Gilt brass full-plate fusee movement of ‘Liverpool runner’ configuration, 22 size, with set-up on barrel-bridge, signed Josh Johnson, 25 Church St 30309 on back-plate and LIVERPOOL on barrel-bridge. Standard English single-table side-lever escapement. Split-rim bimetallic balance with timing and poising screws. Parallel-sided cock with fenestrated table, engraved with swags. Jewelled to centre-wheel (19 jewels), with Liverpool windows in back-plate. Gilt brass dust-cap with cut-out for set-up. Off-white dial with roman chapter-ring and early form of recessed seconds panel in which the numerals are placed outside the recessed area. Gilt Breguet-type hour and minute hands and blued-steel second hand (all non-original). Silver consular case, marked for [Mary] Samuel & [Arthur Guinness] Rogers of 72 Wood Street, Liverpool, and hallmarked Chester 1861.

This relatively late Joseph Johnson lacks much of the elegance of the 1820s; the gilding is dull (probably it was done by electroplating rather than by the old mercury process) and the lettering and cock-engraving are perfunctory. However, it remains a fine, if much-used, example of the traditional English fusee lever at a time when this was not yet declining into obsolescence. Although the case was clearly made specifically for the movement (it bears the same number), I believe that the latter must have already been five or six years old when it was cased, since I have a record of Johnson No. 32478 whose case is dated 1857. Jewelling the centre wheel, whose thick arbor and slow rate of turning would usually be considered enough to assure it against serious wear, is quite unusual in Britain; when it was applied, it was normal to put a hole in the cock-table (as in this example) so that the centre jewel could be seen. It was also regarded as acceptable to announce the number of jewels on the dial, a thing which no British watchmaker would ever do if they were fewer than 19; however, the makers have not asserted this privilege here. The hands are appropriate for the period, although all are a little too short.

At the time of writing (29 May 2020) I carry this watch for everyday use and have been keeping a daily record of its performance for seven weeks. It has never deviated by more than 18 seconds in a day, and on all but eleven days the error per day has been below ten seconds. I have adjusted the minute-hand five times during that period (moving it forward or back by one minute when the cumulative error exceeds 30 seconds) and have not touched the regulator since 8 May. This is of course far below a ‘railroad’ standard of performance, but it is highly creditable for a pre-industrial English watch with no pretensions to be regarded as a chronometer or a candidate for testing at Kew; it is the only watch I have ever owned whose accuracy is close enough to justify maintaining such an analysis. Moreover, the watch is unrestored and has probably not been professionally overhauled for decades.