332 Chapter 7 Electrodynamics

(a) What is the current at any subsequent time t?

(b) What is the total energy delivered to the resistor?

(c) Show that this is equal to the energy originally stored in the inductor.

Problem 7.32 Two tiny wire loops, with areas a1 and a2 , are situated a displacement
r apart (Fig. 7.42).
a2
a1

FIGURE 7.42

(a) Find their mutual inductance. [Hint: Treat them as magnetic dipoles, and use
Eq. 5.88.] Is your formula consistent with Eq. 7.24?

(b) Suppose a current I1 is flowing in loop 1, and we propose to turn on a current

I2 in loop 2. How much work must be done, against the mutually induced emf,
to keep the current I1 flowing in loop 1? In light of this result, comment on
Eq. 6.35.

Problem 7.33 An infinite cylinder of radius R carries a uniform surface charge σ .

We propose to set it spinning about its axis, at a final angular velocity ω f . How much
work will this take, per unit length? Do it two ways, and compare your answers:

(a) Find the magnetic field and the induced electric field (in the quasistatic approx-
imation), inside and outside the cylinder, in terms of ω, ω̇, and s (the distance
from the axis). Calculate the torque
 you must exert, and from that obtain the
work done per unit length (W = N dφ).
(b) Use Eq. 7.35 to determine the energy stored in the resulting magnetic field.

7.3 MAXWELL’S EQUATIONS

7.3.1 Electrodynamics Before Maxwell

So far, we have encountered the following laws, specifying the divergence and
curl of electric and magnetic fields:
1
(i) ∇ · E = ρ (Gauss’s law),
0
(ii) ∇ · B = 0 (no name),
∂B
(iii) ∇ × E = − (Faraday’s law),
∂t
(iv) ∇ × B = μ0 J (Ampère’s law).
7.3 Maxwell’s Equations 333

These equations represent the state of electromagnetic theory in the mid-nineteenth

century, when Maxwell began his work. They were not written in so compact a
form, in those days, but their physical content was familiar. Now, it happens that
there is a fatal inconsistency in these formulas. It has to do with the old rule that
divergence of curl is always zero. If you apply the divergence to number (iii),
everything works out:
 
∂B ∂
∇ · (∇ × E) = ∇ · − = − (∇ · B).
∂t ∂t

The left side is zero because divergence of curl is zero; the right side is zero by
virtue of equation (ii). But when you do the same thing to number (iv), you get
into trouble:

∇ · (∇ × B) = μ0 (∇ · J); (7.36)

the left side must be zero, but the right side, in general, is not. For steady currents,
the divergence of J is zero, but when we go beyond magnetostatics Ampère’s law
cannot be right.
There’s another way to see that Ampère’s law is bound to fail for nonsteady
currents. Suppose we’re in the process of charging up a capacitor (Fig. 7.43). In
integral form, Ampère’s law reads

B · dl = μ0 Ienc .

I want to apply it to the Amperian loop shown in the diagram. How do I deter-
mine Ienc ? Well, it’s the total current passing through the loop, or, more precisely,
the current piercing a surface that has the loop for its boundary. In this case, the
simplest surface lies in the plane of the loop—the wire punctures this surface, so
Ienc = I . Fine—but what if I draw instead the balloon-shaped surface in Fig. 7.43?
No current passes through this surface, and I conclude that Ienc = 0! We never
had this problem in magnetostatics because the conflict arises only when charge

Amperian loop

Capacitor I

Battery

FIGURE 7.43
334 Chapter 7 Electrodynamics

is piling up somewhere (in this case, on the capacitor plates). But for nonsteady
currents (such as this one) “the current enclosed by the loop” is an ill-defined
notion; it depends entirely on what surface you use. (If this seems pedantic to
you—“obviously one should use the plane surface”—remember that the Ampe-
rian loop could be some contorted shape that doesn’t even lie in a plane.)
Of course, we had no right to expect Ampère’s law to hold outside of magne-
tostatics; after all, we derived it from the Biot-Savart law. However, in Maxwell’s
time there was no experimental reason to doubt that Ampère’s law was of wider
validity. The flaw was a purely theoretical one, and Maxwell fixed it by purely
theoretical arguments.

7.3.2 How Maxwell Fixed Ampère’s Law

The problem is on the right side of Eq. 7.36, which should be zero, but isn’t.
Applying the continuity equation (5.29) and Gauss’s law, the offending term can
be rewritten:
 
∂ρ ∂ ∂E
∇·J=− = − (0 ∇ · E) = −∇ · 0 .
∂t ∂t ∂t
If we were to combine 0 (∂E/∂t) with J, in Ampère’s law, it would be just right
to kill off the extra divergence:

∂E
∇ × B = μ0 J + μ0 0 . (7.37)
∂t

(Maxwell himself had other reasons for wanting to add this quantity to Ampère’s
law. To him, the rescue of the continuity equation was a happy dividend rather than
a primary motive. But today we recognize this argument as a far more compelling
one than Maxwell’s, which was based on a now-discredited model of the ether.)20
Such a modification changes nothing, as far as magnetostatics is concerned:
when E is constant, we still have ∇ × B = μ0 J. In fact, Maxwell’s term is hard
to detect in ordinary electromagnetic experiments, where it must compete for at-
tention with J—that’s why Faraday and the others never discovered it in the lab-
oratory. However, it plays a crucial role in the propagation of electromagnetic
waves, as we’ll see in Chapter 9.
Apart from curing the defect in Ampère’s law, Maxwell’s term has a cer-
tain aesthetic appeal: Just as a changing magnetic field induces an electric field
(Faraday’s law), so21

A changing electric field induces a magnetic field.

