486 Pritchard , Urban , M.D. , F.R.C.S. Radcliffe , Charles Bland , M.D. , F.R.C.P. Ransom , Arthur , M.A. , M.D. Ranyard , Arthur Cowper , F.R.A.S. Rawlinson , Robert , C.B. , M.I.C.E. Reinold , Prof. Arnold William , M.A. Rodwell , George F. , F.R.A.S. , F.C.S. Samuelson , Bernhard , M.I.C.E. Spiller , John , F.C.S. [ Mar. 3 , Stoney , Bindon Blood , M.A M.I.C.E. Tidy , Prof. Charles Meymott M.B. , M.R.C.S. , L.S.A. Traquair , Ramsay H. , M.D. Warren , Charles , Lieut.-Col. R.E. W atson , Rev. Henry William , M.A. Williams , Charles Theodore , M.A. , M.D. , F.R.C.P. Wright , Charles R. Alder , D.Sc .
Wright , Edward Perceval , M.A. , M.D. , F.L.S. Sir J. Conroy .
The following Papers were read.:\#151 ; I. " Some Experiments on Metallic Reflexion .
No. II .
" By Sir J OHN Conroy , Bart. , M.A\#187 ; Communicated , by Professor Stokes , Sec. R.S\lt ; Received February 12 , 1881 .
Professor Stokes did me the honour of communicating to the Royal Society a short paper ( " Proc. Roy .
Soc. , " vol. 28 , p. 242 ) giving an account of some determinations I had made of the values of the principal incidence for , and the principal azimuth of , red light , when reflected by a gold plate in contact with air , water , and carbon bisulphide .
These experiments I have continued , using light of different refran-gibilities , and gold and silver plates , and also thin films of these metals .
The observations were made by causing a beam of light which had been polarised elliptically , by passing through a nicol with its principal section at an angle of 45 ' with the plane of incidence , and a " quarter undulation plate , " with one of the neutral axes in that plane , to fall on the metal plate , and then altering the angle of incidence till the reflected light was plane polarised .
The angle of incidence at which this was the case was different , according as one or other of the neutral axes of the quarter plate was in the plane of incidence , and the mean of these two values was taken as the true principal incidence .
The goniometer used in these experiments was described in the former paper , but had an additional vertical divided circle attached to the inner end of the tube fixed to the collimator arm , in order to facilitate the adjustment of the quarter plate , and to permit the plate being changed , without the necessity of determining each time the position of the neutral axes .
Brewster 's law , that the tangent of the angle of polarisation is equal 1881 .
] Some Experiments on Metallic Reflexion .
to the refractive index of the medium , not having been , as far as I was aware , ever experimentally verified in the case of transparent bodies in contact with media other than air , it appeared desirable to make some experiments on this point , though there was every reason for supposing that the law would be found to hold good .
A crown glass prism was fastened to the vertical stage , the quarter undulation plate and the analysing nicol removed , the polarising nicol placed with its principal section in the plane of incidence , and the angle of polarisation determined , a paraffine lamp being used as the source of light .
The thin glass vessel was then placed on the vertical stage , and filled successively with water and carbon tetrachloride , and the angles of polarisation again determined .
Glass Prism .
o / o / Angle of polarisation in air. .
... ... ... ... ... . .
57 14 57 02 , , " water ... ... . .
49 41 49 43 , , , , carbon tetrachloride ... 46 32 46 23 The angles of polarisation in air are the mean of twelve and four observations respectively , those in water of eight and twelve , and those in carbon tetrachloride of twelve observations .
When the surface of the glass was in contact with water or carbon tetrachloride , its reflective power was so much diminished that it was not possible to determine the angle at which the minimum amount of light was reflected , and therefore the angles at which the light ceased to be visible were observed , and the mean of these taken as the angle of polarisation .
The tangents of the angles of polarisation in water and carbon tetrachloride , when multiplied by the refractive indices of these liquids , give the following values for the angle of polarisation in air:\#151 ; O I O / Angle of polarisation in air observed ... ... ... . .
57 14 57 02 Angle of polarisation in air calculated from observations in water ... ... ... ... ... ... ... ... ... ... 57 28 57 33 Angle of polarisation in air calculated from observations in carbon tetrachloride ... ... ... ... ... 57 01 56 52 The distilled water which was used in these experiments had not been freed from air , and this may account for the values of the angles of polarisation in air , calculated from the measurements made in water , being somewhat too high .
These observations show that , within the limits of experimental error , Brewster 's law holds good for glass in contact with water and carbon tetrachloride , as well as air , and that in all probability it is universally true for transparent bodies .
488 Sir J. Conroy .
[ Mar. 3 ; The same gold plate which had been used in the previously described experiments was fastened to the stage of the goniometer , and the principal incidences and azimuths observed with yellow and blue light , when the plate was in air , water , and carbon bisulphide .
