616 REPonT— 1887. 



and determined by experiment. The equation may then be written 



which is the equation to the characteristic curve of a shunt-wound or separately 

 excited machine. Having determined the characteristic when i/ = o the characteristic 

 surface can be determined therefrom by considering the form of v^. (See J. and E. 

 Hopkinson, 'Trans. R. S. ' pt. i. 1886, p. 334.) 



12. On the Froduction of a Gon.stant Current ivith Varying Electromotive 

 Force from a Dynamo. By A. P. Tkotter, B.A. 



The well-known methods are by (1) rocking the brushes ; (2) compound wind- 

 ing ; (3) reducing the strength of the field. 



The first method cannot be adopted with a modern ring or drum armature in a 

 strong field, though it is used with some success in the Thomson-Houston and 

 Hochhausen systems. Compound winding can only produce a very rough approxi- 

 mation to a constant current ; and, lastly, the strength of the field cannot be 

 reduced far without working on the nearly straight part of the saturation-curve, 

 when the electromotive force becomes unstable. 



Mr. Ravenshaw, the' senior electrician of Messrs. Goolden and Trotter, proposed 

 to keep the field saturated, but to weaken its useful eflect by a movable yoke or 

 keeper, which, by oHering a low magnetic resistance, would divert the magnetism 

 from the armature without materially altering the saturation of the magnets. The 

 writer suggested that, instead of moving this keeper, its effect could be annulled by 

 winding on it a coil through which a comparatively feeble current might circulate. 



The general method which, with certain precautions, has been put to practical 

 use with complete success, is, therefore, as follows : 



To a dynamo with a single horseshoe field another magnetic circuit is applied, 

 such as a similar horseshoe, which uuder ordinary circumstances would offer a so 

 much smaller magnetic resistance than the ai'mature and its air-space, that nearly 

 all the lines of force would be diverted through it. 



This is the condition of minimum electromotive force. 



This second magnetic circuit is provided with coils like those of the main 

 magnet, and by the passage of a current through these coils the diversion of 

 the magnetism of tlie main magnets may be obstructed, until, with a certain 

 strength of current, no lines offeree will pass through the second magnetic circuit, 

 and the electromotive force of the armature will be produced solely by the whole 

 useful magnetism of the main magnet. As, however, the second magnetic circuit 

 is similar to the main magnet, it may be used in the same way, and by further 

 increase of the current through its coils may assist the main magnet, until tlie 

 efi'ect of the two is combined, thus doubling the output of the machine as 

 first described. 



This is the condition of maximum electromotive force. 



In a shunt machine the current through the coils of the second magnet may 1>e 

 controlled by the addition of a resistance in series with it. In a series machine 

 the current may be controlled by a resistance arranged as a shunt on the coils, or 

 by dividing the coils into sections. These resistances, whether in series with shunt 

 coils or as shunts on series coils, may be controlled by hand or by automatic 

 regulators. 



13. Fescrij)tion of an Induction Coil. By George Higgs. 



This induction coil was designed and constructed specially for the purposes of 

 spectrum analysis ; the dimensions of the various parts are given as follows :^ 



The core, which is 14 inches long by 1| inch in diameter, is composed of very 

 soft iron wire, No. 20 B. VV.G., but although selected with considerable care the 

 residual magnetism is very perceptible. 



The primarj^ wire is of No. 12 copper, double covered with cotton wound in 

 three layers, and about 40 yards in length, the whole accurately fitting inside an 



