473 1883 .
] Internal Reflexions the Eye .
This last formula is derived from ( 246 ) .
I observe that the fundamental idea by which these integrals are obtained is given by the equation fr(A0 + Ax cos 6+. .
Are cosr\lt ; 9 ) ( Br cos rd + Br+1 cos +1 ) + .
, .
) =-ArB^ .
Jo This method may be much extended .
II .
" Internal Reflexions in the Eye .
" By H. Frank New ALL , B.A. Communicated by Dr. M. Foster , Sec. R.S. Received January 18 , 1883 .
1 The observation I have to record first came under my notice three or four years ago .
Often when working at night by the light of a candle , in a room otherwise dark , my attention was caug y a very faint light some way out of the line of direct vision .
This seemed to defy nearer inspection ; for the instant I turned my eyes towards it , it was gone , thus showing that there was no objective cause , but that the light was due to some internal reflexion in the eye .
Later , however , I found that by keeping the eye fixed and moving the candle , the faint light could be observed at leisure though , as far as I could then make out , never in the line of direct vision .
( See however below , S. 32 .
) r , 2 The best conditions soon became apparent , and I have app ie two methods in later investigations : ( i ) One in which the eye is fixed on a spot on a dark or uniform ground whilst the candle is moved to and fro out of the line of direct vision , ( u ) One m wine i the candle or source of light is kept fixed , whilst the eye follows the regular movement of some point , such as the end of a pencil moved ^3 The first of these methods showed that the ghost , as I may call the faint light , moved roughly speaking in a line drawn through the point of clearest vision and the candle , in direction opposed to that ot the candle 's motion with respect to the point of clearest vision , and with a velocity equal to that of the candle .
4 .
The second method showed what is practically the same thing , namely , that the line of movement of the ghost was just as described ; the direction the same as that of the point of clearest vision over the field in front ; the velocity apparently about double that of the pom 'f 5^'both methods the ghost merged into the candle close to the point of direct vision , and in other positions seemed about equally removed from that point with the candle .
474 Mr. H. F. Newali .
[ Jan. 25 , 6 .
Both methods lead , as I have said , to practically the same results .
The ghost then is independent of the position of the eye-ball in the socket , and hence must be produced by reflexions internal to the eye , and not brought about by its external surroundings .
7 .
Nearer inspection of the ghost itself showed it to be an inverted image of the candle , about equal in size , very faint , and of a slightly dull bluish tint .
8 .
Before saying more I will describe what has proved the easiest way of finding the ghost .
Stand opposite a uniformly dark wall in a darkened room .
Direct the eyes to any point in front ( e.g. , a mark on the wall , a pin in a curtain ) , and keeping the eyes fixed and being ready to perceive any appearance out of the line of direct vision without moving the eyes towards it , hold up a candle at arm 's length , and move it to and fro over about two inches , on a level with the point fixed , and a little to the *right or the left of it .
The ghost ( or rather ghosts , if both eyes are used ) may be seen moving with a motion opposite to that of the candle on the other side of the point of direct vision ; 9 .
The best way that X have found for observing the image , is to set a candle on the -table about a foot from the eye , and place close in .front of it a dark-coloured board two or more feet in length , and of breadth jusc enough to allow the flame of the candle to be seen above when the board is set up breadthwise on its edge .
Now let the eye be fixed on any object which can be moved along the top edge of the board .
When this point moves the eye follows it with a regular steady motion , which I have failed to get by any other method , such as the seemingly more simple one of letting the gaze move along the edge of the board , unhelped by a moving point .
The board in shadow makes the best background for faint images , such as the " ghost , " which will be found in the position above described .
10 .
In either of the methods described in the two last paragraphs , the motion of the candle or of the moving point should be slow , and to and fro , over a short range ; but it should be continued till the -image is found , for if the image rests , it makes less and less impression on the retina .
11 .
Since the ghost depends on the state of the accommodation of the eye , for distinctness and even visibility ( see below , SS 15-20 ) , it follows that people with different -sight may have to look for the ghost in different ways .
A short-sighted person for instance , will probably not be able to see it in the middle of the field at all , and so may fail to recognise the phenomenon , unless indeed his short sight is due to too grea/ t convexity of the cornea only .
12 .
Beyond a mere cursory investigation of the phenomenon , I did nothing at the subject at the time ; but later , having need to consult Mr. B. Liebreich , I asked him if he knew of the phenomenon .
He 1883- ] Internal Reflexions the Eye .
475 was at first incredulous , but being convinced when I showed him how to find the ghost , he attributed it immediately to reflexion from the fundus , and compared it with other retinal images .
