In the table given above will be found a few cases in which the error of collimation is inconveniently large. These arose from erroneous settings of the transit-micrometer, not discovered till the end, or nearly so, of the observations of the evening. At the commencement RADCLIFFE OBSERVATIONS, 1862.

d

of the year the errors of level and azimuth had been (as can be seen by inspection) made very small, but, owing perhaps to the recent erection of the piers, a gradual increase in the error of level took place, till it became inconveniently large. It is probable also that some disturbance took place by the hammer of the mason at the end of April, when the chronographic clock was placed on the eastern pier. Still there seems to have been a tolerably uniform increase of level till September or October, after which the tendency was rather to decrease. I preferred to leave it unaltered till the end of the year, that the law of its changes might be more distinctly seen; and, after that time, it was reduced, and remained during the year 1863 tolerably constant. Occasionally there are some startling changes of azimuth (much greater than in succeeding years), and, if these are not due to errors of observation, they would tend to show that the piers had not acquired sufficient steadiness.

Clock-Error.-The assumed mean right ascensions of the stars used for determining clock-error were, in the year 1862, derived from a list kindly furnished to me by the Astronomer Royal (designated in a preceding table "The Greenwich List"), which is identical with that printed at pages xxxv and xxxvi of the Greenwich Observations for 1862, and is fundamentally based on the Greenwich Seven-Year Catalogue of 2022 Stars for the Epoch 1860, recently published. The equinox to which all objects are referred is therefore identical with the Greenwich equinox for this year, and, for convenient reference in a matter fundamentally affecting the right ascensions, it seems desirable to reprint it here, only premising that the right ascensions of Polaris, Cephei 51 (Hev.), 8 Ursa Minoris, and λ Ursa Minoris, also included in the Greenwich List, have been given already in the table printed on pages x and xi.

Assumed Mean Right Ascensions, for 1862, Jan. 1, of Stars used for determination of Clock-Error, with the corrections to the Nautical Almanac places for Stars contained in that Work.