An alcohol flame with a salted wick and a paraffine lamp , with a glass trough filled with ammonio-sulphate of copper placed in front of it , were used as the sources of light .
The light transmitted by the deep red glass used in the previous experiments and by the ammonio-sulphate was not homogeneous .
The light of the greatest intensity in the one case was slightly more refrangible than C , and in the other of a refrangibility nearly midway between F and Gr .
The polarising nicol was placed with its principal section at an angle of 45 ' with the plane of incidence in each of the four quadrants successively , and two observations made with the nicol in each position ; so that the value of the principal incidence and the principal azimuth for each position of the " quarter undulation plate " is the mean of eight observations .
The following table contains the results of these experiments ; the values for red light , which had been previously obtained , are included in it for the sake of comparison :\#151 ; Gold in air , red light .
Gold in air , yellow light .
Gold in air , blue light .
Gold in water , red light .
Gold in water , yellow light .
Gold in water , blue light .
Quarter plate .
Quarter plate .
Corrected L YI .
numbers .
P.I. P.A. p.i. P.A. P.I. P.A. O / G / o / O t O / o / r Plate at A. 80 13 35 52 78 03 35 43 J Plate at B. 70 45 35 57 73 45 35 21 1 Mean value 75 29 35 54 75 54 35 32 76 0 35 27 r Plate at A. 76 33 38 12 73 59 36 27 | Plate at B. 63 33 37 39 72 55 36 20 L Mean value 72 33 37 55 73 27 36 23 73 28 36 22 r Plate at A. 65 32 30 19 60 27 32 31 s Plate at B. 69 03 29 44 71 51 32 09 L Mean value 67 17 30 01 66 09 32 20 67 24 29 47 r Plate at A. 76 46 37 13 74 46 36 0 Plate at B. 66 46 35 50 70 25 36 49 L Mean value 71 46 36 31 72 35 36 24 72 46 36 23 r Plate at A. 73 47 38 05 70 05 37 28 -j Plate at B. 65 29 38 27 68 51 36 22 36 52 L Mean value 69 38 38 16 69 28 36 55 69 28 f Plate at A. 61 51 28 30 57 09 31 35 s Plate at B. 65 26 29 06 70 59 31 01 1 Mean value 63 38 28 48 64 04 31 18 63 36 28 38 1881 .
] Some Experiment , s on Metallic Reflexion .
489 Quarter plate .
Quarter plate .
Corrected I. VI .
numbers .
p.i. P.A. P.I. P.A. P.I. P.A. o / O / o / O / O / o / Gold in carbon | r Plate at A. 76 10 37 48 71 37 36 59 bisulphide , \lt ; Plate at B. 62 44 37 40 68 26 36 43 red light .
L Mean value 69 27 37 44 70 01 36 51 70 03 36 48 Gold in carbon i r Plate at A. VI 0 39 09 67 02 37 17 bisulphide , \lt ; Plate at B. 61 50 38 24 66 02 37 10 yellow light .
1 L Mean value 66 25 38 46 66 32 37 13 66 32 37 12 Gold in carbon | r Plate at A. 59 51 31 30 54 24 29 17 bisulphide , \lt ; Plate at B. 61 30 33 13 67 56 31 08 blue light .
1 L Mean value 60 40 32 21 61 10 30 12 60 40 32 23 The last column gives the values as corrected by Professor Stokes ' method ( Zoo .
cit. , p. 248 ) .
The table shows , that for red and yellow light the observations made with plate I , and for blue light those made with plate YI , are but of little value ; whilst for red and yellow light the Observations made with plate YI , and for blue light those made with plate I , lie very close to the corrected numbers .
All subsequent experiments were therefore made with plate YI for red and yellow light , and with plate I for blue light .
The gold plate used in these experiments had been polished with rouge , and from time to time rubbed with a little rouge , and then with a clean chamois leather to remove the polishing powder .
From some experiments with a silver plate , it appeared probable that the optical constants of the plate might depend to a certain extent on the nature of the polishing powder used , and therefore the gold plate was dismounted and polished with putty powder , and then well rubbed with a clean chamois leather , and the principal incidences and azimuths determined .
Four observations were determined in each position of the quarter plate , two with the principal section of the polarising nicol on the right , and two on the left of the plane of incidence ; the plate YI being used with red and yellow light , and plate 1 with blue light .
Red light .
Yellow light .
Blue light .
p.i. P.A. P.I. P.A. PI .