20 For the history of this subject, see A. M. Bork, Am. J. Phys. 31, 854 (1963).
21 See footnote 8 (page 313) for commentary on the word “induce.” The same issue arises here: Should

a changing electric field be regarded as an independent source of magnetic field (along with current)?
In a proximate sense it does function as a source, but since the electric field itself was produced by
charges and currents, they alone are the “ultimate” sources of E and B. See S. E. Hill, Phys. Teach.
49, 343 (2011); for a contrary view, see C. Savage, Phys. Teach. 50, 226 (2012).
7.3 Maxwell’s Equations 335

Of course, theoretical convenience and aesthetic consistency are only suggestive—

there might, after all, be other ways to doctor up Ampère’s law. The real confir-
mation of Maxwell’s theory came in 1888 with Hertz’s experiments on electro-
magnetic waves.
Maxwell called his extra term the displacement current:

∂E
Jd ≡ 0 . (7.38)
∂t

(It’s a misleading name; 0 (∂E/∂t) has nothing to do with current, except that it
adds to J in Ampère’s law.) Let’s see now how displacement current resolves the
paradox of the charging capacitor (Fig. 7.43). If the capacitor plates are very close
together (I didn’t draw them that way, but the calculation is simpler if you assume
this), then the electric field between them is

1 1 Q
E= σ = ,
0 0 A

where Q is the charge on the plate and A is its area. Thus, between the plates

∂E 1 dQ 1
= = I.
∂t 0 A dt 0 A

Now, Eq. 7.37 reads, in integral form,

   
∂E
B · dl = μ0 Ienc + μ0 0 · da. (7.39)
∂t

If we choose the flat surface, then E = 0 and Ienc = I . If, on the other hand, we
use the balloon-shaped surface, then Ienc = 0, but (∂E/∂t) · da = I /0 . So we
get the same answer for either surface, though in the first case it comes from the
conduction current, and in the second from the displacement current.

Example 7.14. Imagine two concentric metal spherical shells (Fig. 7.44).
The inner one (radius a) carries a charge Q(t), and the outer one (radius b) an
opposite charge −Q(t). The space between them is filled with Ohmic material of
conductivity σ , so a radial current flows:

1 Q σQ
J = σE = σ r̂; I = − Q̇ = J · da = .
4π 0 r 2 0

This configuration is spherically symmetrical, so the magnetic field hasto be zero

(the only direction it could possibly point is radial, and ∇ · B = 0 ⇒ B · da =
B(4πr 2 ) = 0, so B = 0). What? I thought currents produce magnetic fields! Isn’t
that what Biot-Savart and Ampère taught us? How can there be a J with no
accompanying B?
336 Chapter 7 Electrodynamics

J b
a

FIGURE 7.44

Solution
This is not a static configuration: Q, E, and J are all functions of time; Ampère
and Biot-Savart do not apply. The displacement current

∂E 1 Q̇ Q
Jd = 0 = r̂ = −σ r̂
∂t 4π r 2 4π 0r 2

exactly cancels the conduction current (in Eq. 7.37), and the magnetic field
(determined by ∇ · B = 0, ∇ × B = 0) is indeed zero.

Problem 7.34 A fat wire, radius a, carries a constant current I , uniformly dis-
tributed over its cross section. A narrow gap in the wire, of width w  a, forms
a parallel-plate capacitor, as shown in Fig. 7.45. Find the magnetic field in the gap,
at a distance s < a from the axis.

a +σ −σ
I I

FIGURE 7.45

Problem 7.35 The preceding problem was an artificial model for the charging ca-
pacitor, designed to avoid complications associated with the current spreading out
over the surface of the plates. For a more realistic model, imagine thin wires that
connect to the centers of the plates (Fig. 7.46a). Again, the current I is constant,
the radius of the capacitor is a, and the separation of the plates is w  a. Assume
that the current flows out over the plates in such a way that the surface charge is
uniform, at any given time, and is zero at t = 0.

(a) Find the electric field between the plates, as a function of t.

(b) Find the displacement current through a circle of radius s in the plane mid-
way between the plates. Using this circle as your “Amperian loop,” and the flat
surface that spans it, find the magnetic field at a distance s from the axis.
 7.3 Maxwell’s Equations 337

I I I s I

w
(a) (b)

FIGURE 7.46

(c) Repeat part (b), but this time use the cylindrical surface in Fig. 7.46(b), which
is open at the right end and extends to the left through the plate and terminates
outside the capacitor. Notice that the displacement current through this surface
is zero, and there are two contributions to Ienc .22

Problem 7.36 Refer to Prob. 7.16, to which the correct answer was
μ0 I0 ω a 
E(s, t) = sin(ωt) ln ẑ.
2π s

(a) Find the displacement current density Jd .

(b) Integrate it to get the total displacement current,


Id = Jd · da.

(c) Compare Id and I . (What’s their ratio?) If the outer cylinder were, say, 2 mm in
diameter, how high would the frequency have to be, for Id to be 1% of I ? [This
problem is designed to indicate why Faraday never discovered displacement
currents, and why it is ordinarily safe to ignore them unless the frequency is
extremely high.]

7.3.3 Maxwell’s Equations

In the last section we put the finishing touches on Maxwell’s equations:

1
(i) ∇ · E = ρ (Gauss’s law),
0
(ii) ∇ · B = 0 (no name),
∂B (7.40)
(iii) ∇ × E = − (Faraday’s law),
∂t
∂E
(iv) ∇ × B = μ0 J + μ0 0 (Ampère’s law with
∂t
Maxwell’s correction).

