NavList:

In 1945 Popular Astronomy magazine ran a three-part series by Raynor L.
Duncombe on the history of the "personal equation" in observations with
transit instruments. He begins with the case of the Greenwich
Observatory assistant who persistently observed meridian transits .5 to
.8 second late relative to Maskelyne, who fired the man in 1796 for his
"vicious" habit.

However, by the early 19th century there was growing awareness of
systematic individual differences in "eye and ear" transit timings by
experienced observers. (The observer listens to the ticks of the clock
and keeps mental count of the seconds while watching the star cross the
vertical wire of the telescope.) For instance, a series of tests showed
a 0.8 to 1.0 second offset in the transit timings of eminent astronomers
Bessel and Struve.

Development of the electrical art gave an alternative to eye and ear:
the recording chronograph, in which a pen traced a helical path on
paper, wrapped around a cylinder and rotated at constant speed. An
electromagnet wired to contacts in a clock deflected the pen once per
second. It also recorded the observer's tap on an electrical key at the
moment of transit.

This was an improvement. Early tests of personal equation were purely
relative: one man with respect to another. But now absolute measurements
were possible, via an apparatus which drove an artificial star across
the field of view and recorded the true crossing of the central wire vs.
the observer's key tap. Thus each observer's work could be made free of
his personal equation.

At least, that was the theory. But the growing body of chronograph data
began to show a systematic error as a function of declination in the
work of some observers, but not others. The fault lay in the lack of
standardization in the tapping technique. Some observers were proactive
— they thought it proper to anticipate the instant of transit and begin
the muscular impulse early so the tap occurred exactly on the wire.
Others were reactive. They waited until the star was on the wire, then
tapped. Of course this tap was always late, but reaction time would be
included in that observer's personal equation.

It turned out the latter gave more consistent results. Proactive tappers
had a tendency to develop a timing optimized for one particular
declination. At any other declination the star would move at a different
rate, and despite the observer's best effort the tap would be early or late.

The next weapon against personal equation was the impersonal Repsold
micrometer. In this the wire was movable, driven by a screw operated by
the observer, who tracked the star continuously. A disc with electrical
contacts was attached to the screw. This transmitted pulses to the
chronograph. In some cases a motor rotated the screw, and the observer
controlled its rate.

With this device the transit instrument attained unprecedented accuracy.
In fact, even recent JPL ephemerides depend in part on transit
observations of the outer planets as far back as 1911. (This coincides
with the introduction of the impersonal micrometer at the USNO.)

Nevertheless, eye and ear is still the way we check a clock with a
shortwave radio. When I find a time page on the Web dead accurate (or
not) with respect to ticks from my radio, I wonder how much personal
equation is present. Some of the values measured in the heyday of eye
and ear are almost unbelievably large.

http://adsbit.harvard.edu/cgi-bin/nph-iarticle_query?bibcode=1945PA.....53....2D&db_key=AST&page_ind=0&plate_select=NO&data_type=GIF&type=SCREEN_GIF&classic=YES

http://adsbit.harvard.edu/cgi-bin/nph-iarticle_query?bibcode=1945PA.....53...63D&db_key=AST&page_ind=0&plate_select=NO&data_type=GIF&type=SCREEN_GIF&classic=YES

http://adsbit.harvard.edu/cgi-bin/nph-iarticle_query?bibcode=1945PA.....53..110D&db_key=AST&page_ind=0&data_type=GIF&type=SCREEN_VIEW&classic=YES