]\gt ; On the Refraction of .
435 1881 .
] From this we may draw the conclusion , that either the materials of the earth have about the strength of granite at 1,000 miles from the th surface , or they have a much greater strength nearer to the surface .
This investigation must be regarded as confirmatory Sir William Thomson 's view , that the earth is solid nearly throughout its o hol mass .
According to this view , the lava which issues from volcanoes arises from the melting of solid rock , existing at a very high temperature , at points where there is a diminution of pressure , or es rom lse comparatively snlall vesicles of rock in a molten condition .
XIX .
" " On the Refraction of Electricity.\ldquo ; By ALFRED TRIBE , F.I.C. , Lecturer on Chemistry vich College .
Communicated by Dr. GLADSTONE , F.R.S. Recelved June 7 , 1881 .
On December 15 , 1880 , I had the honour of communicating to the Royal Society the latest results of my work on electric distri bution .
In that paper there is included a description of results lts which form tlJe basis of a graphic and electro-chemical method of inves tigating the field of electrolytic action .
These results may be classed un er ree der th .
Dislribntion of electricity on metallic conductors in electrolytic media .
2nd .
Physical differences in corre.arts o non-homogeneous electrolytic fields .
3rd .
Direction which energy is transmitted .
As the detailed account of these experlmen ts has not yet been published , it is necessary for the appreciation of the evidence to be adduced to give in this place the groundwork of that part method relating to the direction in which the energy is transmitted .
It will be convenient to do this under three heads .
Let it be remembered that a rectangular electrolytic cell was used , that the electro was a solution of copper sulphate , and the electrical relations th liquid were ascertained by immersing in it a rectangular ver lver late ( called an anaiysing plate or analyser ) , on which the ions were deposited .
In all cases the positive ion separates and is distributed on that part of the plate which may be supposed to receive \mdash ; electrl fication , or by which the energy enters the analyser , while the \mdash ; ion separates on that part of the same plate which receives electrification , or from which the energy emerges . .
When the course of the energy*is parallel with any two edges an analyser , and therefore with the sides , the boundary llnes of the ions on both sides of the plate are parallel with the plane of the elecassume that the energy in a homogeneous lield in straight parallel lines isturbed by from one electrode to allother , and that this course is not distur an analyser . .
When the course of the energy is parallel with the sides of an analyser , makes with its edges an oblique angle , the boundary lines of the ions are still parallel with the plane of the electrodes , but necessarily cross the sides of the plate obliquely to its edges .
This obliquity varies propol.tionately with the angle of inclination of the analyser to the energy , so that in all cases the boundary lines ars perpendicular to the direction of its transmission . .
When the course of the energy makes an oblique angle with the sides of the analyser , neither the magnitude nor the boundary configuration of the same electrification is identical on the two sides of the plate .
The boundary of the positive ion on the side in opposition the direct course of the influence is now markedly convex , and greater in magnitude than on the reverse side , where , moreover , the boundary configuration of the sams ion is markedly concave .
Again , the boundary line of the negative ion on the first-named side of the analyser is conca , vs and smaller in magnitude than on the reverse side , where the boundary of this ion is convex .
The convexity and concavity of these several boundaries increase as he direction of the energy approaches a perpendicular to the sides of the analyser .
Both the cIasses of distributions described in and are named nonparallel .
The position of the ions on an analyser , and the character of their boundary lines determine then , and with accuracy , the ection of the energy in the electrolytic field relatively to either side or edge of the analysing plate .
1881 .
] On the Refraction of Electricity .
437 Many questions have occurred to me in the nvestigation of which this method might be expected to afford material assistance .
ut But th one of immediate a , ttraction was whether electricity is endowed , like light , heat , and sound , with the quality of refraction .
From the blance of the fundamental laws of the forms of energy , I general resem ance instituteA experiments in the expec a ng tation of findin an answer in the affirmative to this question , the better conducting electrolytic medium being taken as the electric analogue of the more rare medium in light .
Refraction .
My first trials were made with double convex-shaped bladders .