22 This problem raises an interesting quasi-philosophical question: If you measure B in the laboratory,
have you detected the effects of displacement current (as (b) would suggest), or merely confirmed the
effects of ordinary currents (as (c) implies)? See D. F. Bartlett, Am. J. Phys. 58, 1168 (1990).
338 Chapter 7 Electrodynamics

Together with the force law,

F = q(E + v × B), (7.41)
23
they summarize the entire theoretical content of classical electrodynamics (save
for some special properties of matter, which we encountered in Chapters 4 and 6).
Even the continuity equation,
∂ρ
∇·J=− , (7.42)
∂t
which is the mathematical expression of conservation of charge, can be derived
from Maxwell’s equations by applying the divergence to number (iv).
I have written Maxwell’s equations in the traditional way, which emphasizes
that they specify the divergence and curl of E and B. In this form, they reinforce
the notion that electric fields can be produced either by charges (ρ) or by changing
magnetic fields (∂B/∂t), and magnetic fields can be produced either by currents
(J) or by changing electric fields (∂E/∂t). Actually, this is misleading, because
∂B/∂t and ∂E/∂t are themselves due to charges and currents. I think it is logically
preferable to write
⎫
1 ∂B ⎪
(i) ∇ · E = ρ, (iii) ∇ × E + = 0, ⎪
⎪
0 ∂t ⎬
(7.43)
∂E ⎪
⎪
⎪
(ii) ∇ · B = 0, (iv) ∇ × B − μ0 0 = μ0 J, ⎭
∂t
with the fields (E and B) on the left and the sources (ρ and J) on the right. This
notation emphasizes that all electromagnetic fields are ultimately attributable to
charges and currents. Maxwell’s equations tell you how charges produce fields;
reciprocally, the force law tells you how fields affect charges.

Problem 7.37 Suppose

1 q
E(r, t) = θ (vt − r )r̂; B(r, t) = 0
4π 0 r 2
(The theta function is defined in Prob. 1.46b). Show that these fields satisfy all of
Maxwell’s equations, and determine ρ and J. Describe the physical situation that
gives rise to these fields.

7.3.4 Magnetic Charge

There is a pleasing symmetry to Maxwell’s equations; it is particularly striking in
free space, where ρ and J vanish:
⎫
∂B ⎪
∇ · E = 0, ∇×E=− , ⎪
⎬
∂t
∂E ⎪ ⎪
∇ · B = 0, ∇ × B = μ0 0 . ⎭
∂t
23 Like any differential equations, Maxwell’s must be supplemented by suitable boundary conditions.
Because these are typically “obvious” from the context (e.g. E and B go to zero at large distances from
a localized charge distribution), it is easy to forget that they play an essential role.
7.3 Maxwell’s Equations 339

If you replace E by B and B by −μ0 0 E, the first pair of equations turns into the
second, and vice versa. This symmetry24 between E and B is spoiled, though, by
the charge term in Gauss’s law and the current term in Ampère’s law. You can’t
help wondering why the corresponding quantities are “missing” from ∇ · B = 0
and ∇ × E = −∂B/∂t. What if we had
⎫
1 ∂B
(i) ∇ · E = ρe , (iii) ∇ × E = −μ0 Jm − , ⎪⎪
⎬
0 ∂t
(7.44)
∂E ⎪ ⎪
⎭
(ii) ∇ · B = μ0 ρm , (iv) ∇ × B = μ0 Je + μ0 0 .
∂t
Then ρm would represent the density of magnetic “charge,” and ρe the density of
electric charge; Jm would be the current of magnetic charge, and Je the current of
electric charge. Both charges would be conserved:
∂ρm ∂ρe
∇ · Jm = − , and ∇ · Je = − . (7.45)
∂t ∂t
The former follows by application of the divergence to (iii), the latter by taking
the divergence of (iv).
In a sense, Maxwell’s equations beg for magnetic charge to exist—it would fit
in so nicely. And yet, in spite of a diligent search, no one has ever found any.25
As far as we know, ρm is zero everywhere, and so is Jm ; B is not on equal foot-
ing with E: there exist stationary sources for E (electric charges) but none for B.
(This is reflected in the fact that magnetic multipole expansions have no monopole
term, and magnetic dipoles consist of current loops, not separated north and south
“poles.”) Apparently God just didn’t make any magnetic charge. (In quantum elec-
trodynamics, by the way, it’s a more than merely aesthetic shame that magnetic
charge does not seem to exist: Dirac showed that the existence of magnetic charge
would explain why electric charge is quantized. See Prob. 8.19.)

Problem 7.38 Assuming that “Coulomb’s law” for magnetic charges (qm ) reads
μ0 qm 1 qm 2
F= r̂, (7.46)
4π r2
work out the force law for a monopole qm moving with velocity v through electric
and magnetic fields E and B.26

Problem 7.39 Suppose a magnetic monopole qm passes through a resistanceless

loop of wire with self-inductance L. What current is induced in the loop?27

24 Don’t be distracted by the pesky constants μ0 and 0 ; these are present only because the SI system
measures E and B in different units, and would not occur, for instance, in the Gaussian system.
25 For an extensive bibliography, see A. S. Goldhaber and W. P. Trower, Am. J. Phys. 58, 429 (1990).
26 For interesting commentary, see W. Rindler, Am. J. Phys. 57, 993 (1989).
27 This is one of the methods used to search for monopoles in the laboratory; see B. Cabrera, Phys.

Rev. Lett. 48, 1378 (1982).

