Imágenes de páginas
PDF
EPUB

normal magnetization correspond in this case to the abnormal periods.

Hankel applied himself to the explanation of this phenomenon, and he lays down the following as the fundamental idea:

"It is known from Faraday's researches, that a current at its commencement generates an opposite current in a neighboring conductor; at its cessation, on the other hand, a second current which passes in the same direction with the original one. The electrical sparks must act in both ways, upon a steel needle placed near the wires, as the needle is perpendicular to the direction of the current, the planes of the currents produced in the needle are likewise perpendicular to the length of the needle, and the magnetism of the needle will be in opposite directions according as we consider it to be excited by the action of the beginning or by the cessation of the spark. But the two instants of beginning and end of electrical sparks follow each other so rapidly, that their separate effects cannot be measured; hence magnetization is the result of both of these influences."

This is essentially the fundamental idea to which Wrede (Berzelius' Jahresbericht, deutsch von Wöhler, 20ster Jahrgang, S. 119,) sought to reduce the alternate normal and abnormal magnetism of steel needles by the discharge stroke in main as well as in secondary wires.

As already intimated by Riess, (Dove's Repertorium, VI, 218,) this mode of explanation belongs yet to the domain of conjecture. It is possible that this is the natural process in magnetizing steel needles by the discharge stroke, but it is by no means proved.

On the whole this explanation seems very plausible; but the deduction of the particulars of the phenomenon is not at all convincing, although Hankel expresses himself quite at length upon the subject. We will do well to consider this as still an open question.

Riess remarks, in the place above cited in Dove's Repertorium, that it is better, and more for the furtherance of science, openly to confess the deficiencies of our knowledge, than to attempt to aid it with half explanation and to cover up its defects; and in this connexion he quotes a passage from Franklin's letters, which should be taken to heart by every scientific man:

"I find a frank acknowledgment of one's ignorance is not only the easiest way to get rid of a difficulty, but the likeliest way to obtain information; I think it an honest policy."

In the second memoir Hankel treats of the following points:

1. The number and magnitude of the magnetizing periods, mentioned in the first memoir.

2. The action of different spirals.

3. The action of the conducting wire upon itself.

4. The influence of the thickness of the needles.

5. The influence of the surface of the battery.

6. The changes of the alternations by obstacles interposed.

7. Special influence of particular metals, totally distinct from their conducting capacity.

We will consider these points in succession:

1. As a magnetizing spiral, a spiral of silver wire was employed with coils so close that the introduced needle covered 31 of them.

The charge of the battery was regularly increased by 1 spark of the measuring jar, and at each discharge a new needle was magnetized; the strength of the magnetism communicated was then determined by the time which the needle required to make a given number of vibrations. A copper wire 2.63 metres long and 1.2966 millimetre diameter was used in the circuit together with the spiral.

In this manner Hankel made a series of experiments whose results are represented graphically in fig. 69. The abscissas are proportional

[blocks in formation]
[graphic]

to the strengths of the battery charges, the ordinates to the strengths of the corresponding magnetization. The ordinates above the horizontal 0 correspond to normal, those below to abnormal magnetism.

This curve does not produce the impression of regularity; it seems rather to mask some sort of a law by irregularities which cannot be corrected by computation. But in such cases the law may be represented by averages obtained from numerous experiments.

Hankel says he repeated these experiments with the shortest circuits, to determine the position of the abnormal, or equally significant weak normal periods; from all his experiments with the same kind of needle, using the same battery of nine jars, he found these periods to occur in the following charges: 3, 6, 9, 11, 14, 16, 18, 21, 23, 26, 29, 32, 36, 40.

Hankel says, "we see that the change in the polarity returns regularly;" but I can find in this series of numbers nothing very clearly expressed, and least of all regularity. He says, moreover, that this regularity might have been more clearly represented by the introduction of fractions, but he purposely avoided them, as he had not measured them exactly, but only estimated them.

Now, what does this mean? Does not the above series of numbers represent the means of numerous experiments made under the same condition? If this is the case, why hesitate to introduce fractions? Mean values are generally computed, not observed.

To render it possible for the reader to judge of the value of his results, Hankel should have told how he arrived at the series 3, 6, 9, 11, &c.; and he should have communicated the separate series of experiments in order that one might ascertain how far the separate series differed from the mean on account of accidental disturbances.

2. The series of experiments represented by fig. 69, were compared with two others in which the spirals were so moved in the direction of their length that the needle covered only 28 coils in the second, and only 11 in the third series. The general result was, that the periods were longer in proportion as the needles covered fewer coils.

3. As mentioned above, Riess announced the proposition that, in discharging a battery, no part of the circuit acts inductively upon itself. Hankel contests this proposition. He comes to the opposite conclusion from the following experiments:

A copper spiral of tolerably large diameter was surrounded by a similar spiral, the two being so arranged that the discharge could at pleasure be made to pass through the two, either in the same or in opposite directions.* A magnetizing spiral was also introduced into the circuit. The march of the magnetizing periods for both arrangements being then compared they did not harmonize, and hence Hankel inferred that there was necessarily an interference of effects.

Even if it be conceded that Reiss' experiments are not sufficient to establish his proposition, those of Hankel are still less fitted to overthrow it; for, in the phenomena of magnetism by the discharge stroke, our knowledge of what is regular or what may be accidental is not

[merged small][graphic]

such as to permit a safe conclusion to be drawn from the want of coincidence of two such series of experiments.

