Page images
PDF
EPUB

existence had been foretold by Maxwell's genius; and with suitable apparatus stationary electric waves are now almost as readily made evident as are those of sound. Hertz's brilliant success stimulated his fellow countryman, Otto Wiener, to undertake the apparently hopeless task of producing and studying stationary light waves. Wiener's admirable work* excited great interest on the continent of Europe, but it has been singularly neglected in England and America. It is worth much more than a passing notice.

Assume a plane silvered mirror upon which a bundle of rays of monochromatic light falls normally so as to be reflected back upon its own path. The superposition of reflected and direct waves causes a system of stationary waves, but under ordinary conditions these are wholly imperceptible. The nodes are formed upon a series of planes obviously parallel to the reflecting plane at successive distances of a half wave length. If now we consider a plane oblique to the mirror, it will cut these successive nodal planes in parallel lines whose distance apart will be greater in proportion as the oblique plane approaches parallellism to the mirror. Although a half wave length of violet light is onlyth of a millimeter, it is easy to conceive of the cutting plane forming so small an angle with the mirror that the distance between the parallel nodal lines shall be a thousand times a half wave length. Such would be the case if the inclination of the cutting plane is reduced to a little less than four minutes of arc. The nodal lines would then be th millimeter apart, and readily capable of resolution if their presence can be manifested at all. Imagine a very thin transparent photographic film to be stretched along the oblique cutting plane, and developed after exposure to violet light as nearly monochromatic as possible. Then the developed negative should present a succession of parallel clear and dark lines, corresponding to nodal and anti-nodal bands along the oblique plane, the photographic effect being annihilated along an optical nodal line.

The realization of a photographic film thin enough for such an experiment is quite conceivable when we remember that under the hammer gold is beaten into leaves so delicate that 8000 of them would be required to make a pile one millimeter thick. By electro-chemical deposit Outerbridget has made films of gold whose thickness is only th of a millimeter, orth of a wave length of sodium light. Wiener obtained a perfectly transparent silver chloride film of collodion, whose thickness was about 3th of a wave length of sodium light. This was formed on a plate of glass and inclined at a very

00000

*O. Wiener. Wiedemann's Annalen, xl, p. 203, 1890.
Journal of the Franklin Institute, vol. ciii, p. 284, 1877.

small angle to a plane silvered mirror which served as reflector. From an electric arc lamp the light was sent through an appropriate slit and prism, so that a selected spectral band of violet fell normally on the prepared plate in the dark room. The developed negative presented the alternate bands, in perfectly regular order more than a half-millimeter apart. Various tests were applied to guard against error in interpretation, and the existence of such stationary waves was proved beyond all doubt.

These waves, moreover, when polarized light was employed, furnished the means of determining the direction of vibration with relation to the plane in which the light is most copiously reflected when incident at the polarizing angle, and thus of subjecting to experiment the question as to whether the plane of vibration is coincident with this plane of polarization or is perpendicular to it. The former of these views was held by Neumann and MacCullagh, the latter by Fresnel. Let a beam of polarized light fall upon the mirror at an angle of about 45°. If the vibrations in the incident beam are parallel to the mirror, and hence perpendicular to the plane of polarization, those of the reflected and incident beams will be parallel to each other and hence capable of interference. But if the vibrations of the incident beam are in a plane identical with that of incidence, and hence in the plane of polarization, the vibrations of incident and reflected beams are in mutually perpendicular planes and hence cannot interfere. Wiener obtained interference fringes when the light was polarized in the plane of incidence, while that polarized in the plane perpendicular to this gave no trace of interference. The theory of Fresnel was thus confirmed experimentally. Again, the familiar phenomena of Newton's rings show us that on changing media there is a change of phase of the incident light, else the central spot where the two surfaces come into optical contact would be white instead of black. But there has been difference of opinion as to whether this change of phase occurs at the upper surface of the air film, where the light passes from glass to less dense air or at the lower surface where it passes from air to more dense glass. In the latter event there should be a node at the reflecting surface. Replacing the silvered plane surface by a lens in contact with the photographic film, Wiener obtained circular fringes with no photographic action at the center, showing the nodal point to be at the point of contact, and thus again confirming the theory of Fresnel.

[To be continued.]

ART. XXX.-The Quantitative Determination of Perchlorates; by D. ALBERT KREIDER.

[Contributions from the Kent Chemical Laboratory of Yale College-XLIV.]