340 Chapter 7 Electrodynamics

7.3.5 Maxwell’s Equations in Matter

Maxwell’s equations in the form 7.40 are complete and correct as they stand.
However, when you are working with materials that are subject to electric and
magnetic polarization there is a more convenient way to write them. For inside
polarized matter there will be accumulations of “bound” charge and current, over
which you exert no direct control. It would be nice to reformulate Maxwell’s
equations so as to make explicit reference only to the “free” charges and currents.
We have already learned, from the static case, that an electric polarization P
produces a bound charge density
ρb = −∇ · P (7.47)
(Eq. 4.12). Likewise, a magnetic polarization (or “magnetization”) M results in a
bound current
Jb = ∇ × M (7.48)
(Eq. 6.13). There’s just one new feature to consider in the nonstatic case: Any
change in the electric polarization involves a flow of (bound) charge (call it J p ),
which must be included in the total current. For suppose we examine a tiny chunk
of polarized material (Fig. 7.47). The polarization introduces a charge density
σb = P at one end and −σb at the other (Eq. 4.11). If P now increases a bit, the
charge on each end increases accordingly, giving a net current
∂σb ∂P
dI = da⊥ = da⊥ .
∂t ∂t
The current density, therefore, is
∂P
Jp = . (7.49)
∂t
This polarization current has nothing to do with the bound current Jb . The
latter is associated with magnetization of the material and involves the spin and
orbital motion of electrons; J p , by contrast, is the result of the linear motion of
charge when the electric polarization changes. If P points to the right, and is
increasing, then each plus charge moves a bit to the right and each minus charge
to the left; the cumulative effect is the polarization current J p . We ought to check
that Eq. 7.49 is consistent with the continuity equation:
∂P ∂ ∂ρb
∇ · Jp = ∇ · = (∇ · P) = − .
∂t ∂t ∂t

P
da⊥
+σb
−σb

FIGURE 7.47
7.3 Maxwell’s Equations 341

Yes: The continuity equation is satisfied; in fact, J p is essential to ensure the

conservation of bound charge. (Incidentally, a changing magnetization does not
lead to any analogous accumulation of charge or current. The bound current
Jb = ∇ × M varies in response to changes in M, to be sure, but that’s about it.)
In view of all this, the total charge density can be separated into two parts:
ρ = ρ f + ρb = ρ f − ∇ · P, (7.50)
and the current density into three parts:
∂P
J = J f + Jb + J p = J f + ∇ × M + . (7.51)
∂t
Gauss’s law can now be written as
1
∇·E= (ρ f − ∇ · P),
0
or
∇ · D = ρf, (7.52)
where, as in the static case,
D ≡ 0 E + P. (7.53)
Meanwhile, Ampère’s law (with Maxwell’s term) becomes
 
∂P ∂E
∇ × B = μ0 J f + ∇ × M + + μ0 0 ,
∂t ∂t
or
∂D
∇ × H = Jf + , (7.54)
∂t
where, as before,
1
H≡ B − M. (7.55)
μ0
Faraday’s law and ∇ · B = 0 are not affected by our separation of charge and
current into free and bound parts, since they do not involve ρ or J.
In terms of free charges and currents, then, Maxwell’s equations read

∂B
(i) ∇ · D = ρ f , (iii) ∇ × E = − ,
∂t
(7.56)
∂D
(ii) ∇ · B = 0, (iv) ∇ × H = J f + .
∂t

Some people regard these as the “true” Maxwell’s equations, but please under-
stand that they are in no way more “general” than Eq. 7.40; they simply reflect a
convenient division of charge and current into free and nonfree parts. And they
342 Chapter 7 Electrodynamics

have the disadvantage of hybrid notation, since they contain both E and D, both
B and H. They must be supplemented, therefore, by appropriate constitutive
relations, giving D and H in terms of E and B. These depend on the nature of
the material; for linear media
P = 0 χe E, and M = χm H, (7.57)
so
1
D = E, B,
and H= (7.58)
μ
where  ≡ 0 (1 + χe ) and μ ≡ μ0 (1 + χm ). Incidentally, you’ll remember that
D is called the electric “displacement”; that’s why the second term in the
Ampère/Maxwell equation (iv) came to be called the displacement current.
In this context,
∂D
Jd ≡ . (7.59)
∂t

Problem 7.40 Sea water at frequency ν = 4 × 108 Hz has permittivity  = 810 ,

permeability μ = μ0 , and resistivity ρ = 0.23  · m. What is the ratio of conduc-
tion current to displacement current? [Hint: Consider a parallel-plate capacitor im-
mersed in sea water and driven by a voltage V0 cos (2π νt).]

7.3.6 Boundary Conditions

In general, the fields E, B, D, and H will be discontinuous at a boundary between
two different media, or at a surface that carries a charge density σ or a current den-
sity K. The explicit form of these discontinuities can be deduced from Maxwell’s
equations (7.56), in their integral form
 ⎫
(i) D · da = Q fenc ⎪⎪
⎬
S
 over any closed surface S.
⎪
⎪
(ii) B · da = 0 ⎭
S
  ⎫
d ⎪
(iii) E · dl = − B · da ⎪
⎬ for any surface S
P dt S
  bounded by the
d ⎪
⎪
(iv) H · dl = I fenc + D · da ⎭ closed loop P.
P dt S

Applying (i) to a tiny, wafer-thin Gaussian pillbox extending just slightly into
the material on either side of the boundary (Fig. 7.48), we obtain:
D1 · a − D2 · a = σ f a.
(The positive direction for a is from 2 toward 1. The edge of the wafer con-
tributes nothing in the limit as the thickness goes to zero; nor does any volume
7.3 Maxwell’s Equations 343

