the coils themselves. Now, when the coils DD happen to be at the same moment in that position during their revolution in which they are producing the maximum and minimum amount of current respectively, as must often be the case where there is no synchronism, that current which is at the maximum rushes through the coil which is producing the minimum current, as is shown by the spark at the point where contact is broken between G and H. The effect of this passage of the current from one coil to the other is to accelerate or retard the rotation of the armature (according to the direction of the current) until synchronism is established.

That this influence of one coil upon the other operates in the manner described was easily shown by the following experiment:-The driving-strap of one of the armatures was removed, so that only one of the armatures should be producing a current, while the magnetism of the electromagnets of both machines was, as usual, maintained to the same degree. On placing the stationary armature with its coil in a suitable position in relation to the magnet-cylinder for producing electromagnetic rotation, and setting the other armature in motion, the stationary armature with its coil oscillated rapidly in arcs of very small amplitude, the oscillations corresponding in number with the alternations of the current. As the amplitude of the oscillations in this experiment was limited by the vis inertia of the armature, and in order that the effect of one pulsation only on the armature might be observed, contact was made and broken suddenly between the plate H and the end G of the coil by a sort of tapping motion, when the stationary armature was suddenly jerked round nearly a quarter of a revolution, sometimes in the direction in which it would have been driven by the strap, and at other times in the opposite direction, according as the alternating electrical wave which happened to be passing at the instant of making contact was positive or negative.

We have now seen, in the results obtained with the rotating and stationary armatures, a cause sufficient to account for their synchronism when revolving together, the absence of synchronism observed when the terminals F and H were bridged over by a conductor having comparatively little or no resistance being occasioned by the controlling current traversing the short circuit established between the terminals F and H, instead of the 280 feet of resistance presented by either of the coils when approaching the neutral point of their revolution. The absence of synchronism observed when the direct current was taken from the machines by means of commutators, is caused by the direction of the current being coincident with that which they would receive by induction from the electromagnets, and consequently opposite

to that which tends to impart an accelerating or retarding impulse to the armatures.

Having obtained the full effect of the combined alternating currents from the two machines without any mechanical gearing, it yet remained to obtain the combined direct currents from the machines in the same manner. A pair of rings and a commutator were therefore fitted upon one of the armature-spindles, which was made sufficiently long for the purpose, and metallic connexion was established between the rings of each machine and the commutator on the prolongation of the armature-axis. As the commutator necessarily revolved synchronously with the two armatures, it was found that the combined alternating currents were rectified just as if they had proceeded from only one machine, and were consequently available for electrodeposition, or for any other purpose for which a direct current might be required.

Although this property of synchronous rotation has as yet been observed only in the case of several pairs and a triple combination of machines, yet there is no reason for supposing that it may not be extended to any number of machines that may be conveniently worked together from the same prime mover. It is necessary, however, to observe that as the controlling power of the current is only calculated to correct such minute deviations from synchronism as it is beyond the power of mechanical skill to prevent, the driving and driven pulleys should be respectively as nearly as possible of the same diameters, as the correction of any considerable difference in the number of the revolutions of the armatures, caused by differences in the diameters of the pulleys, must necessarily be attended by a corresponding diminution of the useful effect of the current outside the machines.

Before concluding this communication I wish to direct attention to an important property of the magneto-electric circuit which renders the commonly accepted theory, by which the generation and propagation of the electric influence in voltaic circuits is explained, inapplicable to those circuits which are entirely metallic. Reference to this property is all the more called for at the present time, as I find that a want of acquaintance with it has given rise to no small amount of misconception on the part of several eminent mathematicians and electricians who have examined my experiments on the electric condition of the earth, and the method by which I have thought proper to estimate the magnitude of powerful induction-currents*.

The intensity of a voltaic current, as represented by the mathematical theory of Ohm, is equal to the electromotive force di

*Philosophical Magazine, August 1868.

vided by the internal resistance of the battery; and from this theory it is inferred that an electromotor, in order to overcome a great external resistance, must itself possess a correspondingly great internal resistance. A further consequence deduced from this theory is, that the maximum useful effect of a given electromotor is obtained when the external and internal resistances are equal.

Now this mode of estimating the magnitude of an electric current does not apply to the circuits on the armatures of my machines. Taking for example the results obtained from the quantity-armature of a 10-inch machine :-The dimensions of the coil of this armature may be represented by a bar of pure copper, 67 feet long, and having a sectional area of 1.6 square inch; so that the resistance which this circuit presents to the passage of a current, when compared with that of the liquids in a voltaic battery, is practically null. When the coil is in full action it will melt 15 inches of thin iron wire 035 of an inch in diameter, or the same length of 4-inch iron rod with equal certainty, and will electrolyze acidulated water in at least 16 voltameters in series; so that the resistance outside the circuit, whether estimated by the 15 inches of thin wire melted or by the number of electrolyzing-cells in series, is more than a hundred times as great as that of the coil in which the current is generated.

Moreover I have found that whenever a voltaic battery and a magneto-electric machine will melt an equal length of wire, the power which these electromotors have to overcome external resistance, as measured by the number of voltameters in series, is also equal. And, generally, the power of an electromotor (whether voltaic or magneto-electric) to overcome external resistance is directly proportionate to the length of wire which it will melt.