THE method usually employed for the quantitative determination of perchlorates, by igniting to the chloride and weighing the halogen as the silver salt, is indirect and subject to error, especially as my experience proved, where the free acid is to be determined and where, consequently, an alkali which is apt to contain chloride is used to form the salt for the ignition. To purify the salt for this method only adds to the complication, and therefore a more satisfactory process was sought. In a recent article from this laboratory by Professor Gooch and myself, a method for the detection of alkaline perchlorates associated with chlorides, chlorates and nitrates was detailed, with mention of certain efforts towards a quantitative determination. As throwing light upon the peculiar properties of perchlorates, and as an introduction to the satisfactory method which I have finally developed, some of the results of these earlier efforts will here be given.

In studying the properties of perchloric acid in the form of its potassium salt, we found that when treated with potassium. iodide in the presence of boiling phosphoric acid, no reduction of the perchlorate is effected; unless indeed, the boiling be continued till the temperature rises to 215° to 220° C, where the meta-phosphoric acid begins to form. But when the metaphosphoric acid (made by heating the syrupy ortho-acid to 360° C) is directly applied in the presence of potassium iodide and kept at a temperature of about 200° C, iodine is copiously evolved. To test this reaction quantitatively a number of experi ments were made in an apparatus consisting of a retort, into the tubulature of which a glass tube was carefully ground and prolonged so as to reach to the bottom of the bulb and serve for the passage of a current of carbon dioxide, used to expel the air and carry the iodine into the receiver. The neck of the retort was bent so as to reach to the bottom of an Erlenmeyer receiving vessel, containing a solution of potassium iodide, which was trapped by a side-necked test tube. After introducing the perchlorate with the iodide and meta-phosphoric acid, all air was expelled by carbon dioxide and heat applied. The iodine collected in the receiver was titrated with decinormal thiosulfate, from which the perchlorate was calculated.

Table I gives the results of several experiments performed in this way, which prove that even with a large excess of potas

*This Journal, vol. xlviii, p. 38.

sium iodide the perchlorate is so slowly reduced that the hydriodic acid escapes before the reduction is completed. In order to delay the distillation of hydriodic acid until the perchlorate had been completely reduced, the potassium iodide of experiment (3) was introduced in a short tube sealed at one end, so that the meta-phosphoric acid could attack it only slowly, and the heat quickly raised to about 300° C, but evidently without advantage. In experiment (4) the iodide was introduced in the same way, but the heat was applied gradually and more moderately, with considerably improved results.

[blocks in formation]

A complete reduction of the perchlorate evidently necessitated the means of introducing the iodide in sufficient quantity and at will.

For this purpose the tube serving for the introduction of carbon dioxide was enlarged so as to hold the iodide, which could then be added to the solution at any time by a manipulation of the rubber conducting-tube for carbon dioxide, which would draw the acid up to the iodide and, retreating, would carry back an easily regulated quantity of the latter.

[blocks in formation]

Table I gives a number of results obtained in this way. Experiments (10), (11) and (12) differed from the others only in the employment of a bulb pipette instead of the retort: one end being bent so as to reach to the receiver and the other cut off rather short with a tube ground into it, serving the same purpose of conducting carbon dioxide and holding potassium

iodide; the greater inclination of the potassium iodide tube made possible by this change appearing to offer advantages for the more gradual and regular introduction of the iodide. The amount of meta-phosphoric acid used was in all cases 15 cm3. In experiment (13) heat was applied by means of a bath kept at 230°.

While several of these determinations gave only admissible errors, the irregularity of the remainder and the uncertainty in striking just the proper conditions for good results, proved the method worthless at least in that shape.

The experiments of Table III record the results of adding the acid drop by drop to an intimate mixture of the powdered perchlorate and iodide kept hot.

[blocks in formation]

The high results of this table doubtless point to the dissociation of hydriodic acid or to the partial reduction of the metaphosphoric acid in the temperature, which would naturally rise higher where so small an amount of liquid was present. But when the meta-phosphoric acid was there in greater amount the distillation of the hydriodic acid before the complete reduction. of the perchlorate could not be prevented.

An ordinary mixture having thus been found insufficient to hold the hydriodic acid to the reduction of perchlorates, a search for some compound in which the perchlorate could be fused with an excess of potassium iodide and the mixture thus obtained subjected to the action of meta-phosphoric acid resulted in the employment of zinc chloride. Anhydrous zinc chloride was found to fuse at about 200° C. The perchlorate and iodide could be added to this fusion and the whole melted, thoroughly diffused and cooled without any evolution of iodine. This mass, when treated with meta-phosphoric acid in the apparatus previously employed, melted gradually with a copious evolution of iodine. Table IV shows the quantitative action. The amount of zinc chloride used was roughly taken about equal to that of the iodide.

[blocks in formation]
« PreviousContinue »