D1

a
1

σf

D2

FIGURE 7.48

charge density.) Thus, the component of D that is perpendicular to the interface is

discontinuous in the amount

D1⊥ − D2⊥ = σ f . (7.60)

Identical reasoning, applied to equation (ii), yields

B1⊥ − B2⊥ = 0. (7.61)

Turning to (iii), a very thin Amperian loop straddling the surface gives

d
E1 · l − E2 · l = − B · da.
dt S
But in the limit as the width of the loop goes to zero,
 the flux vanishes. (I have
already dropped the contribution of the two ends to E · dl, on the same grounds.)
Therefore,

E1 − E2 = 0. (7.62)

That is, the components of E parallel to the interface are continuous across the
boundary. By the same token, (iv) implies

H1 · l − H2 · l = I fenc ,

where I fenc is the free current passing through the Amperian loop. No volume
current density will contribute (in the limit of infinitesimal width), but a surface
current can. In fact, if n̂ is a unit vector perpendicular to the interface (pointing
from 2 toward 1), so that (n̂ × l) is normal to the Amperian loop (Fig. 7.49), then

I fenc = K f · (n̂ × l) = (K f × n̂) · l,

344 Chapter 7 Electrodynamics

1
l
2

Kf

FIGURE 7.49

and hence

H1 − H2 = K f × n̂. (7.63)

So the parallel components of H are discontinuous by an amount proportional to

the free surface current density.
Equations 7.60-63 are the general boundary conditions for electrodynamics. In
the case of linear media, they can be expressed in terms of E and B alone:
⎫
(i) 1 E 1⊥ − 2 E 2⊥ = σ f , (iii) E1 − E2 = 0, ⎪
⎪
⎬
(7.64)
1 1 ⎪
(ii) B1⊥ − B2⊥ = 0, (iv) B1 − B2 = K f × n̂. ⎪
⎭
μ1 μ2
In particular, if there is no free charge or free current at the interface, then

(i) 1 E 1⊥ − 2 E 2⊥ = 0, (iii) E1 − E2 = 0,
(7.65)
1 1
(ii) B1⊥ − B2⊥ = 0, (iv) B1 − B = 0.
μ1 μ2 2
As we shall see in Chapter 9, these equations are the basis for the theory of reflec-
tion and refraction.

More Problems on Chapter 7

! Problem 7.41 Two long, straight copper pipes, each of radius a, are held a dis-
tance 2d apart (see Fig. 7.50). One is at potential V0 , the other at −V0 . The space
surrounding the pipes is filled with weakly conducting material of conductivity σ .
Find the current per unit length that flows from one pipe to the other. [Hint: Refer
to Prob. 3.12.]
 7.3 Maxwell’s Equations 345

σ
a a
–V0 2d +V0

FIGURE 7.50

! Problem 7.42 A rare case in which the electrostatic field E for a circuit can actually
be calculated is the following:28 Imagine an infinitely long cylindrical sheet, of
uniform resistivity and radius a. A slot (corresponding to the battery) is maintained
at ±V0 /2, at φ = ±π , and a steady current flows over the surface, as indicated in
Fig. 7.51. According to Ohm’s law, then,
V0 φ
V (a, φ) = , (−π < φ < +π ).
2π

s φ
+V0 /2 a
−V0 /2 x

FIGURE 7.51

(a) Use separation of variables in cylindrical coordinates to determine V (s, φ) in-

side and outside the cylinder. [Answer: (V0 /π ) tan−1 [(s sin φ)/(a + s cos φ)],
(s < a); (V0 /π ) tan−1 [(a sin φ)/(s + a cos φ)], (s > a)]

(b) Find the surface charge density on the cylinder. [Answer: (0 V0 /πa) tan(φ/2)]

Problem 7.43 The magnetic field outside a long straight wire carrying a steady
current I is
μ0 I
B= φ̂.
2π s
The electric field inside the wire is uniform:
Iρ
E= ẑ,
πa 2
28 M.A. Heald, Am. J. Phys. 52, 522 (1984). See also J. A. Hernandes and A. K. T. Assis, Phys. Rev. E
68, 046611 (2003).
346 Chapter 7 Electrodynamics

where ρ is the resistivity and a is the radius (see Exs. 7.1 and 7.3). Question: What is
the electric field outside the wire?29 The answer depends on how you complete the
circuit. Suppose the current returns along a perfectly conducting grounded coax-
ial cylinder of radius b (Fig. 7.52). In the region a < s < b, the potential V (s, z)
satisfies Laplace’s equation, with the boundary conditions
Iρz
(i) V (a, z) = − ; (ii) V (b, z) = 0.
πa 2

I
b
a I
z

FIGURE 7.52

This does not suffice to determine the answer—we still need to specify boundary
conditions at the two ends (though for a long wire it shouldn’t matter much). In the
literature, it is customary to sweep this ambiguity under the rug by simply stipulat-
ing that V (s, z) is proportional to z: V (s, z) = z f (s). On this assumption:

(a) Determine f (s).

(b) Find E(s, z).

(c) Calculate the surface charge density σ (z) on the wire.

[Answer: V = (−I zρ/πa 2 )[ln(s/b)/ ln(a/b)] This is a peculiar result, since E s and
σ (z) are not independent of z—as one would certainly expect for a truly infinite
wire.]

Problem 7.44 In a perfect conductor, the conductivity is infinite, so E = 0

(Eq. 7.3), and any net charge resides on the surface (just as it does for an imperfect
conductor, in electrostatics).

(a) Show that the magnetic field is constant (∂B/∂t = 0), inside a perfect
conductor.

