The physical properties of the nasalized vowels have been investigated from many angles by use of the Sonagraph as the sound analyzer, associating some other measures to ascertain the effects of the characteristic features. The Principal characteristics of nasalization are found as: (1) enforcement of intensity in the region about 250c. p. s. , due to the dull resonance of the pharyngo-nasal cavity. (2) weakening of intensity at about 500c. p. s. , due to the selective attenuation caused by the resonance of the nasal cavity (antiresonance of the nasal cavity as a side channel of the vocal tract). (3) appearance of comparatively weak and diffuse components which are ascribed to the sound emitted from the nostrils. These components are interpreted as representing mostly the sound at pharynx affected by the dull and complex resonance in the nasal cavity. These three features give the perceptional impression of nasalization, and are commonly found usually in any nasalized vowel, though the relative importance of each feature depends on the vowel. If the nasal passage is closed with fingers at the nostrils when the nasalized vowels are pro nounced, the features (1) and (3) disappear, and the frequency of selective attenuation (2) is lowered. This causes an appreciable reduction of energy in the lower frequency region, resulting a distortion of the vowel color.
The characteristics of nasalization reported in the preceding paper by the present authors have been further studied for the case of nasalized vowels with blocked nasal channel, in relation with the charateristics of nasal consonants. The features (1) and (3) are commonly found also at the retention of nasal consonants. The feature (3) which, in this case, is affected by the antiresonance of the oral cavity as a converse effect of (2) in the case of nasalized vowels, carries information about the tongue position, giving distinction of the sound from other nasals, when only the retention is heard. The discontinuous transition from the sound of the retention of the nasal to the following vowel, as usually seen on the sonagram of a syllable e. g. /ma/, is ascribed to the essential difference of the mechanism of formation of the two sounds, and the impression of"plosion"in the glide from the nasal to the vowel is presumably due rather to this abrupt change of timbre, than to the plosion.
The driving-point mechanical impedance describing the steady-state characteristics of mechanical vibratots are considered. When a mechanical vibrator is driven by such a force which is in certain distribution and which is changing sinusoidally with time, the drivint point impedance is defined by the ratio of the total force to the mean velocity at the driving position. If the equation of motion, describing the steady state displacement of the vibrator driven by a concentrated force be solved, and the steady state solution under this condition be obtained, then the driving point impedance is formally calculated according to definition. But this gives sometimes an incorrect value, for the steady state solution with a point source (Green's function or Green's tensor of the problem) may have some singularities in the neibourhood of its source point. The behavior of the solutions in the neighborhood of the source points, and singulalities of the driving-point impedances are considered for various vibrators, namely, stretched strings, longitudinally or torsionally vibrating bars, stretched membranes, laterally vibrating bars, thin plates, stretched plates, etc. All of these solutions are continuous and driving-point impedances have no singularity except the only case of stretched membranes. In the case of the vibrator with an isotropic elastic material however, the driving-point impednce of such a vibrator with a point source terminal or a line or curve source terminal has serious singularities and its value becomes identically equal to zero, because of the singularities of the Green's tensor (diadic) of elastic waves at its source point.
The fundamental vibrations of the short-cylindrical barium titanate vibrators are investigated. Assumptions are: i) The disk polarized in the direction of thickness is isotropic concerning its elastic properties. ii) The electrostrictivities Γ_<11> and Γ_<12> are connected with the relation: Γ_<12>=σΓ_<11> where σ is Poisson's ratio. Under these assumptions, axially symmetric vibrations of free disks are deduced from the general equations of motion. Two cases, which can be well approximated either by a thin disk or a long bar are considered, because of its difficulty in satisfying all boundary conditions. The corresponding curves for the both cases are connected with each other through the point P_0 of (thickness)/(diameter)≌0. 854, where, fortunately, all boundary condition are satisfied. In measuring the electrostrictive constant εΓ_<11>(ε: susceptibility) or electrostrictivity Γ_<11>, it is more convenient to consider the shape factors χ_1 and χ_2 which are determined by vibrational modes and the configuration of electrodes. The factors χ_1 and χ_2 are computed numerically. The characters of electrostrictive vibrators in the first longitudinal mode are elucidated by this research.
Experimental results on the Tuning Rod with circular section, which is to be used as the practical frequency standard, are stated in this paper. Merits of this Tuning Rod are economy of material and simplicity of construction work. Many experiments have been made and its working characterestics were compared with that of the Tuning Bar with rectangular section.
Accuracy in free field calibration of microphones is dependent on signal to noise ratio of its enviromental circumstances, and for carrying out the calibration of microphones in the sea-water, it is necessaty that the magnitude of undersea noise be known beforehand. Measurement of the undersea noise spectrum and the amplitude in the frequency range of 5-60kc/sec at Mito-hama Bay (Numazu, Shizuoka Prefecture) was made by using ADP standard microphone. The spectrum and magnitude of surrounding noise were dependent on the time, the season, the weather, the place, etc. , and our data show that the intensity of noise was in the order of 10^<-16> watts/cm^2//cycles/sec.
It is shown that the elastic constants of metal plates can be measured by rotating plate technique in which the ultrasonic transmission is plotted as a function of the angle of incidence. As ultrasonic apparatus, a commercial "Roflectoscope" was used by which the defects of materials were tested with pulsed ultrasonics. The elastic constants of Al, Cu, Pb and Pb alloys have been measured with this equipment and the value obtained is shown to be in agreement with the value obtained by Marten's mirror method. It has been found that by this method the thickness of metal plates must be more than twice the wavelength of ultrasonics in the plates.