers and
entire genome into the cell, and I was running into some difficulty
in interpretation of the genetic data, but I was finding all kinds of
complicated ways -- it got to be a little bit like Ptolemy's epicycles --
of embracing the data within the framework of that model, and what Jacob and Wollman discovered was that -- originally it was a hypothesis,
and I didn't put much stock in it at first, because at first there wasn't
that strong data for it -- but they thought that the chromosome is transferred
progressively from the male cell to the female cell, and that it takes
about 100 minutes for the process to be completed, and that many of these
transfers are interrupted before it's completed,
so the act of fertilization in
E.coli
involves sometimes only tiny bits of DNA, sometimes larger bits,
sometimes larger ones, and then very very rarely you have pretty much
the whole chromosome coming in. So that was their alternative model. I
should have been both more ingenious, and thinking of it first, and
pertinent to the knowledge that they in fact built a pretty good case for it.
So, all that happened 30 years ago, in the mid-50s when all of that was
ironed out. Beyond that, well, one thinks of all the experiments one
might have done and didn't get around to. I don't know if you call that
failures or not. I can't complain. There was one thing I can recall. I was part
of the planning for the Viking landing mission on Mars, and we were so
imbued with the idea that there would either be no signal, or there would be
a tiny signal. Suppose life there is very sparse, and they only had
a few equivalents of bacteria. We struggled very hard to make our instruments
sensitive enough, so that we could pick up the faintest signs of chemical
activity. It ended up there was a lot of chemical activity, and we have
not adequately modeled what happens when you take a soil and expose it
incessantly to ultraviolet light for a long period of time, and then study
the chemical reactivity of that product, and it turned out under those
conditions, the UV generates ozone, the ozone oxidizes surface materials,
you get superoxides of some of the metals, and bango, they will react
chemically with some of the substrates that we were feeding it,
and we didn't pay enough attention to that side of it. We were so
struggling to, so focused on trying to detect the tiniest signal,
that we didn't think of all the possible artifacts that might have come along.
You know, we were given an impossible task, and that was to design in advance
the one and only experiment, and we had no chance of a sequential process,
in which you look at the preliminary results and say oh, that's interesting,
let's try this, let's try that and so on. It all had to be put into a
capsule and sent, it was a one shot kind of deal. So, to that extent
in some ways the experiments were less conclusive than they ought
to have been, although I think that we made a pretty firm case in
the end with the built-in chemical observation capability that would almost preclude
there being any living organisms at that latitude, and I haven't given up
all of Mars yet, but at least in those latitudes...And, fortunately,
there was other instrumentation, there was a mass spectrometer that
was also sent along, it found no evidence of any organic material in the
Mars surface, and if the oxidated activity that our experiments were
demonstrating were the result of bugs, it would've been a very strong signal
on the mass spectrometer, so I think we were able to come to a
reasonable conclusion, but, you know, you asked me for something that in
retrospect we could have done better, so I gave you an example.