These gave what I took to be a slight evidence of refraction , thoug the result was far from satisfactory .
Trlaugular-shaped cells were next employed , made by placing diaphragms of parchment-paper obliquely across the electrolytic cell near its ends .
In this way unmlstakeable proof was obtained of the bending of the energy pass the line of demarcation of the two media .
When the influence pass from ome medium t other perpendicularly , i.e. , when the dia- phragms were parallel to the plane of the electrodes , no refraction whatever took place .
On further consideration the arrangement which appeared less open to objection , and at the same time the most simple and theoretically the best , was a refracting cell , having parallel sides of some material permeable to the electl'ic in nfluence .
In the first instance parchmentpaper was the material employed .
Two sheets of this substance were fixed in a verticalposition across an electrolytic cell , 380 millims. 128 millims. broad , and 128 millims. deep , at an angle of .
They were parallel to one another , 76 nlillims .
apart , but equidistant from the respective ends of the cell .
An unit current was employed , and copper electrodes of the breadth and depth of the outer or transmitting cell .
A 1 per cent. solution of sulphate was placed in the inner or refracting cell , and a concentrated solution of the same salt in the transmitting cell .
On placing successive analysers eng wlse th in several parts of the central line joining the electrodes , parallel distributions were recorded by all the plates in the transmitting cell , but the one in the re ractlng ce cell recorded a non-parallel distribution of a most pronounce character and it was evident from the degree of curvature of the ions tha the COUl.se of the on passing into the medium of less conductivity , had bent out of its original course some The difficulty of keeping the parchment diaphragms as rigid as was necessary for a more extended study of this phenomenon of *Unless the contrary is stated , it is to be understood that nalysers 4 millims. were used , and placed in the with ortel .
438 Mr. A. Tribe .
[ June 16 , led to their substitution by others of unglazed earthenware .
These consisted of the sides of a large rectangular porous ground to as uniform a thickness as possible .
With this alteration , but with all the other above-mentioned conditions , the following experiments were made .
Analysers were successively placed in the positions ( all perpendicular to the electrodes ) shown in the annexed diagram , exhibiting a horizontal section drawn to about one-fifth.the scale .
The analysers , 2 millims. from electrodes , recorded parallel distributions .
The analysers , 2 millims. from diaphragm , reoorded not absolutely but very nearly parallel distributions .
This slight nonparallelism was not noticed in the analogous experiment with the parchment diaphragms , and 1 am disposed to attribute it to a greater diffusion in this case occasioning a less sharp line of demarcation between the media .
The analyser , midway between the diaphragms , recorded a nonparallel distribution of a most pronounced character .
The of curvature showed that the energy had been refracted through some while the position of the ions proved that the bending was towards a perpendicular to the refracting surface .
Furthermore , the symmetry of the curve showed that the plane of refraction was the same as that of tlJe incident energy .
The , indicated a result almost identical with The analysers , showed a much smaller deviation from the original course of the energy .
The positions of the relatively good and bad conducGing media were now reversed .
The 1 per cent. solution was placed in the transmitting 'cell , and the concentrated in the refracting cell .
An analyser at xecorded a non-parallel , and the cul'vature and position of the ions showed that the course of the energy had been bent about .
But in this case the refraction was from a perpendicular to the refracting surface .
It was now natural to anticipate that the refraction , with a given 1881 .
] On the Refraction of Electricity .
439 incidence , would decrease or increase as the media in the and transmitting cells approached or receded in con ng power .
ductin This was proved to be the case , first , by replacing the 1 per cent. solution of copper sulphate in the refracting cell by a 2 , 3 , 5 , and 10 per cent. solution respectively , when analysers at recorded respectively a decreasing refraction as the media in the cells approached equa equality .
And , secondly , by placing a 5 per cent. solution of copper sulphate in the refracting cell , and successively a concentrated soIution of copper sulphate , a concentrated solution of sodium chloride , and a di lute solution ( 1\mdash ; 11 ) of oil of vitriol in the transml tting ] .
