the instrument; whilst the proportion applicable to each star was indicated by the table purposely prepared for the latitude of Hartwell.

Making due allowance for the lapse of time between the transits, and then applying it to the clock-errors, the difference between the high and low star is divided by the tabular azimuth difference between those two polar distances. The quotient thus obtained is multiplied by the tabular number, and the result is applied to the clock-error. Finally, the difference in this last column from the clock-error resulting from the selected index-stars, indicates the modification required in the place before assigned to the unknown star. For example

Ex. I. Cygni gives clock-error too small; its assigned AR. being too small :-

Examples of correcting AR. of unknown stars, by two Greenwich stars being assumed as
known, a Cyg, and a Capric.

Ex. II. Aqui. gives clock-error too large; its assigned AR.
being too large :-

In the following Catalogue of the Hartwell Stars, the left-hand page contains the copy of Mr. Epps's registered observations, forming, as it were, the raw material of the right-hand or opposite page, where the results of reduction are given, and the difference of those results from the well-known Catalogue of the Royal Astronomical Society; the whole being brought up to the epoch 1850 by Mr. Baily's co-efficients. I have given weights, 0 to 3, on the ultimate determinations, as shewing the comparative value which an investigation of them produced.