get rid of in case of sickness or breakdown of any
sort. The work is paid for when furnished, and the main trouble of
administration saved. It is only necessary to have a brief report
from time to time, showing that the work is actually going on.

I began with a careful examination of the relation of prices to work,
making an estimate of the time probably necessary to do each job.
Among the performers of the annual work were several able and eminent
professors at various universities and schools. I found that they
were being paid at pretty high professional prices. I recall with
great satisfaction that I was able to reduce the prices and, step by
step, concentrate all the work in Washington, without detriment to
the pleasant relations I sustained with these men, some of them old
and intimate friends. These economies went on increasing year by
year, and every dollar that was saved went into the work of making
the tables necessary for the future use of the Ephemeris.

The programme of work which I mapped out, involved, as one branch of
it, a discussion of all the observations of value on the positions
of the sun, moon, and planets, and incidentally, on the bright fixed
stars, made at the leading observatories of the world since 1750.
One might almost say it involved repeating, in a space of ten or
fifteen years, an important part of the world's work in astronomy for
more than a century past. Of course, this was impossible to carry out
in all its completeness. In most cases what I was obliged practically
to confine myself to was a correction of the reductions already made
and published. Still, the job was one with which I do not think
any astronomical one ever before attempted by a single person could
compare in extent. The number of meridian observations on the sun,
Mercury, Venus, and Mars alone numbered 62,030. They were made at the
observatories of Greenwich, Paris, Koenigsberg, Pulkowa, Cape of Good
Hope,--but I need not go over the entire list, which numbers thirteen.

The other branches of the work were such as I have already
described,--the computation of the formulae for the perturbation of
the various planets by each other. As I am writing for the general
reader, I need not go into any further technical description of this
work than I have already done. Something about my assistants may,
however, be of interest. They were too numerous to be all recalled
individually. In fact, when the work was at