He wrote to me after a day or two , saying he had found a complete solution of the problem , and would like , if I had no objection , to publish an account of some experiments he had made in connexion with the subject .
Later he told me he had given up the idea , and I have not seen or heard of any published account .
I had , however , made observations which led me to question his theory , and I determined to make further experiments , measurements , and calculation , to test my own view that the reflexion was from the surface of the lens , and not from the fundus alone .
I was at one time inclined to consider that the ghost was connected with " Sanson 's images , " but was driven from that idea by considerations as to the amount of light that could be reflected back into the eye ; this could only be extremely small .
( See also below , S 30 ' ) .
, .
, 13 .
Many reasons hereinafter related make me believe that the ghost is produced in the following way:\#151 ; An image of the candle flame is thrown on the retina by the crystalline lens ; this image , which I will call the first retinal image , may now be considered as a source of light from which rays proceed outwards from the retina , and are refracted by the crystalline lens and cornea to a focus again outside the eye in such a way that an image of the retinal image would be formed in the place occupied by the candle .
Part of these rays , however , are refleeted at the various surfaces bounding the different media of the eye , and those reflected by the anterior surface of the lens , or what one should more correctly here regard as the posterior surface of the aqueous humor , are brought to a focus somewhere between the lens and the retina .
The rays from this focus have not diverged much before they fall on the retina , and there form a blurred inverted image of the first retinal image .
This second image is " referred " outwards , and looks as if produced by a faint source of light outside the eye , and having a definite position in 14 .
I shall speak in what follows of this imagined source of light as the " ghost , " to distinguish it from the actual image of the retina which gives rise to the sensation , and which I shall call the second retinal image .
The ghost is the analogue of the candle , each being the mental image corresponding to the physical image on the retina .
The name ghost will suggest to the reader similar phenomena in the telescope , which , however , are produced in a different way .
15 .
The ghost is affected by accommodation , so that with the candle about one foot from the eye , and with the eye accommodated for a point six inches from it , the ghost does not become visible till it is 11 degrees from the centre of the field .
These numbers refer to the Mr. H. F. Newall .
[ Jan. 25 , case of my own eyes , and will , of course , vary with the observer .
I may add here that I am what would be called loug-sighted , though having a very large range of accommodation .
16 .
Again , when the eye is focussed for a distant point , the ghost is visible right up to the middle of the field , but invisible when removed more than 17 degrees from the centre of the field .
17 .
Again , if the eye and candle are fixed whilst the eye is focussed for different distances , the ghost seems also to pass through regular gradations of distinctness .
( See also below , S 35 .
) 18 .
Now , in the last case , it is clear that the candle can only give a clear image on the retina in one particular state of accommodation , so that generally speaking there will be a blurred first retinal image , the reflexion of which is the second image .
One would then expect that the ghost would be more likely to be distinct when the candle is at the same distance from the eye as the point focussed .
19 .
This I have found from observation to be the case when the eye is focussed for points not nearer than two feet .
When the eye is focussed for a point six inches from it , the ghost is invisible when the candle is further than one foot from the eye , and in no position of the candle does it become very distinct , but is most so when the candle is close to the eye .
With so close a point of regard the radius of curvature of the anterior surface of the lens would be much diminished , and hence also its focal length when regarded as a concave mirror , so that the rays proceeding from the focus have diverged considerably more before reaching the retina to give rise to the second image than in the case when the eye is focussed for long distances .
20 .
Here it may be noticed that with the point of regard at arm 's length , the ghost increases much in size if the candle is approached to the eye .
And also if , with the eye focussed for a near point and the candle about a foot from it , the point of regard be moved gradually away from the -candle across the field , then the ghost , which is not to be seen in the centre of the field , comes into sight and grows more distinct as it becomes more moved from the centre up to a certain point , beyond which it again becomes indistinct and finally disappears .
The point of distinctness varies also with the state of accommodation .
21 .
From considerations referred to in S 18 , the question arises : Are we to regard the retina in this matter as a concave mirror , or merely as a screen ?
An example will help to make the importance of this-clear .
Suppose the source of light is the sun ; then if the eye is accommodated to see it distinctly , the focus of rays from the sun would fall on the retina , which in this case would act simply as a screen , receiving a distinct image of the sun , whence rays would proceed outwards again as in S 13 .
If , however , the eye was accommodated for near distances , then rays from the sun would be brought Internal Reflexions in the Eye .
1883 .
] to a focus at a point in front of the retina , and would diverge again from the point and form a blurred image on the retina .
Now , will this blurred image be the source of light from which rays diverge , or will the retina act as a concave mirror , so that rays from the blur imacre will be reflected from it so as possibly to converge again to a focus which we may regard as the source of light ; or , to put the matter more briefly , is the reflexion from the retina regular or irregu .
