In Part I the theoretical and experimental conditions which must be satisfied in order to determine lattice constants with a precision of a few parts per hundred thousand are discussed with particular application to the symmetrical focusing type of camera. Cohen's method of calculation of lattice constants for the elimination of ``drift'' and experimental errors is applied. Methods are developed for the evaluation of standard errors and fiduciary limits of results from a single film and from a set of films. The usefulness of x-ray targets made of alloys rather than pure elements for the purpose of securing a larger number and better distribution of lines is indicated. The influence of the number and Miller indices of diffraction lines on the values of lattice constants in noncubic systems is shown. The importance of the methods of sample preparation for precision work is emphasized. In Part II, precision measurements on Al, Ni, Ag, Au, Si, Fe, Mo, W, Mg, Zn, Cd, Sb, Bi and Sn are reported for materials of a high degree of purity. The fiduciary limits of these lattice constants vary between 2 and 7 parts per hundred thousand and are so chosen that the probability of the correct value lying between the stated limits is 19 out of 20.

A series of iron oxides with compositions between 76.08 and 76.72 percent Fe, all of them lying within the single phase solid solution area known as Wüstite has been prepared and the lattice constants and densities of the individual members determined. The NaCl structure found by earlier investigators and considered by them to be the structure of FeO has been found throughout this series. Contrary to the earlier results the size of the unit cell decreases as the iron content decreases. The solid solution is of a complex type; an oxygen ion replaces an FeO group with an accompanying increase in valence of 2 Fe++ to 2 Fe+++. The results are discussed from the chemical and structural points of view.

Carbon produced by the explosion of graphitic acid showed a density of 2.215, and is, therefore, probably graphite.4. Samples of the graphites originally showed no "drift" or increase in weight with time after immersion in the liquid. The natural graphites, however, after being swelled, or increased in volume by treatment with fuming nitric acid and heat, displayed remarkable drifts. Pressure was used to hasten completion of the drift.5. The tendency to drift was largely destroyed by a severe compression of the swelled material previous to the density determination.

VIBRATIONAL INTENSITIES. BrCI AND ICI 365The BrCI was prepared from reagent grade Br2 (Baker) and 99-percent pure tank Cb (Matheson). The Br2 was dried by passing it over P 2 06 three times in a vacuum line, and it was used without further purification. The Ch, similarly dried, was distilled once from a bulb cooled with a dry-ice butanol mixture; the middle third of the distillate was used. The vapors from the Cl2 and Br2 were admitted to the cell at known pressures and allowed to mix by diffusion. The concentration of BrCl can be found from the dissociation constant of BrC}! and the amounts of Br2 and Cb added. The ICI was a commercial product (Eastman) described by THE JOURNAL OF CHEMICAL PHYSICS the manufacturer as "pure." It was dried over P 2 0 5 and used without further purification.Both fundamental bands are in a region of strong water bands, and despite a continuous current of dry air through the enclosed source and monochromator, the traces still showed considerable water absorption. This and the relatively weak signal at long wavelengths are the principal reasons for the large indeterminacy of the results. ACKNOWLEDGMENT

on an iron cylinder, recorded on a recording milliammeter. The compensation apparatus works automatically by means of a system of relays and a drum resistance rotated by a motor. From the relation between time and current the sedimentation curve can be found and from that curve the distribution curve can be calculated.4. To illustrate the procedure, the sedimentation and distribution curves of a mercury hydrosol and a gold hydrosol have been given.5. A method has been described for determining the distribution of size of particles, depending on the variation of concentration with height in a sedimenting system. As an example, a gold hydrosol was studied, the concentration at different heights being measured by means of the light' absorption.6. The theory for the development of an analogous method depending on the use of centrifugal force has been discussed.