P.A. Gold plate in air .
r Plate at A. \#171 ; \lt ; Plate at B. L Mean value o / 75 57 71 45 73 51 O / 41 17 41 29 41 23 O / 72 18 71 21 71 49 O / 41 40 40 57 41 18 o / 67 38 66 34 67 06 O / 34 44 34 55 34 49 After an interval of two months the measurements were repeated , the plate being first rubbed with a clean chamois leather .
490 Sir J. Conroy .
[ M " ar .
3 , Bed light .
Yellow light .
Blue light .
p.i. P.A. p.i. P.A. p.i. P.A. o / o / o , / O / o / O t Gold plate in 4 r Plate at A. 76 20 42 40 72 16 41 07 67 59 37 12 J Plate at B. 71 49 42 05 70 59 41 14 66 29 35 50 air .
1 Mean value 74 04 42 22 71 38 41 10 67 14 36 31 The measurements were then made with the plate in water and carbon bisulphide .
Bed light .
Yellow light .
Blue light .
p.i. P.A. p.i. P.A. p.i. P.A. o / O / o / O / o / O / ( Gold plate in I r Plate at A. 72 43 42 36 69 23 42 10 63 50 36 44 Plate at B. 68 07 42 39 67 28 41 01 62 39 35 12 water .
I. L Mean value 70 25 42 37 68 25 41 35 63 14 35 58 Gold plate in J r Plate at A. 73 17 42 10 67 11 39 25 64 27 36 07 Plate at B. 68 18 42 24 66 36 42 25 62 25 36 41 water .
II .
1 L Mean value 70 24 42 17 66 53 40 55 63 26 36 24 Gold plate in i f Plate at A. 71 52 42 37 66 58 41 26 61 41 36 19 carbon \lt ; Plate at B. 66 57 42 29 66 14 41 56 58 29 37 35 bisulphide .
1 L Mean value 69 24 42 33 66 36 41 41 60 05 36 57 The mean values of these two series of determinations are\#151 ; Red light .
Yellow light .
Blue hght .
P.I. Gold in air ... ... ... ... . .
73 57 " water ... ... ... ... .
70 24 " carbon bisulphide. .
69 24 P.A. P.I. PA .
P.I. P.A. O / 41 52 o / 71 43 O / 41 14 O / 67 10 O / 35 40 42 27 67 39 41 15 63 20 36 11 42 33 66 36 41 41 60 05 36 57 The values of the principal azimuths are considerably higher ( about 6 ' ) than those previously observed ; and in the case of red and yellow light , when the gold plate was in air or water the principal incidences are lower ( about 2 ' ) than with the plate polished with rouge ; but with blue light , and also when the plate was in contact with carbon bisulphide , there was but little difference in the values of the principal incidences .
The plate was again cleaned with rouge , and the measurements in air repeated .
Bed light .
Yellow light .
Blue light .
p.i. P.A. p.i. P.A. p.i. P.A. ) O / O / o / \#171 ; / o / Q / f Plate at A. 76 41 41 07 Gold in air .
-s Plate at B. 72 05 41 16 1 Mean value 74 26 41 11 f Plate at A. 76 31 40 47 72 46 40 39 69 07 34 44 Gold in air .
^ Plate at B. 72 11 40 15 72 06 39 40 67 55 33 05 L Mean value 74 21 40 31 72 26 40 09 63 31 33 54 Sortie Experiments on Metallic Reflexion .
1881 .
] Polishing the plate with rouge appears to have had the effect of increasing the values of the principal incidences and diminishing those of the azimuths , but not to such an extent as might have been anticipated from the difference between the values obtained with the plate before and after it had been polished with putty powder .
This may be due to the original smoothing and polishing with rouge having altered the surface of the plate to a greater extent than could be effected by repolishing with rouge the already polished plate .
The gold plate , which was a long narrow one , having been constantly rubbed in one direction whilst being polished , it was thought possible that the direction of the furrows produced by the polishing might influence the result : the plate was therefore turned so that these were perpendicular to the plane of incidence , instead of parallel with it .
This alteration in the position of the plate did not appear to alter the value of the constants .
Gold in air .
Tied light .
PI .
P.A. O / O / Plate at A ... . . .
76 42 40 44 Plate at B. . . . . .
72 23 39 54 Mean value. . . . .
74 32 40 19 In order to obtain a gold surface free from the influence of any polishing powder , a sheet of gold leaf was placed between two pieces of paper , and the paper and gold cut into strips with a sharp pair of scissors , the gold leaf transferred to glass slips , floated out with water , in the manner described by Faraday , and the water allowed to drain off , leaving the gold leaf stretched out on the glass .
Some of the pieces of gold leaf were thinned by being floated on a dilute solution of potassium cyanide , the cyanide solution washed away with water as soon as the thickness of the gold had been sufficiently reduced , and the gold leaf left adhering to the glass .