22 .
I have sought for an answer to these questions m allowing e first retinal image to fall on the depression of the fovea centralis , -where the radins of curvature is mueh smaller .
In this case there is no perceptible alteration in the appearance of the ghost .
This pom s o the idea that the retina acts simply as a screen and not as a concave mirror , and that we may regard the light coming from the first retinal image ( whether this be clear or blurred ) as diverging from \#171 ; \lt ; " \#187 ; .
23 The facts recounted in SS 19 , 20 lead me to believe that the second image is the reflexion of the first .
One more weighty argument may here be added .
If the positions of the candle and ghost be noted hi any one position of the eye , and the candle be moved into the apparent position of the ghost , then the ghost will be observed in thePplace at first occupied by the candle : that is , the first and second images of the candle on the retina are conjugate foci with respect to some reflecting surface in front of the retina .
24 The ghost is an inverted image of the candle ; there or , physical cause , the second retinal image , must be erect , and .
hence an inverted image of the first retinal image of the candle .
The reflecting surface is therefore concave .
The possible reflecting surfaces are the posterior surface of the cornea and the anterior surface of the lens .
25 A priori the cornea seems least probable for two reasons .
1st .
Light from the retina must pass through the lens twice before returning to the retina , and calculation shows that the rays wou proceed finally , after leaving the lens , towards the retina , as if coming from a virtual focus within the lens , 2'39 millims m front of the posterior surface .
2nd .
The light proceeding outwards from retina and reflected back by the cornea would , in cases when the rs image was far from the centre of the field , be reflected on to the anterior surface of the iris , and so would not reach the retina again 26 .
I have attempted to find whether there was any illumination of the iris that would correspond to this ; but two observers have failed to find light on the iris , which could not be traced to other causes .
The circumstances in the case were as favourable as possible for I made use of the sun as prime source of light , so that the ig \#153 ; very bright and the pupil therefore much contracted .
Again , light is intercepted by the iris in its return to the retina , the ghost would disappear part by part , the edge of the iris mtercep mg more 478 Mr. H. F. NeWall .
TJan .
25 , and more of the figure of the ghost : but I find , on the contrary , that it dies away gradually , getting " out of focus , " as it were , as it passes from the field .
27 .
On the other hand , the anterior surface of the lens seems most likely to give the solution of the problem .
Calculation shows that the light proceeding outwards from the first image on the retina would be , after refraction , at the posterior surface of the lens , reflexion at the anterior surface , and second refraction at the posterior surface , focussed at a point in the vitreous humor .
This point , when the eye is focussed for long distances , will be about 24 millims. from the posterior surface of the lens : and when the eye is focussed for a near point , it will be about 1 millim. from that surface .
The distance of these points from the retina certainly forms a very great difficulty in the explanation .
28 .
As to the brightness or rather faintness of the image , it seems , at first sight , improbable that the surface between the lens and aqueous humor should be capable of reflecting enough of the light coming from the first retinal image , to excite the retina .
But the ease is comparable with the appearance of " Sanson 's Images , " and the brightness of the ghost , as compared with the candle , might well be described as about the same as the two faint images as compared with the bright one in Sanson 's phenomenon .
We might expect to find some change in the brightness of the ghost when the first retinal image falls on the more opaque part of the retina at the entry of the optic nerve .
But I have failed to get certain results .
I have allowed the image of the candle to fall on the blind spot : the ghost does not lose in brightness , though the whole field of vision seemed to become brighter , in consequence , no doubt , of the greater diffusion of light through the eyeball from the image on that part .
That the ghost does not seem to lose in brightness as the ground becomes more illuminated , may , perhaps , be taken as a sign of increased brightness^ but nothing definite can be gained from this point towards the explanation .
29 .
The fact that the ghost is visible in the centre of the field ( see below , S 32 ) shows that the reflexion cannot be from the fundus alone , that is from the fundus direct , on to another part of the retina , as , I understood Mr. Liebreich to have imagined .
This is clear , moreover , from the fact that if the fundus were the only reflector in the case , the ghost would move in the same direction as the candle .
This would also be the case if the first reflexion outwards were from a surface in front of the retina instead of the retina itself .
30 .
If , for instance , the reflexions were entirely within the lens ( as I at first thought possible ) or within the aqueous humor , then the ghost and candle ought to move in the same direction , the former with greater angular velocity .
In fact , the case would be comparable 1883 .
] Internal Reflexions in the Eye .
with the reflexion sometimes to be seen on the scale of a reflecting galvanometer .
