258 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1951 



material now known to be molybdenum oxide. Both of these intrud- 

 ing constituents form in the boundaries of the individual grains — 

 the carbide increasing cohesion, and the oxide destroying it. Here 

 one is accordingly looking, as a special case, at the external surface 

 of an individual internal grain — not its internal surface as in previous 

 figures — and under the remarkable circumstance of finding both of 

 tlie counteractive phases present. The carbide and oxide react, of 

 course, to form carbon oxide gas, which is removed by the vacuum 

 treatment. Here one can actually visualize the oxide caught in the 

 act of invading a region of carbide feathers, destroying them as it 

 advances. 



THE MICELLAR THEORY 



Wliile such discoveries as the preceding readily lead to practical 

 applications, a matter of far greater scope and interest is highlighted 

 by f ractographic patterns. 



A century and a half ago, a great French mineralogist and crystal- 

 lographer, Haiiy, established what is now known as the Law of Ra- 

 tional Indices in crystallography and laid down a description for the 

 physical constitution of crystals which endured for many years. Haiiy 

 spoke of the "molecules integrantes," which were presumed to be 

 minute building blocks — perfect microscopic crystals — which fitted 

 together to comprise the macroscopic crystal. Virtually every scien- 

 tist of that period accepted the theory that crystals were built of tiny 

 crystallite units. The impact of atomic theory and space-lattice theory 

 in the latter nineteenth century, and particularly X-ray diffraction 

 in 1912, temporarily shattered this picture to replace it with a con- 

 ventional concept of regular atomic structure extending from the atom 

 individual up to the boundary of the crystal or grain. 



Nevertheless, in the past several decades, this picture of the homoge- 

 neous atomic lattice has come under sharp criticism from many angles 

 of research in which crystalline substances persist in showing a mark- 

 edly subdivided structure on a scale far more minute than the indi- 

 vidual grain, yet much greater than the atom. Many theories have 

 been advanced to explain this anomaly, and these can be reviewed in 

 most current textbooks on physics. It is now becoming widely agreed 

 that most crystals, if not all, have a finely subdivided structure. The 

 nature and the origin of that structure, however, constitute one of 

 the most hotly argued problems in metallurgy and physics today. 



Briefly, the principal contention rests upon the question whether 

 the subdivision results from imperfections and accidental submicro- 

 scopic cracking, or whether it is a fundamental result of the surfaces 

 of previous submicroscopic units that come together at the time of 

 freezing to form the solid. 