From a consideration of these results, it will be seen that one of the fundamental elements which enters into the theory of Ohm is found wanting when that theory is applied to the estimation of the magnitude of currents generated in circuits entirely metallic.

MM. Jamin and Roger, in a recent Number of the Comptes Rendus of the Academy of Sciences*, have also pointed out the discrepancy here referred to in the application of Ohm's theory to magneto-electric circuits. I am, however, by no means prepared to admit the correctness of the views advanced by these physicists in their endeavours to reconcile the facts observed with established theory; besides which, other anomalies present themselves when the customary formulæ are applied to magnetoelectric circuits, a consideration of which must ultimately lead

* Philosophical Magazine, October 1868.

to the enunciation of laws much more general in their application than those with which we are at present familiar.

Manchester.

P.S. Since this paper was read, it has occurred to me that a comparison might be attempted to be drawn between the controlling power of the magneto-electric current over the rotations of a number of armatures, and that of the voltaic current over the oscillations of a number of pendulums. Beyond the fact that synchronism is produced in both cases through the agency of an electric current, there is no further resemblance between the two actions. In the case of the armatures the synchronism is produced by the mutual action of several rotating bodies upon one another, or by the dominant influence of several bodies upon one; whereas in the case of the pendulums the synchronism of the system is produced by the influence of one body alone upon several. Again, the synchronism of a number of pendulums is only accomplished by the skilful adaptation of means to an end, while the synchronous rotation of a number of armatures is a phenomenon which exhibits itself without the exercise of any ingenuity whatever; and, so far as I have studied this peculiar electromechanical action, no amount of ingenuity can produce the synchronous rotation of the armatures by means of the voltaic current, as magneto-electric currents and circuits seem absolutely essential to the attainment of this result.

IX. Notices respecting New Books.

A Manual of Elementary Chemistry, Theoretical and Practical. By GEORGE FOWNES, F.R.S., late Professor of Practical Chemistry in University College, London. Tenth Edition. London, Churchill: 1868. (Pp. xxviii & 1020.)

The Elements of Heat and of Non-metallic Chemistry. Especially designed for Candidates for the Matriculation Pass Examination of the University of London. By FREDERICK GUTHRIE, B.A. (Lond.), Ph.D., F.R.S.E., F.C.S., late Professor of Chemistry and Physics, Royal College, Mauritius. London, Van Voorst: 1868. (Pp. x & 210.)

THE

HE different chemical manuals which appear from time to time seem to be written from two distinct points of view-the author desiring either to display some original mode of considering his subject, or to make an average statement of the chemical knowledge which is accepted, in a given year, as useful to the student. Manuals of the former class are, from their nature, not very frequently written; while those belonging to the latter constitute the great majority of such publications. To each kind a special merit apper

tains; but while the one has the narrower limit of an individual effort, it is included in the other, which hands down to the historian the general character of a time.

The present edition of Fownes's Manual of Chemistry' has the curious property of blending both these distinctions. It does this by adhering to the design of the late Professor Fownes, which, if interpreted from the first edition and preface, was to lead to the principles of chemistry by an inductive ascent, and convey as complete an impression of the entire range of the science as could fairly be expected. Hence we find a special place allotted to the analytical characteristics of the elementary bodies, brief notices of laboratory operations, of higher researches, of the relations of chemistry to physics and biology. Such a plan was not then a novelty abroad; but no students' book of this kind had, so far as we are aware, appeared in this country. The sale of nine editions in twenty-four years, and the almost universal approval of English teachers, are gratifying proofs both of the value of the original conception and of the manner in which it has been carried out at subsequent intervals. Though succeeding editions have thus been invariably prepared upon the primitive model, the progress of science has made considerable readjustments and amendments necessary to them; and the influence of all the great ideas which have arisen in chemistry since 1844 may be readily observed here at the appropriate epoch. But, unfortunately, in this process the modest octavo of less than six hundred pages has so far overgrown as to contain at present more than a thousand. We cannot help thinking that some part of this growth is excessive, and that means might be taken to repress it with advantage. The physical introduction, for example, is no longer justifiable, when physics is beginning to be taught (even in elementary schools) by a distinct official, or as a distinct subject from chemistry. If that were omitted, about one-eighth of the entire volume (or one hundred and twenty-five pages) would be removed, referring to subjects which, as their very able writer occasionally admits, cannot be satisfactorily treated in so small a compass. For the electricity and crystallography afterwards described, and occupying about thirty pages, the student might also be referred to other and appropriate quarters. Most of the Tables, too, at the end of the book are usually sought for in larger works, and seldom noticed in their present position. An additional reason for this curtailment is to be found in the increased length of the sections which are devoted to chemistry proper. On account of the growing attention which is now paid to inorganic chemistry, there is much fresh matter to summarize and record in that department. Still, the number of those who pursue organic research preponderates; and it is here consequently that we notice the greatest enlargement in the size of the manual.

The best mode of treating the multitudinous detail which the science continues to produce is a point upon which chemists either doubt or disagree. The philosopher laments a dreary desert of facts fruitless and even dangerous for want of law; the teacher bewails each serious trifle that he is compelled to read, as adding only to the