for navigators to become
familiar with the proposed change in time-reckoning before they were
called upon to employ it in their calculations.

I believe that they would soon come to think it more convenient and
natural to reckon according to civil time than according to the
present astronomical time. I am told that this practice is already
universally adopted in keeping the log on board ship. To avoid any
chance of mistake, it should be prominently stated on each page of the
ephemerides that mean time reckoned from mean _midnight_ is kept
throughout.

Whether or not astronomers agree to adopt the civil reckoning, I think
we ought to adopt the instant of midnight on the initial meridian as
the commencement of the universal day.

The relation between the local time at any place and the universal
time would then be expressed by the simple formula:

Local time = universal time + longitude.

Whereas, if the proposition of the Roman Conference were adopted, we
should have to employ the less simple formula:

Local time = universal time + longitude - 12 hours.

In recommending the mean noon at Greenwich as the commencement of the
universal day and of cosmopolitan dates, the Roman Conference refers
to this instant as coinciding with the instant of midnight, or with
the commencement of the civil day, under the meridian situated at 12
h. or 180 deg. from Greenwich. Now, this reference to the civil day and
date on the meridian opposite to Greenwich appears not only to be
unnecessary and to be wanting in simplicity, but it may also lead to
ambiguity in the date, as expressed in universal days, unless this
ambiguity be avoided by making an arbitrary assumption. No doubt the
Greenwich mean noon of January 1 coincides with midnight on the
meridian 12 h. from Greenwich, but with what midnight. What shall be
its designation and the corresponding date given to the universal day?
Shall we call the instant above defined the commencement of the
universal day denoted by January 1 or by January 2? Each of these
dates has equal claims to be chosen, and the choice between them must
clearly be an arbitrary one, and may, therefore, lead to ambiguity.

By adopting Greenwich mean midnight as the commencement of the
universal day, bearing the same designation as the corresponding
Greenwich civil day, all ambiguity is avoided, and there is no need to
refer to the opposite meridian at all.

Those are the ideas I wish to express with regard t