A possible method of specific charge identification from profile measurements in nuclear emulsions

Abstract

This is a discussion of a method of specific charge identification in nuclear emulsions (e.g. 1200 µm thick K-5 pellicles processed unmounted) requiring ∼ 50 µm projected range. Using a constant cell lengtht, the width of the trackTn was measured at each residual rangent, yielding a distribution {Tn}. Typically {Tn} consisted of 100 measurements made at a basic cell lengtht ≈ 0.57 µm. Each track was subjected to about three such measurements in order to estimate the measurement errors. The meanσ and standard spreadσ of this distribution, which are conventional charge sensitive parameters, often require normalizations for best results. A search amongst the dimensionless parameters associated with the distribution {Tn} indicates that the third moment of the distribution,α3 ≡\(\alpha _3 = \sum\limits_n {(T_n - \bar T)^3 /\sigma ^3 } \) (T −σ)3/σ3, is a charge sensitive parameter. A truncated distribution {T′n} with meanσ′, spreadσ′ and skewnessα(pσ′) is derived from {Tn} by replacing allTn >σ′ +pσ′ byσ′ +pσ′.α(pσ′) withp=3.75 is found to be the most sensitive parameter of those tried. It is presumed that its ability to discriminate is based on the presence of «subδ-rays» and the tendency to form gaps. The principal limitation arises at present from the rather large measurement errors. The observed efficiency of discrimination (∼ 80%) is consistent with what would be expected from these errors only. The possibility of utilizingα together with the mean track thickness in order to infer the mass of the particle, is also discussed.