Professional Documents
Culture Documents
Science 165 3895 757
Science 165 3895 757
least with regard to statistics, which The unit of pure magnetic charge has
g/e=tan 0 must be Fermi-Dirac if baryons are a magnitude given by
Then the integer in any charge-quanti- to be built from them. It would not do
to have only one value of magnetic g2/hc 36(137)
zation condition vanishes, and no fur-
ther restrictions appear. Of course, as charge, for then magnetically neutral We come now to a very important
the use of the angle 0 signals, this situa- composites could be produced in only question. What name shall we give to
tion is just a charge-rotated version of one way, namely, by the combination the fundamental dual-charged particle
pure electric charges, where no charge of particle and antiparticle. That would (6)? The particle ending -on is obliga-
quantization exists. But it emphasizes only manufacture mesons. But it is tory. As evidenced by the use of the
the weakening of charge quantization enough to have just two different values provisional phrase "dual-charged par-
that the consideration of dual-charged of magnetic charge, which we take to ticle," the basic aspect that should be
particles entails. Suppose there are two be 2go and - go. Now a magnetically commemorated in the name is the du-
kinds of dual-charged particles with a neutral composite is also formed from alistic or dyadic character of the charge
common magnetic charge go, but dif- three constituents, of magnetic charges that the particle bears. There are vari-
ferent electric charges, el and e2. The 2go,-go,-go, and this is a Fermi- ous short Greek and Latin combining
charge quantization condition is Dirac particle. It is satisfactory that forms that could be applied: bi-, di-,
this pattern of magnetic charge is un- duo-, dyo-, as well as longer words such
(ei-e2) go/hc = 2n symmetrical, in contrast with the meson as dyadikos-, of two. Dyadikon surely
22 AUGUST 1969 759
has a ring to it. But being mindful that This formula can be compared with Charge Exchange
mesotron became shortened to meson, empirical meson mass spectra (7).
I believe that dyon is a better choice. Simple but accurate representations of Are other known phenomena omitted
The symbol D will not often lead to the mass splittings within the known in this dynamical scheme? Although
confusion with deuterium, particularly families of 9 = 3 X 3 particles enable conventional electromagnetic and strong
if we add labels that indicate the elec- one to remove this mass structure, and interactions have been given a general-
tric and magnetic charges, for which the resulting squared masses are pro- ized electromagnetic interpretation,
we use e and g as units. Thus FD½ portional to an integer with actual there is no reference to the so-called
is the dyon with electric charge -3e values of 0, 1, 2, or 3. The scale is weak interactions. This type of inter-
=-eo and magnetic charge 2/3g- 2go. supplied by the mass of the p-meson, action can be viewed as a mechanism
What is the mass of a dyon? Let which gives the identification of electric-charge exchange among
us be clear about this; any estimate is members of the same particle family,
sheer guesswork. We do not have the (2MD)2 = 1/2 nompo_ including the lepton family (L) of
wit to connect the known properties of The specific value of no noted above electron, muon, and neutrino. It is pos-
the composites-hadrons-with the un- refers to the individual magnetic charge sible, but not necessary, to regard this
known properties of the constituents- magnitude g0. If 2go were considered, charge exchange as proceeding through
dyons.The interaction strength far sur- no would be four times larger. We the intermediary of an unknown, heavy,
passes anything for which such skill shall use a weighted mean of these charged boson, as has been proposed
exists. But a beginning must be made. values, which effectively equates no to several times under different names.
Consider the nonrelativistic behavior 4(137), and then The commonly employed symbol is
of two widely separated dyons, of W(weak), which we use in writing
common mass MD, that are combined MD (137/2)½ mp -
some typical particle reactions
in a hydrogenlike structure. The energy 6 billion electron volts
I would not risk more than three L+ *L + W+, -½D%Y-½D-D± +W+
expression is
groschen on the likelihood of this esti- There is a striking analogy with elec-
21 Pr,+
-
(h)21 + mate, but at least it is an optimistic tromagnetic emission and absorption of
one, in relation to current accelerator photons (which motivated my own ear-
(giga + e1e2)/r plans. ly speculations in this direction) since
where m is the reduced -mass, 1/2MD. Let us return to what was termed the observed interactions are essentially
Hydrogen energy levels depend only the fine and hyperfine structure of the vectorial in character and have a cer-
upon the principal quantum number mass spectrum. The hyperfine structure tain universality in strength. In more
n =nr + I + 1, where I is here given is an electric-charge dependence which detail, these weak interactions are
by causes the interaction strength to vary known to be CP-conserving, but C-
by a fraction and P-violating, in a way that depends
1(1 + 1) =j( + 1)-a upon the sign of the electric charge
As an initial approximation, let us
e2/gO that is exchanged. It would seem that
ignore the fine structure of order ego/ The corresponding change in squared they destroy the charge rotational in-
hc 1, and the hyperfine structure of mass is variance of the Maxwell equations and
order e2/lhc- 1/137. The appearance thereby help to establish the absolute
of the formulas will also be simplified
~ MD2 e2/go- (1/137) mp' distinction between electric and mag-
by the adoption of atomic units for That is indeed the magnitude observed netic charge.