The light reflected from the mirror is partially re-fleeted at the surfaces of the worked glass which closes the mirror tube : and that part reflected within the glass at the surface remote from the mirror is again reflected within the glass at the surface next the mirror , so that a second and very faint image of the slit is sometimes formed on the scale .
This second image moves in the same direction as the ordinary bright one , but with greater angular 31 That the ghost is not due to reflexions within the lens is also made very probable from calculations which show that in any state of accommodation the focus would be virtual and situated in the aqueous humor That it is not due to reflexions within the aqueous humor is made more probable from the fact that I have discovered a second ghost , which is almost certainly due to this cause ; it moves m the same direction as the candle , but the faintness and indistinctness of this image is too great to allow of any accurate measurements .
It is very blurred , and I have never been able to bring it even approximately to a focus on the retina . .
- .
32 The wish to see the ghost , if possible , m the hoe of d , rect vision has led me to further very interestmg observations .
The method was suggested by an accidental sight of the ghost in^ bright sunlight .
I was standing on the terrace at Heidelberg , and looking towards the sunset ; in front of me on the other side of the ravine was the Castle , standing out dark against the bright sky beyond .
As my eye wandered slowly over the outline of the buildings my attention was caught by the ghost flitting down below in the shadow .
Here were all the necessary conditions , a bright light and a dark screen .
There was no difficulty in moving the eye till the ghost fe 1 exactly on the point of direct vision .
But , strange to say , the sun was still some way out of the line of direct vision ; that is , instead of seeing the sun and its ghost equidistant from the centre of the field , and finally , as the eye moved further , merging into one at the centre of the field , the ghost had got to the centre first , and had moved on to the same side of the centre as the sun before they coincided .
33 .
Moreover the eccentricity , as I may call it , is different for my two eyes both in amount and direction .
I can best describe the difference by saying that with my right eye , when the ghost was m the line of clearest vision , the sun was still to the right and above ; and when the line through my eye was horizontal , the line through the sun 's position and the ghost , or the point of clearest vision , was inclined at an angle of 36 degrees to the horizon , the angular distance of the sun from the ghost being 8 ' 32 ' .
With my left eye when the ghost was in the line of clearest vision , the sun was to the left on the same level , and at an angular distance from the ghost of 6 25 .
480 Capt. Abney and Lt.-Col .
Festing .
[ Jan. 25 , 34 .
This points to something anomalous in the centering of the eye surfaces .
The amount and direction of the eccentricity in my eyes as shown by these means are in somewhat remarkable correspondence with those obtained by the entirely different method of Helmholtz ( " Phys. Opt .
, " pp. 86 , 87 ) .
I may add below the angles for my eyes corresponding to numbers given by Helmholtz , loc. cit.:_ Light came from Nasal side .
Temporal side .
Right eye ... , ... ... ... . .
9 ' 27 ' ... .
3 ' 49 ' Left eye---- ... ... ... ... .
7 ' 36 ' ... .
7 ' 36 ' 35 .
One more observation of some interest I may record .
With the sun as source of light and with a fixed direction of vision ( I have generally taken that direction in which the ghost comes into the line of direct vision , so as to avoid the constant tendency of the eye to turn towards any object of interest in the field ) , the ghost passes through a series of forms as the state of accommodation alters .
There are three marked forms , a horizontal bar with a near focus ; a circle for a slightly more remote focus ; and a vertical bar for long distance .
There may be signs of a double astigmatism ; but as they are the only signs that I have had of such defect in my eyesight , I have rejected this explanation in favour of the equally simple one , that we have here the focal lines and circle of least confusion such as are observed in oblique reflexion by a concave mirror .
The explanation suggested in this paper is attended with difficulties , the greatest of which is perhaps that referred to in S 27 .
It is , however , the one which on the whole explains best the greatest number of the facts to be accounted for .
The calculations made on various points mentioned above are based on measurements given by Helmholtz for his schematic eye .
It seems likely that individual variations from the normal eye should be such as to determine the visibility of the ghost ; but I have found as yet no obvious relation in \amp ; bout fifteen cases between the power to see the ghost and the kind of sight , as defined by the ordinary terms , long and short sightedness .
Two cases in which I have , notably failed to show the ghost , are opposed in this respect , one having short sight , the other normal ; whilst of those to whom the ghost is visible there are as many short sighted as long sighted .
III .
" Note on the Absorption Spectrum of Iodine in Solution in Carbon Disulphide .
" By Captain Abney , R.E. , F.R.S. , and Lieut.-Colonel Festing , R.E. Received January 18 , 1883 .
In 1876 , Sir John Conroy ( " Proc. Roy .
Soc. , " vol. 25 , p. 46 ) , described the absorption spectrum of a solution of iodine in carbon