Analysers at showed , in the case of the copper sulphate , a retraction of abou about in the case of the sodium chloride , , and in that of the dilute sulphuric acid , .
The order of the increase in refraction is here the as that of the conductivities of the liquids .
bici ated that the refraction would decrease as the Again , it was an lclpa incidence decreased .
This was proved to be true by setting the diaphragms so that the angle of incidence should be 45 , 30 , and 15 respectively .
With a 5 per cent. solution of copper sulpbate in the refracting cell , and dilute sulphuric acid in ransml ng ce , transmittin cell , refraction of about was obtained when the incidence was about when at , and about when at Relation between Electric Incidence and That part of the electro-chemical method hich has been already described supplies thres means of determlnlng the ben lng din of the course of the energy on its passage into the refracting cell .
First , by the curvature of the ions on an analyser with the shorter edges of the plate upright , and its length perpendicular to the electrodes .
But the practical impossibility of obtaining accurate measuremen ts of these curves , renders this plan only roughly quantitative . .
econ , by placing an analyser at various angles to a right line nlng the electrodes , until it records a distribution .
This is the case en the course of the energy is parallel with the sides of the analysing plate , and the angle enclosed between the plate and the right line to consequelltly expresses the amou1lt of refraction .
Th objections to this plan are twofold .
It is very tedious , and it is also difficult to determine with sufficient accuracy the angle of the plate as it stands in the cell .
The refraction numbers given above were obtained by this second method , and for last of the reasons just mentioned I regard them only as approximately correct .
The third plan , based on the fact that the course of the influence , parallel to the sides of an analyser , is perpendicular to the boundary lines of the ions set free , is the one I have employed for the more accurate determination of the relation veen electric incidence and refraction .
way that their edges were parallel with the respective sides and ends of the rectangular electrolytic cell ; the disks , so that a line passing through their csntres was coincident with a line joining the electrodes .
As it has been already shown that the course of the energy in the transmitting and refracting cells was in the same horizontal plane , it was to be anticipated that the boundary lines of the ions on the analysers placed in the positions just described would be straight .
Such is the case with the boundary of the negative ion , which consequently is taken for purposes of measurement .
Were the course of the energy in thelnner cell identical with that in the transmitting cell , the boundary lines , in accordance with the rule in , would be perpendicular to a right line joining the electrodes , and this might be named the zero line , or line of no refraction .
And were the analysers in the position described , the zero line would be perpendicular to the csntral line on the disks to the edges of the squares , which are parallel with the sides of the cell .
It follows that , were the course of the energy diverted from a right line joining the electrodes ( its original course ) , the boundary line of the ion would deviate passu from the line of no refraction .
The numbers expressing this obliquity or amount of refraction in five series of trials are given in the annexed table .
In Series I , II , III , square analysers were used , and for and V the disks already described .
ationIII I 1881 .
] On the Refraction of Electricity .
441 These data give the following ratios between the angles of incidence and refraction\mdash ; and the ratios between the sines of the angles of incidence and sin Taking the results as thus appear , I think they justify the conclusion that the more probable relation between electr ca ence 1 incid and refraction is as the sines of their respective angles .
And further , as the experimental difficulties are overcome and the inherent sources of error appreciated , this relation may be still more rigidly foun to obtain .
This electro-chemical method being new , it would be well to supplement , if possible , the evidence it has furnished of the laws of electric 1'efraction by one based on more familiar principles .
I am engage .
perfecting a method founded on those employed by De la Rive in 1825 , and by Professor Adams in 1875 , for investigat ng the laws of electric distribution in electrolytes , which I have reason to believe will demon strate in another way the main facts set forth in this communication .
Conclusions .
I. Electricity passes without alteration of direction from one electrolytic medium to another differing from it in conductivity , when the course is perpendicular to the surfaces of contact .
II .
Electricity , on passing obliquely from one medium um to the other 442 On the Refraction of .
Electricity .
