In his letter Donohue further defended his argument, presented in an article in the September 12, 1969, issue of Science (vol.
165, p. 1091), that Fourier syntheses of X-ray data produced by Maurice Wilkins and others fit a model with alternative base-pairing,
not just the canonical A=T and C=G pairs of the Watson-Crick model. In particular, Donohue once again insisted that Fourier
syntheses could not conclusively prove any model because this method for inferring the structure of molecules relied on assumptions
about the angles of bonds between atoms which were based upon the proposed model of the molecule itself. Donohue saw in this
a theoretical fallacy which, as such, could not be used to rule out other molecular structures.
The resolution achieved by contemporary X-ray diffusion techniques--a resolution to about 3 angstroms, as Donohue mentioned--was
yet too low to allow for a conclusive proof of Watson's and Crick's model (an angstrom is a measure for the distance
between atoms; one angstrom equals one ten-millionth of a millimeter). Such proof was not offered until the late 1970s.
A dyad is a form of symmetry by which one half of a structure matches the other half in reverse, meaning that if the structure
is rotated by 180 degrees, it comes back into symmetry with itself. In the case of DNA, evidence for a dyadic structure,
produced first by Rosalind Franklin, suggested to Crick that the molecule consisted of two (not, as other researchers had
proposed, three) chains, and that the two chains were antiparallel.
Number of Image Pages:
2 (138,927 Bytes)
1970-06-09 (June 9, 1970)
Original Repository: Wellcome Library for the History and Understanding of Medicine. Francis Harry Compton Crick Papers
Courtesy of the Wellcome Library for the History and Understanding of Medicine.
The National Library of Medicine's Profiles in Science
program has made every effort to secure proper permissions
for posting items on the web site. In this instance, however,
it has either not been possible to identify or contact the
current copyright owner.
If you have information regarding the copyright owner,
please contact us at
Medical Subject Headings (MeSH):
Embryology and the Organization of DNA in Higher Organisms, 1966-1976
Letter from Francis Crick to Jerry Donohue [ca. 15 July 1970]
Letter from Francis Crick to Jerry Donohue (May 20, 1970)
I have before me your letter of May 6. Although you may find it amusing to believe that I have not appreciated the point about
the dyads in the DNA model, I find it nothing less than scandalous that even when this point is explained to you, you still
maintain your original erroneous position. The dyads in the model do not lead to numerous effectively centric reflections,
and if you don't believe me ask Struther for the calculated A's and B's - you will find that there are not numerous
F's with B = 0. Hypersymmetry may affect the distribution of the magnitudes of the F's (the N(z) test) but it will
not lead to what you call numerous effectively centric reflections.
However, as I have already pointed out, even if the above were true, my original point is still valid, viz., that presentation
of a Fourier based on a model does not provide proof of that model, whether or not the structure is centric or acentric.
You now bring up an entirely new point and state that it is far harder to get false structures from polymers than from single
crystals of small molecules. I note that you merely assert it is far harder, but not impossible. This is fortunate, because
a structure proposed for a rather simple polymer in Physica 23, 746 (1957) and Acta Cryst. 20, 341 (1966) was later shown
to be incorrect in J. Chem. Phys, 51, 348 (1969) and Acta Cryst. B25, 2168 (1969). Thus, even if it were true that a polymer
imposes greater restrictions on a Patterson these restrictions are not severe enough to prevent arriving at a false solution.
However, is it not a fact that these restrictions do not apply to the low resolution DNA data? At 3A most crystals are polymers.
Of course I can explain Fig. 2 on p. 1701, as well as Fig. 1 which comes from another source. What I cannot explain is figures
such as occur in JMB 11, 391 (1965) and elsewhere. Perhaps it is significant that the latter were made using experimental
data, whereas 1 and 2 above were made with synthetic error-free F's.
But enough of these red herrings. The question is Does model building followed by electron density calculation furnish proof
of a structure?