演者が一年有餘の米國滞留中、見聞調査した事柄のうち、最近歸朝した諸氏の視察談と重複せぬ方面のことを述べられたものが即ち本篇である。大體下の諸點につき有益な面白い事柄が平易に述べてあろ。 第一、微粉炭の使用微粉炭使用の利益、操作の順序方法、自然發火の恐れ、實驗の結果。 第二、水力發電所の連結運轉米國では需要家のランプを消す塲合の殆んど絶無なこと、Δ對Y接續、リレーの重要視されて居る事、Southern Power Co.の概況、Central Main Power Co.では大小21個の發電所を連絡運轉して居る事。 第三、自働發電所Turners Power Co.及Iowa Railway and Light Co.の自働發電所の詳細な説明。 第四、懸垂碍子Cap and Pin type, Hewlett type及JD. typeの比較評論、懸垂碍子の不良化率、機械的強度を大にする必要、Tunnel Kiln。 第五、避雷器aluminum避雷器の手入法、oxide film型の一缺點。 第六、弧光抑制器G. E.製Creighton氏新案に成る彈丸順次發射型抑制器の良好な事。
Various methods are suggested to predetermine a load saturation curve at power facter zero, however it seems almost impossible to predetcrmine an armature reaction at full voltage from the short circuit and open circuit tests. Torda-Hcymann's method is widely adopted for the predetermination of a load saturation curve at power-f_??_ctor zero, in which the armature reaction at full voltage and power-factor zero is supposed to be is f(R), where is is the field current to over come a armature reaction on short circuit. Dr. Torda-Heymann has suggested that the value of f(R) can be represented by (R/Rs)2, where R is the rcluctance corresponding to the actual saturation and Rs is the lowest limit of reluctance with no saturation. The auther has tested few actual machines and found that f(R) can not be represented in a form (R/Rs)2, or even in the general form (R/Rs)n, that the value of r is not constant (varies from 1.2 to 2.3) even for the same machine and depends upon a degree of saturation. But we can find very intersting character of the saturation curve by a close observation of our experimental results that the actual value of f(R) is almost independent of armature reaction and only depend upon the actual point of saturation at which the resultant flux is working. From this result, we can determine the full load saturation curve at power factor zero by measuring a saturation curve at lower load current at power factor zero, or obtaining armature reaction at lower load current, and increasing this armature reaction by the ratio of the load current to the supposed full load current.
As a general case of an alternating current problem, an alternating electromagnetic current field is treated in which the rectangular components of each vector quantity, such as the density of electric and magnetic current, are considered as complex hormonic functions of time. Special attention is given to the problem of complex average power or vector power. As a method of analysis of such a problem, the symbolism of electrical vector power product, complex scalar product, complex vector product, etc, are proposed. The analysis is reduced to the calculation of complex effective vector fields and complex effecive scalar fields, that is, symbolic vector and scalar fields. This is an extension of the symbolic method to a space problem. The equations expressing the law of continuity of true electric current and true magnetic current and also Maxwell's fundamental equations of electromagnetic field are given in symbolic form. The phenomena of dielectric and magnetic hysteresis in revolving ellipsoidal electric and magnetic field is discussed under the conception of fundamental ellipse of the hysteresis loop, and the expressions of hysteresis loss and average stored energy in such fields are given in symbolic form. The symbolic vector fields of electric intensity and magnetic force are decomposed into curl field, divergence field and impressed field. This result is applied to the decomposition of vector power density and thus the fundamental equation of vector power in a complex harmonic electromagnetic current field is obtained in its symbolic differential form. Applying Gauss' theorem in the symbolic vector field of the vector power equation, it is converted into the following integral form. This equation, by the author's opinion, can be best explained as a principle of continuity of vector power in an alternating electromagnetic current field. As a simple case of technical application of this principle, the vector power equations for network of linear electric circuit, linear magnetic circuit and linear electromagnetic circuit are deduced. (June, 1921)