Gold leaf treated in this way has a very fairly smooth surface , sufficiently good to act as a tolerable mirror , and both the ordinary leaf , and that thinned by the action of potassium cyanide , is transparent , the transmitted light being , as is well known , green .
In order to determine the thickness of the gold , the glass slips were carefully cleaned and weighed before the gold was transferred to them , and then , after being dried , again weighed .
The area of the gold was measured , and the specific gravity being taken as 19'36 ( " Watts ' Dictionary , " vol. ii , p. 926 ) , the thickness calculated .
The average thickness of gold leaf is stated in " Watts ' Dictionary ( loc. cit. ' ) not to exceed an ( '0001270 millim. ) , and in " Roscoe and Schorlemmer 's Chemistry " to be about '0001 millim. , 492 Sir J. Conroy .
[ Mar. 3 ; numbers that agree very fairly well with those determined in the above-mentioned manner , especially when the actual weight of the gold ( in no case exceeding *0029 grm. ) and the uncertainty as to its true specific gravity in the state of gold leaf is taken into account .
The glass slips were fastened to the goniometer , and the constants determined for the gold-air surface ; four observations being made in each position of the quarter plate , two with the principal section of the polarising nicol on the right , and two on the left , of the plane of incidence ; the plate VI being used with red and yellow light , and plate I with blue light .
No. of film .
Thickness .
2 ... ... ... ... 0 -0000645 millim. 7 ... ... ... ... 0 -0000709 5 0-0000837 " 4 ... ... ... ... 0-0001107 " 3 ... ... ... ... 0-0001135 Red light .
Yellow light .
Blue light .
p.i. P.A. P.I. P.A. P.I. P.A. 0 ?
O ?
O / O / O / O / f Plate at A. 74 43 40 50 71 05 40 51 68 18 37 21 No. 2 .
Plate at B. 70 39 41 14 69 26 40 11 65 54 37 02 L Mean value 72 41 41 02 70 15 40 31 67 06 37 11 f Plate at A. 75 16 40 42 72 32 40 10 68 56 35 11 No. 7 .
v Plate at B. 71 48 40 26 70 45 40 21 66 20 35 36 L Mean value 73 32 40 34 71 38 40 15 67 38 35 23 f Plate at A. 76 44 43 27 73 59 41 45 69 51 38 30 No. 5 .
1 Plate at B. 72 54 43 30 72 41 42 22 67 51 38 16 L Mean value 74 51 43 28 73 20 42 03 68 51 38 23 f Plate at A. 76 27 43 32 73 06 42 39 69 54 38 59 No. 4 .
\ Plate at B. 72 39 42 42 71 23 42 21 67 21 38 52 \lt ; \#166 ; Mean value 74 33 43 07 72 14 42 30 68 37 38 55 f Plate at A. 76 34 42 56 73 20 42 02 69 49 39 09 No. 3 .
\ Plate at B. 72 22 42 55 72 16 41 59 67 16 38 31 t Mean value 74 28 42 55 72 48 42 0 68 32 38 50 Numbers 2 and 7 had been thinned with potassium cyanide , the other three ( 3 , 4 , and 5 , being all cut from different leaves ) were of their original thickness .
The table shows that , as the thickness of the leaf increases , the principal incidence and the principal azimuth increase , and to about the same amount for all three kinds of light , although only portions of the same leaf appear to be strictly comparable with each other .
The values obtained with the gold leaf and with the gold plate cleaned with putty powder agree fairly well together ; the principal 1881 .
] Some Experiments on Metallic Reflexion .
493 incidences and azimuths being somewhat higher , especially with blue light , in the case of the gold leaf .
It therefore appears probable that of the two sets of values for the gold plate , those obtained after the plate had been polished with putty powder are nearest the truth .
A large number of determinations of the principal incidences and principal azimuths were made with a silver plate in air , water , and carbon tetrachloride , with both the quarter undulation plates .
The silver plate had been polished with rouge , and during the course of these experiments , which were carried on at intervals during some months , rubbed with a little rouge from time to time , and although afterwards well rubbed on a clean chamois leather , in certain lights it appeared to have a reddish tinge , although the surface was most brilliant .
In some few of the earlier experiments sixteen observations were made in each position of the quarter plate , four with the polariser in each quadrant ; but most of the numbers are the means of eight observations , two for each position of the polariser .
The following table contains the measurements made with the quarter plate Vi for red and yellow light , and with quarter plate I for blue light:\#151 ; Red light .
P.I. P.A. Yellow light .
P.I. P.A. Blue light .
P.I. P.A. Silver in air .
