Further investigations with a Wilson Chamber. III.—The accuracy of the angle determination
I.—The question of the precision of the determination of the angles of forked tracks is of considerable importance, in particular owing to the possibility of determining nuclear energy levels from Wilson photographs. In an earlier paper some experiments were described in which an artificial track consisting of a bent glass fibre was photographed in different positions. The average error of determination of the angle was found to be 10 minutes of arc. This error was attributed to the lack of perfect adjustment of the camera. That the error of measuring actual tracks could be nearly as small as this was shown by measurements of three collisions in which the difference between the calculated and expected mass ratios was consistent with a probable error of about 10 minutes of arc for the angle measurements. Only such forks were used for these calculations of the mass ratio, for which the three arms appeared unusually straight and for which the test for coplanarity was accurately satisfied. Subsequently two collisions with hydrogen nuclei were described in which the error of the angles was held to be as low as 6 minutes of arc. It was pointed out at the time that many tracks did not, in fact, satisfy these conditions, but sufficient data were not then available for a statistical analysis of the distribution of calculated mass ratios, from which a reliable estimation of the probable error of measurement of an average fork could be made. Since then a great many more photographs have been taken with a larger chamber and an improved camera and such a statistical test is now possible. To test the camera itself five photographs were taken of two black lines ruled on a card. The angles calculated from the photographs were:—