which h = c = 1. With the specific for the charge dependence as illustrated I find it natural to imagine a mag-
choice by the K-mesons netic analog of these processes, with a
correspondingly stronger coupling, that
g9g2 =
-gO0 mKo - mK*' 0.7 x 10-' mP could be mediated by a boson of unit
the Bohr formula supplies the total but the sense of the splitting is oppo- magnetic charge, S (strong). A typical
mass, or better, the squared mass as site to what one would expect from the magnetic charge-exchange process for
simple mechanism considered. The fine the dyon is
MS = (2MD)' [-1
I structure is represented by a variation
in n or k that is of the order of unity. %D-½ <4 -½D-½ + +S
This result is valid only when the sec- Such is the qualitative empirical situa- It is entirely possible that both S and
ond term is small compared to unity, tion, as illustrated by a comparison of W are fictitious and should be under-
corresponding to very large quantum the K-meson, which belongs to the stood only as a shorthand for the direct
numbers, n > no, where k = 0 nonuplet, with the p-meson, a exchange of charge between pairs of
member of the k = 1 nonuplet: particles of various types. We shall not
n, = ½/2 go2 2(137) dwell on the conceivable magnetic
mK' 0.4 mp2 counterparts of leptons, except to won-
But, faute de mieux, let us abandon der if the neutral neutrino(s) could be
caution and extrapolate down to zero But the quantitative details are wrong.
Only the value of the electric charge common to both families. The mecha-
mass! That is reached at n =nO. The nism represented by the magnetic par-
neighboring states identified by n = no would seem to be relevant, whereas
+ k, where k= 1, 2, . . ., are approx- observed mass spectra are labeled by ticle S produces a rapid exchange of
imately represented by and give meaning to the properties of magnetic charge among the dyons that
isotopic spin and hypercharge. Some- constitute a hadron. It may be that the
Ma = (2MD)2(2/no)k thing is missing. result is a very short time scale aver-
760 SCIENCE, VOL. 165
aging out of the magnetic charge on that sets the scale, and one empirical erected on the basis of the speculative
an individual dyon. Indeed, that is factor of the order of unity gives a but theoretically well-founded hypoth-
what is suggested by a naive view of reasonable account of neutron and pro- esis that electric and magnetic charge
the empirical baryon situation. The ton magnetic moments. The analogous can reside on a single particle. I hope
pattern of low-lying multiplets is cor- electric dipole moment is proportional that these suggestive, if inadequate, ar-
rectly represented if we unite a three- to the sum of dyon magnetic charges guments will be sufficiently persuasive
valued electric label with a two-valued multiplied by spin vectors. Since the to encourage a determined experimental
spin index and consider only totally total magnetic charge is zero, the elec- quest for the portal to this unknown
symmetrical arrangements of three such tric dipole moment would vanish if all new world of matter, for
index pairs as though three constituents three dyons were dynamically the same
were in a symmetrical orbital state and and therefore had identical average spin Nothing is too wonderful to be true, if
obeyed Bose-Einstein statistics. Conflict vectors. But surely the dyon of mag- it be consistent with the laws of nature,
with the physical Fermi-Dirac statistics netic charge 23 is in a different environ- and in such things as these, experiment
of dyons is avoided if we recognize ment than a dyon of magnetic charge is the best test of such consistency.
the additional three-valued magnetic -1/3 and should have a somewhat dif- Faraday
labels and combine them in a totally ferent average spin; this would lead References and Notes
antisymmetric arrangement. This im- to an unacceptable electric dipole mo- 1. P. A. M. Dirac, Proc. Roy. Soc. London Ser.
