Phys. Educ 21 1986.

Prlnted In Great Brltaln

l eachlng
air wedge
interference
M

H S Fricker

SeveralA-levelsyllabusesincludetheexperiment
shown in figure 1. An air wedge is formed between l I
two glass plates G inclined at a small angle, a , by G
means of a thin spacer, X . Light from an extended Figure 1
monochromatic source, S. such as a sodium lamp,is
reflected onto the wedge by the glass plate R, and
the light reflected up from the wedge is observed
through a microscope M. Partof the light is reflected
by each surface of the wedge, and the two reflec- believe that, forall but the weakest students, details
tions, which are coherent, have a geometrical path are morelikely to recede into the background if they
difference of 2d, where d is the wedge thickness. have first been clearly explained.
One therefore observes interferenceof a type which Figure 2c, where the law of reflection is obeyed
depends on d . If the surfaces bounding the wedge and the two reflected rays are not parallel, is more
areopticallyflat, d varieslinearly with distance realistic. But even here there are difficulties if one
across the wedge, which gives evenly spaced fringes attempts to discuss the localisation of the fringes.
of separation A/2a, where 1, is the wavelength of the Clearly,oneshouldobserveinterference if the
light. microscope is focused at P, where the rays intersect.
The essential physics is thus very simple. How- But suppose one looks elsewhere? In fact, the frin-
ever, problems arise when one tries to explain more ges are not sharply localised, but may be seen for
fully what is happening. A glance at some popular some distance above the wedge, depending on the
A-leveltextbooksrevealsdiagramssuch as those angle of the wedge and how it is viewed.
shown in figure 2 . (In eachcaserefractionatthe Moreadvancedtextbookssolvetheseproblems
aidglassboundary is ignored, as it will be in this by abandoningthe‘ray’approach,and discussing
article.) To a teacher, it will be clear that figures 2a sets of intersectingwavefronts reflected fromthe
andbarenotattemptingtoshowwhatactually two faces of the wedge. Although this approach is
happens to the incident ray, but any student who seldom used at school level, it is within the scope of
remembersthelaws of reflection is likely tobe good ‘S’ level students, particularly if it is related to
confused. Presumably the intention behind such dia- Young’s fringes, with which they are very familiar.
grams is that, if details are avoided, the essential Such a treatmentis outlined in the last section of this
physics will standoutmoreclearly.However, I article. However, what is needed for everyday pur-
poses is to make the ‘ray’ treatment work properly,
and this is done in the next section.
Hugh Fricker is a physics teacherat Bradford Gram- No originality is claimedfortheapproaches
mar School with additional responsibility f o r higher adoptedhere.However,theydonotseemtobe
education advice and applications. H e both gradu- common in textbooks, and they may of be interest to
ated and obtained his PhD from Cambridge Univer- other teachers.
sity-theoretical
physics-and was IC1Research
Fellow at ManchesterUniversity f r o m 1970-2. His Improving the ray approach
interests includequantumchemistryandpolymer The trouble with the diagrams in figure 2 is that, by
theory. asking what happens to a single incident ray, they
0031-9120/86/010048+05$2.25 01986 The Institute of Physlcs
Figure 2

approach the problem from the wrong end. In prac-

tice, one can focus the microscope on any point one
wishes, and all rays which pass through that point
into the eye will be superposed on the retina. We
should therefore beginby specifying the observation
point, and then ask what light arrives there.
Consider first a point source S (figure 3). With the
apparatus shown infigure 1, S will bethevirtual
image formed by plate R of a point source in the Figure 3
lamp. If the microscope is focused on an arbitrarily
chosen piont P, then the two rays which superpose
are those shown in figure 3. Each obeys the law of
The path difference=BY+YA
reflection, but they are not, in general, part of the
=XY [COS ~ + C O S ( ~ Y + ~ ) I
same incident ray. Since S is a point source the rays
= ~ X cos(y+e)
Y cos
arecoherent,andtheresult of superpositionde-
pends on their path difference.
=2d COS y ,
To find thepathdifference,assumethat S is a where y is theangle of incidenceonthebottom
distant source, so that SX and SY are approximately surface,and d is thewedgethicknessat X. For
parallel. Then P X and PY are also nearly parallel, near-normalincidence(cos y = l ) , thepathdiffer-
because a is very small. The important part of figure ence is 2d, as is clear from figure 4b.
3 then looks like figure 4a. Having discussed the details, one can summarise

