This method is intended for the measurement of colour temperature of tungsten lamps. Using a comparison lamp, its luminous flux and that of a test lamp are alternately projected on the window of an integrator by a rotating sector, and two multiplier phototubes, which are respectively combined with a red and a blue filter, are fixed on the other sides of the integrator. Therefore, the a. c. output of each multiplier phototube corresponds to the amount of unbalance between the two lamps. To detect the condition of the balance, the a. c. output is amplified, and then fed through an electronic switch and into the vertical axis of a cathoderay tube oscillograph, the horizontal axis of which is applied by the time sweep. As the result, the amplified output of each multiplier phototube is represented on the cathoderay tube, and two straight lines are drawn when the balance is made by adjusting the distance of the comparison lamp from the window and its colour temperature. Consequently, the colour temperature of the test lamp is directly obtained from the voltage of the comparison lamp in the balance point, if the voltage of the comparison lamp is calibrated beforehand for the colour temperature on the test lamp side, by substituting the standard lamps of the colour temperature for the test lamps. In spite of only a few occational calibrations, the colour temperature of tungsten lamps could be measured with the reproducibility of about ±2°K in as short a time as 1_??_5 minutes, even if the characteristics of light sensitive devices and colour filter were not ideal.
When the illuminator and receptor axes of goniophotometer are fixed orthogonally and a glossy surface of specimen is so rotated that its normal approaches the latter axis from the initial position of specular reflection, the reflected flux I from the specimen decreases rapidly at first and then slowly as the angle of rotation θ increases. Using a photometer capable of measuring the relative brightness, the slowly decreasing part of I-θ curve is found to be expressed approximately in the form of I2 exp(-βθ2) for many papers. The gloss number GNF1) is defined as the ratio of I0/I2 where I0 is the value of I at the initial position θ=0. The values of GNF for a series of 14 papers, white and coloured, are compared with the visual gradings of Japanese observers. And it is found that, the logarithms of GNF are linearly correlated with the visual gradings of 13 observers except for three white papers, but the correlation holds good for all the papers when the gradings of those two observers who have participated in similar tests several times in the past are taken as the values of visual gloss.
The effect of water vapour on oxide cathode is experimented. Water vapour makes an irreversible poisoning action on the activity of oxide cathode and the relation between log i/i0 and log P is found linear, where i0 is the original activity of the cathode, i that after the water vapour treatment and succeeding reactivation and P the water vapour pressure. This relation is discussed by.applying the theory of sintering in the case of surface diffusion.
The relation between the evaporation rate and the activity of oxide cathode is studied by experiment. The evaporation rate has no relation to the activity, but a high temperature treatment is found to affect the latter as well as the activation energy. Therefore, the evaporated material is in all probability the oxide molecules and, in one case, the activation enerev of evaporation changed from 3.2 eV to 3.9 eV by the heat treatment.
The internal friction Q-1 of three kinds of tungsten filaments is measured at tempera-tures between a room temperature and 2900°K, using a simple apparatus newly designed. It increases with temperature, but there are no noticeable differences among its values below 1000K°. Above this temperature, however, it becomes quite sensitive to the crystal structure. Thee value of Q-1 of a thoriated filament with polycrystalline structure reaches 4×10-2 at 2000K°, whereas that of a nonsag wire with well developed long grains is 0.2×10-2 at the same temperature and remains at as small a value as 0.4×10-2 even at 2900°K. The value of a filament composed of both long and small grains is in between. A nonsag wire before the grain growth gives high Q-1, which decreases rapidly when exaggerated grain growth takes place in the filament by suitable heat treatment. The internal friction of tungsten filaments at high temperatures mostly originates in grain boundaries and is sensi-tive to the area of boundaries or the size of grains. This phenomenon can be used as a non-destructive method either to estimate the size of grains or to follow the grain growth in the filaments.