The differences which occur in magnetizing steel needles, according

*Hankel gives the thickness of the wire to the Tooooo of a millimetre, which appears to me an unnecessary accuracy, considering the other relations of this series of experiments.

as a long wire introduced into the circuit is extended in a straight line or wound into a spiral, will be considered below under No 6.

4. It appears in general, as Hankel infers from his experiments, that with coarse needles the phenomena do not change; the anomalous periods occur only with stronger charges, and also appear to have lost in strength.

5. New experiments on the influence of the surface of the battery, corresponding to the previous ones, indicated that a diminution of the surface brought about the anomalous periods with decreasing charges, but so shortened them that, with a certain size of the battery, they ceased to appear as abnormal magnetization; weak and strong normal periods only were then observed.

6. Besides the short insertion, with which the results in fig. 69 were obtained, Hankel made experiments with inserted copper wires extended in a straight line 0.23 millimetre diameter, and varying between 0.375 and 96.4 metres in length. The curves 1 and 2, fig. 70, represent the results which he obtained with the wires 12 and then 96.4 metres long. These curves seem to indicate that with longer insertions the separate small periods disappear, until at last only a large normal period is observed with stronger magnetism, after which follows a very broad negative period, (from 30 to 100,) in which, however, very weak magnetism is observed.

With reference to the disappearance of the smaller periods, these experiments do not admit, in my opinion, of any certain conclusion, because the charge of the battery was increased from 5 to 5 for the longer insertions, and from 2 to 2 for the medium, while they increased only by 1 in the shortest. Where is the guarantee that in the longer wires single periods are not passed over? Hankel preserves silence on this point.

In relation to the influence of the coils, Hankel compares the result represented by the second curve of fig. 70 with those which are given by 103 metres of the same wire wound into 70 coils. While, with straight wires, a normal period extends to 30, and is then followed by a long negative weak one, he observed, with coiled wires, 3 normal and 3 abnormal periods.

When 26 metres of a very thick (30.76 square millimetres in section) quadrangular copper wire were inserted, no change was seen in the succession of the periods, but they were generally feebler. When, in addition, 113 metres of a round (1.3 millimetre) wire were inserted, stretched in a straight line, the results represented in the third curve of fig. 70 were obtained. Nearly all reversions disappeared, the needles seemed but feebly magnetic.

When 94 metres of the thick wire were coiled into a spiral and inserted in the circuit, the results presented in the fourth curve of fig. 70 were obtained. The enfeebling of the magnetism appeared here in the thick coiled wires still more strikingly than in that extended at length.

The influence of the coiling upon the thick and the thin copper wires is evidently very different; yet, says Hankel, (page 336 of his 2d Memoir,) the influence is the same in both cases. The discussion, by means of which he seeks to prove this, is incomprehensible to me; indeed, I cannot call Hankel's reasoning in general clear and precise.

7. The insertion of iron wires yields remarkable phenomena, producing anomalous periods of very considerable strength. Hankel found them particularly striking with thick, long iron wires. While a thick copper wire greatly weakens the magnetism, the latter is considerably strengthened by a thick iron wire. On introducing an iron wire 1.27 millimetre diameter and 131 metres long it gave, for instance, the result for a charge 6, a normal maximum 11; for a charge 36, an anomalous magnetism of the strength 9, taking for unity the magnetizing strength adopted in constructing the above curves.

§ 71. Leyden jars of thick glass.-Winter, of Vienna, constructs Leyden jars which have a much greater striking distance than those in general use, and he accomplishes this by using vessels with very thick sides, (over 1 line,) and by leaving a very wide uncoated border. Spontaneous discharge is prevented by the width of the uncoated border, and perforation of the glass is prevented by its thickness. In such jars the tension of the free electricity on the inner coating can reach a far higher degree than in the ordinary thin jars, in which, if a spontaneous discharge does not occur, a fracture of the glass is to be feared.

The mutual induction of opposite electricities of the two coatings, in consequence of the great thickness of the glass, is less perfect than with thinner glass. With the same quantity of coating, and with the same density of the free electricity on the inner coating, less electricity will be accumulated in thick glass jars than in those of thin glass; in general, therefore, the quantity of electricity which a thick glass jar can receive is less, but the tension of the free electricity on the inner coating, and consequently the striking distance, is greater. It is to be expected that with the greater striking distance, other effects of the discharge will also suffer a change. All effects of the discharge stroke, in which it is chiefly desirable that a great quantity of electricity should be sent through a body, can be produced better with large, thin glass jars, but where the force of the shock is the main object, thick glass jars serve the purpose better; hence it appeared to me probable that the perforation of glass plates should take place much more easily with thick jars than with ordinary thin ones. Trial perfectly sustained my supposition. Formerly, in using large, thin jars, a great number of revolutions of the machine were necessary to charge the battery sufficiently for the perforation of glass, and even then the experiment did not always succeed satisfactorily; now, 20 revolutions of a very moderate electrical machine suffice to charge a thick glass jar so as to produce this effect with certainty.

Fig. 71.

good results.

The thickness of the glass jar, fig. 71, is about 1 line; each coating has a surface of about 9 square decimetres, and the uncoated border is 22 centimetres in height.

I have not studied carefully the influence of the thickness of the glass upon the effects of the discharge stroke, and only make this notice in order to draw the attention of other physicists to the point. It is much to be wished that Riess would take up this subject, since he has already labored in this field with such generally acknowledged

[graphic]
« AnteriorContinuar »