From S

Figure 4

49
 through P. However if themicroscope is focused
higherup(figure 6b), theeffectiveregion of the
wedge S T is enlarged, and so is the variation of d
acrossit. If thisvariation is large enough, the in-
' " p terferenceeffectsfromdifferentparts of ST will
average out, and no fringes will be seen.
Wedge This leads to a rough criterion for fringevisibility.
P The change in d between S and T is aST, giving a
Figure 5 change in path difference of 2aST. Fringes will not
be observed if

U
2aST>A
or
ST>1/2a
i.e.

d
S T 2 t h e fringe spacing.
If P is a distance z above the wedge, then
objective
Microscope
ST+z,
\"p
\ I so fringes will not be observed if
z>1/2a/?=(fringe spacing /p).
This result is easily tested experimentally. I used
the arrangement shown in figure 1, with a sodium
lamp as the extended source S. The travellingmicro-
scope had an objective of diameter 6mm, focused
on a level 47mm below, givingB=6/47=0.13 radian.
Figure 6 Observing fringes of separation 1mm, I found that
the microscope could be raised about lOmm above
the result with reference to figure 5 , which takes due the position where they were clearest before they
account of the smallness of the wedge angle. Light disappeared. This agrees quitewell with the value of
will arrive at P from 'point' Q on the wedge if there (fringe separation//?) which was UO.13-8 mm. With
is a point source S in the appropriate direction. The a wedge of smaller angle, or a viewing arrangement
light will consist of apair of coherentrays with which ensuredmorenearlynormalincidence,the
geometrical path difference 2d cos y , where d is the fringes would have been less localised.
wedge thickness and y the angle of incidence at Q. It is not, of course, suggested that the detailed
Fringe localisation is now readily explained. With treatment given here should be expected of an A-
an extended source, light strikes all parts of the level candidate.MostA-levelexaminerswould,
wedge with a range of angles of incidence. There- rightly, be satisfied with the account given in the first
fore pairs of coherent rays arrive at any observation paragraph of this article, together with a mention of
point fromall parts of the wedge. Howeverit is clear theneedfornear-normalincidence,thephase
from figure 6a thatnot all such rays enterthe change at the bottom reflection, and the calculation
microscope. The microscope objective subtends an of the fringe spacing. However I have found that,in
angle /? at the point P on which the microscope is teaching the topic, a presentation basedon figures 3,
focused, and /3 defines the region of the wedge, ST, 4b, 5 and 6 is accessible to students, and improves
which contributes to the image. Only light reflected their understanding of the phenomena.
from within STcanenterthemicroscopeand so
The air wedge as a case of Young's fringes
contributetotheimage of P. For all suchrays,
y@/2, and B is usually small enough for the 'normal An interesting alternative, for able students, is to
incidence'approximationtobevalid.However d takeadvantage of theirfamiliarity with Young's
varies across ST.If S T is small enough, this variation experiment. In figure 7a, S is again a point source, a
is negligible (i.e. <<A), and the path difference is distance D from the wedge apex 0. Each reflecting
essentially the same for all coherent pairs of rays surface produces a virtual image (I, and I>). Simple
passing through P into the microscope. In that case, geometry shows that S, I, and I2 all lie on a circle of
fringes will be observed in a horizontal plane passing radius D centred at 0, and that I] and I2 subtend an

50
 a I2 \ I

\ ' I,
b
I2
*Y\
Figure 7 a and b

angle of 2a at 0. Their separation a=2aD. Since I, from the wedge, the fringe systems get increasingly
and I2 are coherent sources, Young's fringes will be out of step with each other.
produced in the regionabove
the
wedge.In
figure 8 the full and
broken lines
represent
the
Assuming, as before, that D is large ( S a distant loci of bright fringes for the two extreme angles of
source), and remembering that 2a is small, we can
usetheusualformula ADia forthefringespacing, Figure
and obtain a spacingof AD/2Da=A/2a in the vicinity
of the wedge. D has cancelled out because, as long
as it is large, we simply have two setsof plane waves
intersecting at an angle of 2a.
Figure 7b shows the result, taking account of the
factthat a is verysmall.Theorientation of the
fringe loci depends on the angleof incidence y of the
light. The fringe spacing along the wedge is L/2acosy,
which reduces to 1/2a for near-normal incidence.
With an extended light source, each direction of
rise
incident
but
system,
gives
fringe
light
own its to I

in everycasetheline of zerogeometricalpath
difference passes through 0. If, as before, the view-
ing arrangements restrict y to small values (</3/2),
stems
fringe
all the o Wedge
spacing at the wedge, and so the fringes coincide
c _ _