J L 1 [ Plate at A. O / 78 13 O / 32 50 O / 76 06 O / 35 40 O / 70 06 O / 38 43 Plate at B. 75 04 31 58 75 40 35 44 73 16 39 26 L Mean value 76 38 32 25 75 53 35 42 71 41 39 04 Silver in air .
r Plate at A. 77 51 33 01 75 42 35 10 72 16 38 45 Plate at B 74 0 32 47 75 21 34 59 72 24 38 46 IL I L Mean value 75 55 32 54 75 31 35 04 72 2C 38 45 Silver in I - Plate at A. 75 13 33 42 73 11 36 47 67 09 40 06 Plate at B. 71 16 33 12 72 20 35 27 69 08 40 07 water .
l Mean value 73 14 33 27 72 45 36 07 68 08 40 06 Silver in r Plate at A. 75 04 32 52 70 27 35 41 64 32 39 36 carbon \lt ; !
Plate at B. 71 01 32 34 70 50 35 22 67 32 39 10 tetrachloride .
L Mean value 73 02 32 43 70 38 35 31 66 02 39 23 The values of the principal incidences and principal azimuths were also determined by a somewhat different method .
The principal section of the polarising nicol was placed successively in the plane of incidence , and at right angles to it , and the neutral axes of the quarter undulation plate ( plate VI being used ) at an angle of 45 ' with this plane .
The angle of incidence , and the azimuth of the analysing nicol by Sir J. Conroy .
[ Mar. 3 , which the light was reduced to a minimum , were observed ; two such observations were made , and then the quarter plate turned through an angle of 90 ' and two more observations made , so that the values for each position of the polarising nicol are the means of four observations .
The retardation produced by the plate not being equal to 90 ' for the wave-length of the light employed , the resultant beam was elliptically polarised , the major and minor axes of the ellipse being in the plane of incidence , and in the plane perpendicular to it , or versa .
The relative position of the axes depending on the position of the polarising nicol , and on the amount of retardation produced by the plate ( i.e. , whether it was less or more than 90 ' ) .
An elliptical vibration being equivalent to two rectiiinear vibrations in the planes of the axes , differing in phase by a quarter of an undulation , and of different amplitudes , this method affords a ready means of producing with the same retarding plate a difference of phase of exactly 90 ' between the vibrations in the plane of incidence and the plane perpendicular to it with light of different wave-lengths .
Hence the angle at which light polarised in this manner is reflected by a metallic plate as plane polarised light is the true angle of principal incidence ; the azimuth , however , of the reflected light is not the principal azimuth .
The azimuths are different , both in sign and in amount , for the two positions of the polarising nicol , according as the major or the minor axis of the ellipse is in the plane of incidence , the vibration in the plane of incidence being the one retarded and reduced in amplitude by the act of reflexion .
The ratio of the semi-axes of the ellipse being called 7 , and the observed azimuths and a2 , the ratio of the amplitudes of the vibrations in the plane of , and perpendicular to , the plane of incidence after reflexion by the metallic plate ( the amplitudes in the incident beam being equal ) , or , in other words , the principal azimuth 8 , is determined by the equations\#151 ; The table contains the determinations made in this way with the silver plate in air , water , and carbon tetrachloride ; the position of the principal section of the polarising nicol , with reference to the plane of incidence , being stated in the first column , and the mean values of the principal incidences , and the values of the principal azimuths , calculated by the above-mentioned formula , in the last line of each portion of the table\#151 ; which give tan cq tan a2=\#163 ; 2 .
1881 .
] Some Experiments on Metallic Reflexion .
Red light .
Yellow light .
Blue light .
P.I. A. P.I. A. P.I. A. Silver in air .
f pll .
O / 76 03 O / 36 11 O ?
75 20 O / 35 52 O / 72 07 O / 31 44 L ppd .
76 19 29 26 75 35 35 30 72 32 47 38 76 11 32 43 75 27 35 41 72 19 39 28 Silver in rpll .
L ppd .
73 58 37 34 72 56 36 34 69 48 31 14 water .
74 03 30 57 73 13 35 52 69 59 48 14 74 01 34 11 73 05 36 13 69 53 39 29 Silver in carbon r pll .
73 04 38 50 72 59 37 48 68 49 31 44 tetrachloride .
1 ppd .
73 09 31 17 72 38 37 07 68 19 49 44 73 67 34 58 72 47 37 27 68 34 40 30 The azimuths for silver in air and water with yellow light were so close to one another , that the mean was taken in these two cases as the principal azimuth .
The determinations of the reflexion constants of the silver plate made by the two different methods agree fairly well together , hut are not in accordance with subsequent observations made with films of chemically deposited silver , or with M. Jamin 's experiments ( " Course de Physique , " edit .
1866 , vol. iii , p. 693 ) , the principal incidences being somewhat too high , and the principal azimuths having different values for light of different refrangibilities , instead of being nearly the same for all kinds of light .
