STUDY OF $\pi$-$\mu$-$e$ DECAY USING SCINTILLATION COUNTERS

The work under review¹ represents one of the first applications of scintillation counters to the study of $\pi$-meson decay. It is known that $\pi$-mesons are unstable—their mean lifetime is about $10^{-8}$ sec. In the decay of a $\pi$-meson at rest, whose mass, according to the most accurate measurements, is equal to $276\,m_e$, a light neutral-

...particle and an ordinary \(\mu\)-meson of mass \(210\,m_e\), possessing a kinetic energy of about \(4\,\text{MeV}\) and, correspondingly, a range of about \(0.1\,\text{g}/\text{cm}^2\) (\(\pi\)-\(\mu\) decay). The \(\mu\)-meson produced in \(\pi\)-\(\mu\) decay is also unstable. It decays into a positron (or an electron, depending on the sign of the \(\pi\)-meson charge) and two neutral particles.

Fig. 1. Principle of measuring the lifetime of a resting \(\mu\)-meson.

Let us recall that the measurement of the \(\mu\)-decay time, first carried out by Rossi and Nereson* with \(\mu\)-mesons from cosmic rays and repeated by many investigators, consisted in principle of the following. Some fraction of the fast \(\mu\)-mesons, entering from the air into plate \(P\) (Fig. 1), is stopped in it. The slowing-down process occurs very rapidly—in a time of the order of \(10^{-12}\,\text{sec}\). The slowed-down meson is then either captured by the nucleus or decays. A meson entering the plate causes one of counters \(I\) to operate, while the decay particle emerging from the plate causes one of counters \(II\) to operate. By measuring the time interval between two successive pulses in counters \(I\) and \(II\), for a large number of decay events Rossi and Nereson obtained for the mean lifetime of \(\mu\)-mesons the value

\[ \tau_\mu = 2.15 \pm 0.07\ \mu\text{sec}. \]

Fig. 2. Diagram of the experiment for studying \(\pi\)-\(\mu\)-\(e\) decay.

The application of such a method to measuring the time of \(\pi\)-\(\mu\) decay encounters difficulties connected, first, with the fact that ordinary Geiger–Müller counters are not inertialess detectors of charged particles passing through them: they have a certain delay in the onset of the discharge relative to the moment at which the ionizing particle passes through the counter, a delay which, as a rule, is greater than the lifetime of a \(\pi\)-meson. Secondly, the small range of \(\mu\)-mesons arising in the decay of \(\pi\)-mesons makes their detection difficult. Therefore the first measurements of the lifetime of \(\pi\)-mesons generated in an accelerator\(^{3,4}\) were carried out with the aid of photographic plates. The idea of these experiments, described in detail at the Rutherford Laboratory\(^{5}\), consisted in measuring the relative attenuation of the intensity of beams of monoenergetic \(\pi\)-mesons traveling along different paths. These measurements gave, for the mean lifetime of \(\pi^+\)-mesons, the value

\[ \tau_{\pi^+} = \left(2.97^{+0.14}_{-0.17}\right)\times 10^{-8}\ \text{sec}, \]

and for the mean lifetime of \(\pi^-\)-mesons the value

\[ \tau_{\pi^-} = \left(1.11^{+0.31}_{-0.22}\right)\times 10^{-8}\ \text{sec}. \]

The use of scintillation counters, which do not have the delay in the beginning of the pulse characteristic of Geiger–Müller counters, makes it possible to implement the same principle for determining the time of \(\pi\)-\(\mu\) decay that Rossi used for measuring the time of \(\mu\)-decay. The experimental arrangement is shown in Fig. 2, where \(T\) is a hydrogen target on which a beam of \(\pi\)-quanta with a maximum energy of \(315\) MeV is incident. Some of the \(\pi\)-mesons born in the target pass successively through absorber \(A_1\), scintillation counter \(A_2\), and are stopped in the second scintillation counter \(X_2\). Practically all \(\mu\)-mesons arising in the decay of \(\pi\)-mesons stopped in the last crystal remain in it and, decaying with a lifetime of \(2.15\) \(\mu\)sec, form positrons. The entry of a \(\pi^+\)-meson into the second crystal, the appearance of a \(\mu^+\)-meson in the decay of the \(\pi^+\)-meson that has occurred in the second crystal, and the appearance of the decay positron cause proportional scintillations in the second crystal, registered by photomultiplier \(PM_2\). The pulses appearing at the output of \(PM_2\) are registered by a fast amplifier and are fed through a delay line to a fast oscilloscope. The total resolving power of the electronic circuit and the crystal makes it possible to begin measurements of times from \(3 \times 10^{-8}\) sec. Pulses from \(PM_2\) are also fed to a second, slower oscilloscope (a double tube is used), whose purpose is to register the pulses produced by the positron in \(\mu\)-decay. The sweeps of both oscilloscope beams are triggered by the coincidence of pulses in photomultipliers \(PM_1\) and \(PM_2\). During 12 hours of operation of the apparatus (the synchrocyclotron operates with 100-microsecond pulses at a repetition frequency of \(0.5\) sec), about 2500 oscillograms were taken, on 100 of which two adjacent pulses are visible. The authors examine a group of 57 photographs on which three pulses satisfying the following requirements are visible.

The distance between the first two pulses obtained on the fast sweep lies within \((3 \div 8)\cdot 10^{-8}\) sec. These two pulses are accompanied by a third pulse, recorded on the slow sweep and lying within \(5\)—\(7\) \(\mu\)sec, which is interpreted as the pulse from the decay positron. Owing to this selection, these photographs represent photographs of pulses arising in a double decay,

\[ \pi^+ \to \mu^+ + n. \]

\[ \mu^+ \to e^+ + 2n, \]

where \(e^+\) is a positron, and \(n\) is a light neutral particle.

Evidence for the correctness of such an interpretation is the fact that the distribution in time of the third pulses, lying within \(5\)—\(7\) \(\mu\)sec, gives the correct value of the \(\mu\)-decay time \(\tau_\mu = 2.5\) \(\mu\)sec. In Fig. 3 the distribution of 57 time intervals between the first two pulses is given. The lifetime of the \(\pi^+\)-meson, determined from the slope of this curve, is equal to

\[ \tau_{\pi^+} = (1.65 \pm 0.33)\cdot 10^{-8}\ \text{sec}. \]

The question of the lifetime of \(\pi^+\)- and \(\pi^-\)-mesons is of great importance for the following reason: from works 3 and 4 it is known that these times are different. If this fact is confirmed by further measurements, then \(\pi^+\)- and \(\pi^-\)-mesons cannot be regarded as identical particles, differ-

differing only in the sign of the charge, and one should expect differences in their properties. Unfortunately, the error of this measurement is too large, and the value obtained, \((1.65 \pm 0.33)\cdot 10^{-8}\) sec, does not contradict both values of the lifetime \(\tau_{\pi^+}\) and \(\tau_{\pi^-}\) obtained in works 4.

Fig. 3. Distribution of time intervals between two pulses.

V.

CITED LITERATURE

1. W. L. Kraushaar, J. E. Thomas, J. and V. P. Henri, Phys. Rev. 78, 486 (1950).
2. Rossi and Nereson, Phys. Rev. 64, 199 (1943).
3. Martinelli and Panofsky, Phys. Rev. 77, 465 (1950).
4. Richardson, Phys. Rev. 71, 1720 (1948).
5. UFN 41, issue 2, 219 (1950).