The transition of amorphous alumina to aalumina was studied by X-ray diffraction, electron diffraction, DTA, TGA, and microscopic observation. The amorphous alumina was prepared by condensing vapor from evaporating molten alumina in vacuo onto the glass envelope of the vacuum chamber. The amorphous alumina was transformed to a poorly crystalline material by heating for 16 hr between 570° and 670OC. Between 670° and 1200°C, the poorly crystalline alumina was converted to a-alumina via two parallel series of transition aluminas. The principal series was ?-alumina to 6-alumina to a-alumina. A minor amount of &alumina developed from the initial crystallization and persisted throughout the duration of the principal series as a parallel path. Some conversion of 6-to 6alumina was detected above 900°C. DTA produced an unexplained exothermic peak at 320°C and a second exothermic peak at 860°C which corresponded to formation of metastable aluminas. 574Library, Microreproduction Service Room 14-0551.

No abstract

The va por press ure of t un gsten was measured by the L a ngmuir met hod in t he te mper ature rano-e 2574 to 3183 OK us ing a vaec um mi crobalance. The least-squares Illl e through t he dat'~ 10gP(at m) = 7.790 -44680/ T, y ields a second l aw tJl~ (2850 OK) = 204.~ ± 2.7 kcal/ mol a nd t.S~( 2850 ° K) = 35.64 ± 0.97 cal/ mol deg. Our re commended equ atlOn. for t he va por press ure is 10gP(atm) = 7.933 -45087/ T based on our expen rn entally de te~' mln e d m ean th ird l aw t. H~ (298 OK ) = 203.5 kcal/ 11101 and tab ul ated valu es for the entropies and e nt halp ies.

Six metal oxides (A120r, TiO,, NbzO5, Ta205, La203, and Dy20a) were melted in a solar furnace and the apparent temperatures at their melting points were measured by means of a conventional optical pyrometer. The apparent temperatures of three of them (A120s, Ti02, and NbzOa) were 2025O, 1840°, and 1496°C. respectively, in close agreement with the melting points as determined by other investigators using other methods. The apparent temperatures of the other three varied widely and bore no relation to their melting points. The method therefore does not have general utility for the measurement of unknown melting points.

The sodium content of portland cement is routinely determined by flame photometry. Some flame photometers were found consistently to indicate sodium contents of less than zero for certain cements.In addition, different brands of flame photometers almost never gave identical results for any cement.A study of the method showed that flame photometers having wide effective band widths gave low, often negative results, because of the inhibiting effect of silica on the emission of calcium oxide in the region of the sodium line. Increased accuracy was obtained by using an auxiliary multilayer interference filter to narrow the effective band width of the instrument or by removing the silica from the cement prior to the determination of the sodium.Both techniques together converted the current routine procedure into one which gave results substantially identical with those obtained by the gravimetric J. Lawrence Smith method, when using the Barclay, the Beckman DU, or the Perkin-Elmer flame photometer.