The plate having been polished with rouge , and having , as has already been mentioned , a reddish tinge , due in all probability to minute particles of the powder having become embedded in its surface , it appeared possible that the difference in the values of the azimuths for the three kinds of light might be due to the rouge ; the plate was therefore dismounted and well polished with putty powder , which was chosen as being nearly white , and then the measurements repeated in the usual manner .
Red light .
Yellow light .
Blue light .
p.i. P.A. P.I. P.A. P.I. P.A. o ?
o / O / o / o / O / f Plate at A. 78 26 44 02 75 0 44 11 71 28 42 56 Silver in air .
^ Plate at B. 74 33 43 41 74 14 43 16 71 39 43 04 ^ Mean value 76 29 43 51 74 37 43 22 .
71 33 43 0 0*1 \#151 ; .
f Plate at A. 76 18 44 27 72 38 44 54 67 15 43 10 Oliver m j Plate at B. 71 37 43 46 71 22 43 20 68 23 43 05 water .
i Mean value 73 57 44 06 72 0 44 07 67 49 43 07 f Plate at A. 75 53 48 69 73 02 44 37 66 27 43 51 Silver in " j Plate at B. 71 53 43 59 71 59 43 46 67 42 43 41 water .
1 Mean value 73 53 43 59 72 30 44 11 67 04 43 46 Sir J. Conroy .
[ Mar. 3 , Silver in r Plate at A. carbon tetra- s Plate at B. chloride .
I. L Mean value Silver in r Plate at A. carbon tetra- \lt ; Plate at B. chloride .
II .
L Mean value Silver in air .
Silver in water .
Silver in carbon tetrachloride .
Bed light .
Yellow light .
Blue light .
p.i. P.A. p.i. P.A. P.I. PA .
o / O / o / O / O / 74 09 44 26 72 05 43 31 65 20 45 25 70 43 44 12 71 12 43 09 67 33 43 21 72 26 44 19 71 38 43 20 66 26 44 23 75 11 44 34 72 13 43 56 67 10 45 44 70 36 44 09 71 10 44 05 67 51 43 37 72 53 44 21 71 41 44 0 67 30 44 40 Mean value .
76 29 43 51 74 37 43 22 71 33 43 0 73 55 44 02 72 15 44 09 67 26 43 26 72 39 44 20 71 39 43 40 66 58 44 31 The values of the principal incidences are nearly the same as when the plate had been cleaned with rouge , but the principal azimuths , in addition to being considerably higher , differ but little for the three kinds of light instead of increasing with the refrangibility of the light .
Films of silver chemically deposited on glass were prepared by Martin 's process ( " Ann. de Chimie , " 4th Series , vol. xv , p. 94 ) .
Glass slips , similar to those used for mounting microscopic objects , were well cleaned , polished with rouge , washed with water , a little nitric acid poured over them , and then again washed and dried .
A rectangular frame ( about 15 centims. by 7 centims. ) , formed of two glass rods , the ends of which were fixed into pieces of wood , was placed in a shallow earthenware dish partly filled with distilled water ; the surfaces of the pieces of glass which were to be silvered were wetted with a mixture of equal parts of alcohol and potassium hydrate solution , and then the glass slips placed with their ends resting on the glass rods , and with their wetted surfaces downwards .
There was just sufficient water in the dish for the lower surfaces of the glass slips to be in contact with it , whilst the upper remained dry .
The silvering solutions were mixed in another dish , care being taken that the depth of the silvering solution was the same as that of the water , and the frame with the slips resting on it , transferred from one dish to the other .
The room in which the silvering was carried on being very cold the action was slow , and till at least five minutes had elapsed there was hardly any deposit ; the slips were removed successively after 8 , 11 , 14 , and 18 minutes , well washed with water , and placed on edge to dry .
The reflexion constants in air , water , and carbon tetrachloride were observed , and then after the area of the silver had been measured , and the slips dried and weighed , the silver was rubbed off with a damp cloth , and the glass again dried and weighed , and from the loss of weight and the area , the thickness of the films calculated .
The 1881.1 Some Experiments on Metallic .
density of the silver being taken as 10*62 , that being the value for silver , finally divided by precipitation , given in " Watts ' Dictionary , " vol. v , p. 277 .
Thickness .
Film I , removed after 8 minutes , 0 *00004035 millim. II , 99 9911 99 0 *00006848 " III , " 14 99 0 *00007772 " IV , 99 " 18 99 0 *00008995 " These values for the thickness of the silver films agree fairly well , being of the same order , with those determined by Quincke ( " Pogg .
Ann. , " vol. cxxix , p. 177 ) , who found that similar films varied from 0*0000039 millim. to 0*000075 millim. Silver films in air .
Red light .
Yellow light .