(b) Show that the magnetic flux through a perfectly conducting loop is constant.
A superconductor is a perfect conductor with the additional property that
the (constant) B inside is in fact zero. (This “flux exclusion” is known as the
Meissner effect.30 )

29 This is a famous problem, first analyzed by Sommerfeld, and is known in its most recent incarna-
tion as Merzbacher’s puzzle. A. Sommerfeld, Electrodynamics, p. 125 (New York: Academic Press,
1952); E. Merzbacher, Am. J. Phys. 48, 178 (1980); further references in R. N. Varnay and L. H. Fisher,
Am. J. Phys. 52, 1097 (1984).
30 The Meissner effect is sometimes referred to as “perfect diamagnetism,” in the sense that the field

inside is not merely reduced, but canceled entirely. However, the surface currents responsible for this
are free, not bound, so the actual mechanism is quite different.
 7.3 Maxwell’s Equations 347

(c) Show that the current in a superconductor is confined to the surface.

(d) Superconductivity is lost above a certain critical temperature (Tc ), which varies
from one material to another. Suppose you had a sphere (radius a) above its
critical temperature, and you held it in a uniform magnetic field B0 ẑ while cool-
ing it below Tc . Find the induced surface current density K, as a function of the
polar angle θ .

Problem 7.45 A familiar demonstration of superconductivity (Prob. 7.44) is the lev-

itation of a magnet over a piece of superconducting material. This phenomenon can
be analyzed using the method of images.31 Treat the magnet as a perfect dipole m,
a height z above the origin (and constrained to point in the z direction), and pretend
that the superconductor occupies the entire half-space below the x y plane. Because
of the Meissner effect, B = 0 for z ≤ 0, and since B is divergenceless, the normal
(z) component is continuous, so Bz = 0 just above the surface. This boundary con-
dition is met by the image configuration in which an identical dipole is placed at
−z, as a stand-in for the superconductor; the two arrangements therefore produce
the same magnetic field in the region z > 0.

(a) Which way should the image dipole point (+z or −z)?

(b) Find the force on the magnet due to the induced currents in the superconductor
(which is to say, the force due to the image dipole). Set it equal to Mg (where
M is the mass of the magnet) to determine the height h at which the magnet will
“float.” [Hint: Refer to Prob. 6.3.]

(c) The induced current on the surface of the superconductor (the x y plane) can
be determined from the boundary condition on the tangential component of B
(Eq. 5.76): B = μ0 (K × ẑ). Using the field you get from the image configura-
tion, show that
3mr h
K=− φ̂,
2π(r 2 + h 2 )5/2
where r is the distance from the origin.

! Problem 7.46 If a magnetic dipole levitating above an infinite superconducting

plane (Prob. 7.45) is free to rotate, what orientation will it adopt, and how high
above the surface will it float?

Problem 7.47 A perfectly conducting spherical shell of radius a rotates about the
z axis with angular velocity ω, in a uniform magnetic field B = B0 ẑ. Calculate the
emf developed between the “north pole” and the equator. [Answer: 12 B0 ωa 2 ]

! Problem 7.48 Refer to Prob. 7.11 (and use the result of Prob. 5.42): How long does
is take a falling circular ring (radius a, mass m, resistance R) to cross the bottom of
the magnetic field B, at its (changing) terminal velocity?

31 W. M. Saslow, Am. J. Phys. 59, 16 (1991).

348 Chapter 7 Electrodynamics

Problem 7.49

(a) Referring to Prob. 5.52(a) and Eq. 7.18, show that

∂A
E=− , (7.66)
∂t
for Faraday-induced electric fields. Check this result by taking the divergence
and curl of both sides.

(b) A spherical shell of radius R carries a uniform surface charge σ . It spins about
a fixed axis at an angular velocity ω(t) that changes slowly with time. Find the
electric field inside and outside the sphere. [Hint: There are two contributions
here: the Coulomb field due to the charge, and the Faraday field due to the
changing B. Refer to Ex. 5.11.]

Problem 7.50 Electrons undergoing cyclotron motion can be sped up by increasing

the magnetic field; the accompanying electric field will impart tangential acceler-
ation. This is the principle of the betatron. One would like to keep the radius of
the orbit constant during the process. Show that this can be achieved by designing
a magnet such that the average field over the area of the orbit is twice the field at
the circumference (Fig. 7.53). Assume the electrons start from rest in zero field,
and that the apparatus is symmetric about the center of the orbit. (Assume also that
the electron velocity remains well below the speed of light, so that nonrelativis-
tic mechanics applies.) [Hint: Differentiate Eq. 5.3 with respect to time, and use
F = ma = q E.]
z

B I
Electron
orbit v
s

y
x

FIGURE 7.53 FIGURE 7.54

Problem 7.51 An infinite wire carrying a constant current I in the ẑ direction is

moving in the y direction at a constant speed v. Find the electric field, in the qua-
sistatic approximation, at the instant the wire coincides with the z axis (Fig. 7.54).
[Answer: −(μ0 I v/2π s) sin φ ẑ]

Problem 7.52 An atomic electron (charge q) circles about the nucleus (charge Q)
in an orbit of radius r ; the centripetal acceleration is provided, of course, by the
Coulomb attraction of opposite charges. Now a small magnetic field d B is slowly
turned on, perpendicular to the plane of the orbit. Show that the increase in kinetic
energy, dT , imparted by the induced electric field, is just right to sustain circular
motion at the same radius r. (That’s why, in my discussion of diamagnetism,
I assumed the radius is fixed. See Sect. 6.1.3 and the references cited there.)
7.3 Maxwell’s Equations 349

