Leeds University Library

Summary

Although much of the data in the notebook is hard to interpret (after all, the Braggs were not consciously recording their results for posterity), we have clear evidence, in these pencil-written pages, of the experimental records which were transcribed into their published papers and book and which provided the basis of their Nobel prize-winning work on the analysis of crystal structures. Their debt to Dr Hutchinson in lending crystals of sufficient size and perfection cannot be overestimated and p100 (figure 16) lists the crystals thus lent and (perhaps by 1914), those returned.

There is one curious entry (shown at the top of figure 16): "Cobaltine. Octahedron. Pentag.dodecahedron". Cobaltine (cobaltite, CoAsS) has a cubic crystal structure consistent with an octahedral habit. But a pentagonal dodecahedron is not consistent with a crystal structure since five-fold symmetry is forbidden. Was the crystal multiply-twinned, which can give rise to pseudo-pentagonal symmetry, or was it a quasi-crystalline form?

There remains one query which we believe has not been addressed before: why did the Braggs choose comparatively complicated crystals to study - the spinels and the carbonates for example - and not crystals of the metallic elements which have simple crystal structures and the advantage of single, strongly scattering, atomic species? Surely sufficiently large crystals of iron, chromium, silicon, zinc, etc. were readily available? It was not until the work of Albert Wallace Hull in the USA from 1917 onwards and Arne Westgren in Sweden from 1921 onwards that the structures of a large number of metals were elucidated. They used powder diffraction techniques, pioneered by Debye and Scherrer, in which case of course large single crystals were not needed.

Did perhaps the Braggs feel, in 1913, that it was only in well-formed crystals that the atomic arrangements of the atoms were also clear and well-formed - unlike metals which rarely provide well-formed crystals and whose characteristic property of malleability indicated a disintegration of any possible pre-existing atomic structure? After all it was another twenty years or so before the concept of dislocations explained both the malleability and the invariance of the crystal structures of metals.