Blue light .
p.i. P.A. P.I. P.A. P.I. P.A. o ?
0 4 o / O / O / O / f ' Plate at A. 72 55 30 14 68 57 28 24 66 30 31 02 I. \ Plate at B. 68 35 30 02 68 31 26 32 66 29 30 54 l - Mean value 70 45 30 08 68 44 27 28 66 30 30 58 ( ' Plate at A. 73 39 40 16 70 23 39 26 66 56 40 31 II .
i Plate at B. 69 03 39 39 69 36 38 05 66 43 38 39 l - Mean value 71 21 39 57 69 59 38 45 66 49 39 35 ' Plate at A. 74 04 41 25 70 53 42 05 66 40 40 55 III .
\ Plate at B. 69 42 41 10 70 01 41 39 67 17 40 15 \ ' Mean value 71 53 41 17 70 27 41 52 66 58 40 35 1 ' Plate at A. 74 32 41 21 71 0 41 39 66 36 40 35 IV .
j Plate at B. 69 35 41 08 70 37 41 17 67 19 41 0 1 L Mean value 72 03 41 14 70 48 41 28 66 57 40 47 Silver films in water .
Red light .
Yellow light .
Blue light .
p.i. P.A. p.i. P.A. p.i. P.A. o / O / o ?
O / o i O / r Plate at A. 69 53 29 32 67 07 33 25 61 39 33 44 I. \ j Plate at B. 66 42 31 09 67 08 34 40 63 32 34 20 t Mean value 68 17 30 20 67 08 34 01 62 35 34 02 f Plate at A. 70 22 39 34 66 56 39 07 62 30 39 01 II .
j Plate at B. 65 41 39 05 66 47 38 44 62 43 39 14 L Mean value 68 01 39 19 66 51 38 55 62 36 39 07 f Plate at A. 69 37 41 27 66 06 40 59 62 0 39 09 III .
| Plate at B. 65 13 41 11 66 02 40 24 62 32 38 42 L Mean value 67 25 41 19 66 04 40 41 62 16 38 55 f Plate at A. 70 34 41 02 66 41 41 01 62 44 39 12 IY .
'I Plate at B. 65 23 40 26 66 0 41 11 63 12 39 14 Mean value 67 58 40 44 66 20 41 06 62 58 39 13 [ Mar. 3 , Sir J. Conroy .
Silver films in carbon tetrachloride .
r Plate at A. I. -s Plate at B. I Mean value Bed light .
P.I. P.A. 69 20 35 30 6'5 59 65 13 33 24 64 20 67 16 34 27 65 19 Blue light .
*.I .
P.A.* ' / O / O / 34 19 60 55 34 59 34 54 61 31 34 40 34 36 61 13 34 49 Yellow light .
P.I. P.A. f Plate at A. II .
" I Plate at B. *- Mean value f Plate at A. III .
t Plate at B. * .
Mean value 69 38 39 51 66 29 64 30 38 59 65 20 67 04 39 25 65 54 68 21 40 42 64 30 63 20 40 22 64 09 65 50 40 32 64 19 38 49 61 49 37 54 37 25 62 13 38 14 38 07 62 01 38 04 39 12 60 26 37 39 40 14 60 57 38 22 39 43 60 41 38 0 f Plate at A. IV .
" j Plate at B. *- Mean value 40 0 61 05 38 21 39 45 61 46 40 50 39 52 61 25 39 35 Five similar glass slips were silvered in the same manner , being removed from the silvering solution after 7,10,13,16 , and 19 minutes , and the reflexion constants for air determined .
The thickness of the films was not ascertained , as they were required for some other experiments , and were therefore kept .
Silver films in air .
Red fight .
P.I. P.A. Yellow fight .
P.I. P.A. Blue fight .
P.I. P.A. 1 r Plate at A. O ?
73 0 O / 30 25 O / 69 37 O / 31 32 66 48 o / 33 11 V ' 1 Plate at B. 68 27 28 10 70 02 30 52 66 24 32 17 1 1 Mean yalue 70 43 29 17 69 49 31 12 66 36 32 44 1 r Plate at A. 73 18 37 10 69 50 38 07 66 42 38 29 VI .
1 Plate at B. 68 39 37 36 68 47 37 31 66 04 37 35 1 L Mean value 70 58 37 23 69 18 37 49 66 23 38 02 1 r Plate at A. 74 06 41 24 70 49 41 14 67 32 40 0 VII .
J Plate at B. 69 53 40 49 70 10 40 47 66 32 40 04 1 L Mean yalue 72 0 41 06 70 30 41 0 67 02 40 02 1 r Plate at A. 74 43 41 20 71 25 41 54 67 41 41 02 VIII .