I
b b
V1 R1 Solenoid R2 V2
a a

FIGURE 7.55

Problem 7.53 The current in a long solenoid is increasing linearly with time, so the
flux is proportional to t: = αt. Two voltmeters are connected to diametrically op-
posite points (A and B), together with resistors (R1 and R2 ), as shown in Fig. 7.55.
What is the reading on each voltmeter? Assume that these are ideal voltmeters
that draw negligible
 b current (they have huge internal resistance), and that a volt-
meter registers − a E · dl between the terminals and through the meter. [Answer:
V1 = α R1 /(R1 + R2 ); V2 = −α R2 /(R1 + R2 ). Notice that V1 = V2 , even though
they are connected to the same points!32 ]

P V Q
90°
r

FIGURE 7.56

Problem 7.54 A circular wire loop (radius r , resistance R) encloses a region of uni-
form magnetic field, B, perpendicular to its plane. The field (occupying the shaded
region in Fig. 7.56) increases linearly with time (B = αt). An ideal voltmeter (infi-
nite internal resistance) is connected between points P and Q.

(a) What is the current in the loop?

(b) What does the voltmeter read? [Answer: αr 2 /2]

Problem 7.55 In the discussion of motional emf (Sect. 7.1.3) I assumed that the
wire loop (Fig. 7.10) has a resistance R; the current generated is then I = v Bh/R.
But what if the wire is made out of perfectly conducting material, so that R is zero?
In that case, the current is limited only by the back emf associated with the self-
inductance L of the loop (which would ordinarily be negligible in comparison with
I R). Show that in this régime the loop (mass√m) executes simple harmonic motion,
and find its frequency.33 [Answer: ω = Bh/ m L]

32 R. H. Romer, Am. J. Phys. 50, 1089 (1982). See also H. W. Nicholson, Am. J. Phys. 73, 1194 (2005);

B. M. McGuyer, Am. J. Phys. 80, 101 (2012).

33 Fora collection of related problems, see W. M. Saslow, Am. J. Phys. 55, 986 (1987), and R. H.
Romer, Eur. J. Phys. 11, 103 (1990).
350 Chapter 7 Electrodynamics

Problem 7.56

(a) Use the Neumann formula (Eq. 7.23) to calculate the mutual inductance of the
configuration in Fig. 7.37, assuming a is very small (a  b, a  z). Compare
your answer to Prob. 7.22.

(b) For the general case (not assuming a is small), show that
 
μ0 πβ  15
M= abβ 1 + β 2 + . . . ,
2 8
where
ab
β≡ .
z 2 + a 2 + b2

Secondary
Primary (N2 turns)
(N1 turns)

FIGURE 7.57

Problem 7.57 Two coils are wrapped around a cylindrical form in such a way that
the same flux passes through every turn of both coils. (In practice this is achieved by
inserting an iron core through the cylinder; this has the effect of concentrating the
flux.) The primary coil has N1 turns and the secondary has N2 (Fig. 7.57). If the
current I in the primary is changing, show that the emf in the secondary is given by
E2 N2
= , (7.67)
E1 N1
where E1 is the (back) emf of the primary. [This is a primitive transformer—a
device for raising or lowering the emf of an alternating current source. By choosing
the appropriate number of turns, any desired secondary emf can be obtained. If you
think this violates the conservation of energy, study Prob. 7.58.]

Problem 7.58 A transformer (Prob. 7.57) takes an input AC voltage of amplitude

V1 , and delivers an output voltage of amplitude V2 , which is determined by the turns
ratio (V2 /V1 = N2 /N1 ). If N2 > N1 , the output voltage is greater than the input
voltage. Why doesn’t this violate conservation of energy? Answer: Power is the
product of voltage and current; if the voltage goes up, the current must come down.
The purpose of this problem is to see exactly how this works out, in a simplified
model.
7.3 Maxwell’s Equations 351

(a) In an ideal transformer, the same flux passes through all turns of the primary
and of the secondary. Show that in this case M 2 = L 1 L 2 , where M is the mutual
inductance of the coils, and L 1 , L 2 are their individual self-inductances.

(b) Suppose the primary is driven with AC voltage Vin = V1 cos (ωt), and the sec-
ondary is connected to a resistor, R. Show that the two currents satisfy the
relations
d I1 d I2 d I2 d I1
L1 +M = V1 cos (ωt); L2 +M = −I2 R.
dt dt dt dt

(c) Using the result in (a), solve these equations for I1 (t) and I2 (t). (Assume I1 has
no DC component.)

(d) Show that the output voltage (Vout = I2 R) divided by the input voltage (Vin ) is
equal to the turns ratio: Vout /Vin = N2 /N1 .

(e) Calculate the input power (Pin = Vin I1 ) and the output power (Pout = Vout I2 ),
and show that their averages over a full cycle are equal.

Problem 7.59 An infinite wire runs along the z axis; it carries a current I (z) that is
a function of z (but not of t), and a charge density λ(t) that is a function of t (but
not of z).

(a) By examining the charge flowing into a segment dz in a time dt, show that
dλ/dt = −d I /dz. If we stipulate that λ(0) = 0 and I (0) = 0, show that
λ(t) = kt, I (z) = −kz, where k is a constant.

(b) Assume for a moment that the process is quasistatic, so the fields are given
by Eqs. 2.9 and 5.38. Show that these are in fact the exact fields, by confirm-
ing that all four of Maxwell’s equations are satisfied. (First do it in differential
form, for the region s > 0, then in integral form for the appropriate Gaussian
cylinder/Amperian loop straddling the axis.)