[ June 16 , suifers reffiction , and in the same plane ; towards the perpendicular , when from a better to a worse conductor , and from the perpendicular when from a worse to a better conductor .
III .
The refraction increases or decreases as the media recede from or approach one another in conductivity .
IV .
The refraction increases as the incidence increases .
[ NOTE.\mdash ; It is known that the flow of electricity in an electrolyte ollows the same laws as the flow of heat in a conductor when the condition has become permanent .
It readily follows , that when two electrolytes of different specific resistance , suppose a strong and a weaker soIution of sulphate of copper , are in contact , a change in the direction of the flow takes place in passing out of the one into the other .
Let be the potential in the first medium , the angle of incidence , or in other words , the inclination of the tangent plane at any point in the surface of separation to the tangent plane to the equipotential surface in the first medium which passes through the same point , the specific conducting power , an element of a section of the common surface by the plane of incidence , an element of a normal the equipotential surface , the flux of electricity , and let , be for the second medium what v are for the first .
Then we have But if we take an elementary closed curve in the surface of separation and make it the base of elementary tubes in the two media bounded by lines of flow , since the same quantity of electricity must flow through the two tubes , and the areas of their sections are as , we have and eliminating between these two equations , we get , since and therefore tan so that according to theory the tangents , not sines , of the angles of incidence and refraction are in a given ratio .
If the potentiaIs in the two media , instead of being equal at the surface of separation , differ by a constant quantity , as may conceivably be the case when the two electrolytes are diflerent in nature , so that different chemical actions on in them ' .
the above relation would not be disturbed .
The results obtained by Mr. Tribe agree rather better with tlJe law of sines than with the law of tangents .
But several lces 1881 .
] Capt. Abney .
On the Spectrum of Sodium .
tend to introduce greater or less errors .
For example , the flow in the second medium must have been somewhat sturbed by the circumstance that the non-conducting walls of the interme ce diate cell were parallel to the direction of the incident , not refracted flow , which may possibly disturbed cour the course even as far as the middle of the 11 .
A , there would have been a muc lar er resistance in the ce .
galn , porous wall than a stratum of equal thickness of one of the electrolect lytes , and the thickness or porosity of the wall , or the proportion of the two electrolytes imbibed by it , may have vat.led somewhat in a lateral direction .
The numbel .
S obtained cannot therefore be deemed sufficient to decide between two such laws as that of sines and that of tangents .
In case of the second medium being better conductor , it is evident that the law of sines would lead to extravagant resu ts , as lts there can be no such thing as total internal .
The alteration in the direction of the equipotenlial surfaces , and , therefore , of the lines of flow , in passing Irom one metal into another of different conducting power , has already been investigated experimentally by Quincke , and the results of experiment compared with theory .
Pogg .
Ann. , vol. xcvii [ 1856 ] , p. ) .
G. S. , 14 June , 1881 .
] .
" " Note on the pecfrum of Sodium By ) W. de W. ABNEY , R.Hi .
, F.R.S. Received June 14 , 1881 .
On examining the spectra of different metals , there is one point which is striking in the extreme , , the absence of any very marked lines in the region between and , which latter num er we may take as the visible limit of the spectrum .
With the except on of the well-known pair of lines of potassium , I am not aware that any lines in metallic spectra , which have been carefully studied , have eeu found below this limit , though recently , in the spectra of some th rarer earths , I believe some few lines have been recorded .
Having photographed the emission spectra produced in the arc of several metals , it appears , so far as examination has been made , that only those which can be volatilised at a low temperature have aIly lines in the infra-red region .
Sodium is an example of this .
It has a pair of lines at wave.length of about 8187 and 8199 of an intensity of about 3 , taking the intensity of lines as 10 .
It no 11 be nofed that the difference in wave-length between this pair is greater than that of the lines .
They do not seem to have any dark lines in the ] spectrum , though there arc faint lines which lie close to these wave-lengths .
In the calcium spectrum there is a pair of very faint lines which lie between 8500 and 8000 .
Their exact wave-lengths have not at present been determined .
XXA11 .