J ; Plate at B. 70 13 41 25 70 51 40 24 66 56 40 52 1 L Mean value 72 28 41 22 71 08 41 09 67 18 40 57 1 r Plate at A. 75 21 42 20 71 47 42 24 66 36 42 09 IX .
J 1 Plate at B. 70 42 42 32 70 47 42 19 67 25 41 52 1 L Mean value 73 01 42 26 71 17 42 21 67 0 41 0 1881 .
] Some Experiments on Metallic Reflexion .
The table for the silver film in air shows that with the increase of the thickness of the film the principal incidence increases for all three kinds of light , but the increase is greatest with red and least with blue light .
The principal azimuth also increases with the thickness , but does so more rapidly , the difference between the values for films I and II and between V and VI being very considerable , whilst the values for films II , III , and IY and VII , VIII , and IX lie close together .
With films II , Y , and VI the azimuths increase from red to blue , whilst with the thicker films they decrease ; film I , however , furnishes an exception , for , although very thin , the azimuths decrease from red to blue .
In this respect the thicker films behave like the silver plate cleaned with putty powder ( p. 496 ) , and the actual values of the azimuths in both cases are about the same ; the principal incidences , however , are much lower with the films than with the plate .
The difference in the values of the principal incidences may possibly be due to the effect produced on the plate by the pressure necessary to polish it .
When the films were , in water and carbon tetrachloride , the principal incidences decreased to a slight extent with the increased thickness , but the variation was small in amount .
The azimuths in the case of film I in water increased from red to blue , but in carbon tetrachloride , and with the other films in both liquids , the azimuths decreased from red to blue .
When the films were in contact with the liquids* the principal incidences were lower , except with film I , than in air , and the principal azimuths were also lower , instead of being higher , as had been the case with the gold and silver plates .
Quincke ( " Pogg .
Ann. , " vol. cxxix , p. 177 ) made a large number of observations with silver films and red light , and found that the values of the principal incidences increased with the thickness* and tended to a constant value , and also that as soon as the thickness of the metal fexceeded 0*00002 millim. there was but little further change in the value of the principal incidence .
In the short paper which appeared in the " Proc ; Roy .
Soc. , " vol. 28 , p. 242 , it is stated that the numbers obtained by multiplying the tangents of the angles of principal incidence for red light and the gold plate in water and carbon bisulphide , by the refractive indices of these media , were somewhat higher than the tangents of the angles of principal incidence for air .
This is also the case for the determinations made with the gold plate and yellow and blue light* and for those made with gold leaf , the silver plate , and the silver films with all three kinds of light .
Thus for the silver plate cleaned with putty powder the values are as follow :\#151 ; o ' ' Sir J. Conroy .
[ Mar. 8 , Bed light .
Yellow light .
Blue light .
o / JO ?
O / Principal incidence in air observed 76 Id9 74 37 71 33 Calculated from observations made in water ... ... ..----------- -------- 77 47 76 30 72 46 Calculated from observations made in carbon tetrachloride ... ... ... , .
77 54 77 13 73 52 In the case of the silver films the difference between the observed and calculated values is greatest with the thinner films ( amounting to nearly 4 ' for the thinnest , with red light ) and diminishes as the films increase in thickness .
The values for the refractive index for gold were also calculated by the formula given by Lindquist ( " Pogg .
Ann. , " vol. clii , p. 405 ) , = tan2A(1\#151 ; sin3A .
sin220 ) , and the values for the relative indices for water and carbon bisulphide multiplied by the refractive indices of these media , but the numbers thus obtained did not agree with those deduced from observations of the principal incidence and azimuth in air .
These experiments appear to show , first , that with glass in contact with media other than air , the tangent of the angle of polarisation is equal to the relative refractive index of the media .
Second , that the optical constants of a polished metallic plate depend , to a certain extent , on the substance with which it has been polished , and that this surface condition is a fairly permanent one , not being destroyed by contact with liquids or by a considerable amount of rubbing with a clean chamois leather , Third , that when the plate is in contact with media other than air , the principal indices qre lower , and the principal azimuths higher than in air ; but that there is no simple relationship discoverable between the change in the valq.es of the constants and the indices of the media .
Fourth , that with metallic films the principal incidence and the principal azimuth increase with the thickness of the film , and that therefore more than one layer of molecules is concerned in the act of reflexion .
The fact that with the increase Qf the thickness of the films the principal azimuth increases appears to show that the surface layers reflect light polarised in the plane of incidence more abundantly than light polarised perpendicularly to that plane , and that as the number of reflecting layers increases , the amoupt of light polarised perpendicularly to the plane pf incidence which is reflected increases also ; or , in other words , the light polarised perpendicularly to the plane of incidence penetrates to a greater depth than that polarised in the plane of incidence .