Problem 7.60 Suppose J(r) is constant in time but ρ(r, t) is not—conditions that
might prevail, for instance, during the charging of a capacitor.

(a) Show that the charge density at any particular point is a linear function of time:

ρ(r, t) = ρ(r, 0) + ρ̇(r, 0)t,

where ρ̇(r, 0) is the time derivative of ρ at t = 0. [Hint: Use the continuity

equation.]
This is not an electrostatic or magnetostatic configuration;34 nevertheless, rather
surprisingly, both Coulomb’s law (Eq. 2.8) and the Biot-Savart law (Eq. 5.42)
hold, as you can confirm by showing that they satisfy Maxwell’s equations. In
particular:

34 Some authors would regard this as magnetostatic, since B is independent of t. For them, the Biot-
Savart law is a general rule of magnetostatics, but ∇ · J = 0 and ∇ × B = μ0 J apply only under the
additional assumption that ρ is constant. In such a formulation, Maxwell’s displacement term can
(in this very special case) be derived from the Biot-Savart law, by the method of part (b). See D. F.
Bartlett, Am. J. Phys. 58, 1168 (1990); D. J. Griffiths and M. A. Heald, Am. J. Phys. 59, 111 (1991).
352 Chapter 7 Electrodynamics

(b) Show that


μ0 J(r ) × r̂
B(r) = dτ 
4π r2
obeys Ampère’s law with Maxwell’s displacement current term.
Problem 7.61 The magnetic field of an infinite straight wire carrying a steady cur-
rent I can be obtained from the displacement current term in the Ampère/Maxwell
law, as follows: Picture the current as consisting of a uniform line charge λ mov-
ing along the z axis at speed v (so that I = λv), with a tiny gap of length , which
reaches the origin at time t = 0. In the next instant (up to t = /v) there is no real
current passing through a circular Amperian loop in the x y plane, but there is a
displacement current, due to the “missing” charge in the gap.
(a) Use Coulomb’s law to calculate the z component of the electric field, for points
in the x y plane a distance s from the origin, due to a segment of wire with
uniform density −λ extending from z 1 = vt −  to z 2 = vt.

(b) Determine the flux of this electric field through a circle of radius a in the x y
plane.

(c) Find the displacement current through this circle. Show that Id is equal to I , in
the limit as the gap width () goes to zero.35
Problem 7.62 A certain transmission line is constructed from two thin metal “rib-
bons,” of width w, a very small distance h  w apart. The current travels down
one strip and back along the other. In each case, it spreads out uniformly over the
surface of the ribbon.
(a) Find the capacitance per unit length, C.

(b) Find the inductance per unit length, L.

(c) What is the product LC, numerically? [L and C will, of course, vary from one
kind of transmission line to another, but their product is a universal constant—
check, for example, the cable in Ex. 7.13—provided the space between the con-
ductors is a vacuum. In the theory of transmission lines, this product
√ is related
to the speed with which a pulse propagates down the line: v = 1/ LC.]

(d) If the strips are insulated from one another by a nonconducting material of per-
mittivity  and permeability μ, what then is the product LC? What is the propa-
gation speed? [Hint: see Ex. 4.6; by what factor does L change when an inductor
is immersed in linear material of permeability μ?]
Problem 7.63 Prove Alfven’s theorem: In a perfectly conducting fluid (say, a gas
of free electrons), the magnetic flux through any closed loop moving with the fluid
is constant in time. (The magnetic field lines are, as it were, “frozen” into the fluid.)
(a) Use Ohm’s law, in the form of Eq. 7.2, together with Faraday’s law, to prove
that if σ = ∞ and J is finite, then
∂B
= ∇ × (v × B).
∂t

35 For a slightly different approach to the same problem, see W. K. Terry, Am. J. Phys. 50, 742 (1982).
7.3 Maxwell’s Equations 353

P′
R
S′

S vdt

FIGURE 7.58

(b) Let S be the surface bounded by the loop (P) at time t, and S  a surface bounded
by the loop in its new position (P  ) at time t + dt (see Fig. 7.58). The change
in flux is
 
d = B(t + dt) · da − B(t) · da.
S S

Use ∇ · B = 0 to show that

  
B(t + dt) · da + B(t + dt) · da = B(t + dt) · da
S R S


(where R is the “ribbon” joining P and P ), and hence that
 
∂B
d = dt · da − B(t + dt) · da
S ∂t R

(for infinitesimal dt). Use the method of Sect. 7.1.3 to rewrite the second inte-
gral as

dt (B × v) · dl,
P

and invoke Stokes’ theorem to conclude that

  
d ∂B
= − ∇ × (v × B) · da.
dt S ∂t
Together with the result in (a), this proves the theorem.

Problem 7.64

(a) Show that Maxwell’s equations with magnetic charge (Eq. 7.44) are invariant
under the duality transformation
⎫
E = E cos α + cB sin α, ⎪ ⎪
⎬
cB = cB cos α − E sin α,
 (7.68)
cqe = cqe cos α + qm sin α, ⎪ ⎪
⎭
qm = qm cos α − cqe sin α,
√
where c ≡ 1/ 0 μ0 and α is an arbitrary rotation angle in “E/B-space.” Charge
and current densities transform in the same way as qe and qm . [This means, in
354 Chapter 7 Electrodynamics

particular, that if you know the fields produced by a configuration of electric

charge, you can immediately (using α = 90◦ ) write down the fields produced
by the corresponding arrangement of magnetic charge.]

(b) Show that the force law (Prob. 7.38)

 
1
F = qe (E + v × B) + qm B − 2 v × E (7.69)
c
is also invariant under the duality transformation.