- h
there and are visible. However, as one moves up 2a

51
 Phys Educ 21 1986 Prlnted ~nGreat Britain

incidence, i p / 2 . At any point in the shaded region

there will beabrightfringeduetooneormore
angles of incidence in between.Clearlyonecannot HAVE YOU READ THIS?
expect to observe interference at a distance above
the wedge greater than roughly z , where 112az=/3,
i.e. z=A/2ab, which is the same condition as derived
in the previous section. Abstracts written by J A Champion. E Deeson and
W H Jarvis
Aside
It is an interesting exercise to consider interference Ring interferometrictests of gravity
in a paral'el-sided from the Young's fringes This article considers ring interferometersof various types
Point of view. Again one obtains a Young's fringe and the wav in which thev
, mav , in exoeriments for
be used
system fom each point on an extended source, only the measurement of quantum and gravitational effects.
time this infinity. JA C
The fringes are therefore best viewed by a relaxed G EStedman ContemPorarY Physics 1985 26 PP311-32
eye,as is usually assumed.Theconditionsunder
which they may be seen by an accommodated eye
Radio amateurs' exam attacked
can be worked out by methods similar to those used
This examination, runby the City and Guildsof London
above. Institute, has been strongly criticisedby the head of physics
at Harrogate College-reflecting comments made over
Acknowledgment severalyears in Electronics and Wireless World.WHJ
I am grateful to Dr I R Gomersall for some helpful P Hawker Elecrronicsand Wireless World 1985 October
suggestions,
andto my sixth-formpupils
whose PI1
scepticism prompted this work.
Logarithmic and exponential circuits
In part 2 of an article on 'the long tailed pair' class
of
electronic circuits, applications include analogue
multipliersand
dividers
andfast
logic
gates. WHJ
J Lidgey Electronic and Wireless World 1985 October
~ ~ 2 7 1- 3

Future of British engineering

In this. the last articleof a series. the author looks at the
hidden strengthsof British engineering, where they are
Queries in physics being suppressed and where they could be broughtto the
surface. The reusable unmanned horizontal take-off and
Q601 (from Q I P 66) From a fairly recent land satellite launcher(HOTOL) suggested by British
examination: a spaceship has an external tank of Aerospace is taken by way of example. WHJ
volume V . which contains oxygen gas at pressurep , , . R E Young Electronics and Wireless World 1985 October
pp45-8,64
A meteorite punctures the tank producing smalla
hole area A in the side. Obtain an expression for the
pressure in the tank after timet in terms of P o , A , V One box for data from space
and t . ASTRID is a complete satellite receiver and decoder for
use with the University of Surrey satellites and a standard
Q602 In Deschanel's NaturalPhilosophy translated tape recorder. Suitable softwareis available from AMSAT
and modifiedby J D Everett (1901), there is a UK. ASTRID costsE149 including aerial, feeder, power
supply unit, test tape, manual, and connecting leads. WHJ
mention of Mr Wimshurst later (1892) showing a Electronics and Wireless World 1985 October p82
version of his machine which produced alternating
charges. Can anyone throw further light on this?
A new 'cure' for poor contacts
The above items were selected from Q I P , a thrice- A liquidcalledTWEEKworksby filling in surface
yearly broadsheet. It is available on subscription at a imperfections and improving metal-to-metal contact-at
of rate f2.75 ( f 4 overseas
airmail,
surface
f3.50 ml. 7€15 for WHJ
mail) from the Editor, Mr W H Jarvis,Salewheel Electronics and Wireless World 1985 October p86
House, Ribchester, Preston PR3 3XU.All
correspondenceconcerningthisfeatureshouldbe European electronics 'dead in 10 years'
addressed
to Mr Jarvis. Since 1978 Europe's
share of electronics
worldequipment

52