plies that each of the three magnetic ment, unless a mechanism restores the A 133, 60 (1931); Phys. Rev. 74, 817 (1948).
assignments is equally probable for an equivalence of all dyons by a rapid ex- 2. The term hadron has been introduced in op-
position to lepton, which designates particles,
individual dyon, thus giving an aver- change of magnetic charge which effec- other than the photon and graviton, that do
not have strong interactions. Lepton was well
age magnetic charge of zero. tively destroys the correlation between chosen since the Greek combining form lepto-
A mechanism for magnetic charge spin and magnetic charge. includes "small, weak" among its meanings.
But, unfortunately, the meanings of hadro- are
exchange is also indicated if large vio- The same mechanism for magnetic limited to "ripe, thick," which this is, a bit.
lations of CP invariance are to be avoid- charge exchange will tend to suppress 3. The argument was presented for particles with
either electric or magnetic charge in Sym-
ed. Particularly relevant is the remark- those effects of order ego/hc that were metry Principles at High Energy, A. Perlmut-
able precision with which it is known called fine structure. The exchange ter, J. Wojtaszek, G. Sudarshan, B. Kursuno-
glu, Eds. (Freeman, San Francisco, 1966). Also
that the neutron does not have an elec- mechanism itself produces mass split- noted there is the charge-quantization condi-
tric dipole moment (8). The associated tion for dual-charged particles, under the as-
tings, however. Among the conse- sumption that particles carry both charges e,g
length is measured to be S 10-22 centi- quences of these couplings is a displace- and e,-g. I was not yet ready to face the
apparently strong violation of CP invariance
meter, in contrast with 10-14 centi- ment in the masses of the individual that occurs if only one of these particles exists.
meter for the magnetic dipole moment. dyons. There is a plausible expression 4. Both arguments are presented in the language
of the new theory of sources [J. Schwinger,
We compare this situation with an for the exchange interaction that pro- Phys. Rev. 173, 1536 (1968)]. A discussion that
elementary model in which a dyon, duces a mass splitting of a threefold employed the more conventional and more
cumbersome methods of operator field theory
considered to be a particle with spin of electric multiplet into a doublet and a was given some time ago by T.-M. Yan
1/2, possesses intrinsic magnetic and singlet, which gives an elementary ac- [thesis, Harvard University (1968)]. It has
been duplicated recently by D. Zwanziger
electric dipole moments proportional count of the empirical properties of [Phys. Rev. 176, 1489 (1968)], although this
author does not recognize that two different
to its spin vector and its electric and isotopic spin and hypercharge. These factors of 2 are involved.
magnetic charge, respectively. Such considerations are too quantitative and 5. Comments in this direction were made by A.
Goldhaber [Phys. Rev. 140, B1407 (1965)]. I
models have often been applied to nu- too uncertain to merit further comment came to this approach from the opposite di-
cleon magnetic moments, with the para- here. Suffice it to say that the general rection, by asking how the spin of a particle
could be removed in favor of its helicity, the
doxical result that the constituents have outlines of a mechanism have appeared, spin component along the momentum direction.
masses smaller than the nucleon mass, which may meet the challenge posed The mathematical problem is the same, with
position and momentum vectors interchanged.
rather than the much larger values that by the regularities observed in the 6. Unfortunately, the field of choice is not free
are required physically. This is an arti- of prior incursions. In the interests of an
properties of hadronic strong, electro- obscure literary reference that celebrates the
fact of an overly naive nonrelativistic magnetic, and weak interactions. empirical aspect of triadism, an untraditional
attitude, however, and it is removed and unmellisonant term was introduced and
has found favor in some circles. I prefer to
if the magnetic (and electric) energy is respect tradition and, more important, to
incorporated as an added term in the Summary emphasize the theoretical basis of the other-
wise mysterious empirical characteristics.
relativistic expression for the squared 7. J. Schwinger, Phys. Rev. Lett. 20, 516 (1968).
mass of the composite particle. Then it A conceivable dynamical interpreta- 8. W. Dress, J. Baird, P. Miller, N. Ramsey,
Phys. Rev. 170, 1200 (1968); C. Schull and
is the mass of the composite particle tion of the subnuclear world has been R. Nathans, Phys. Rev. Lett. 19, 384 (1967).