True solar time

Clocks are set by zone, the sun is not

Everyone inside one time zone reads the same clock, but the sun does not arrive overhead at the same instant across the whole width of that zone. A true solar time correction is the step that carries a clock reading over to the reading the sun would give at the exact spot where the birth was recorded.

The size of the step follows from one number. A clock is set to one meridian, and every degree of longitude away from that meridian is worth four minutes. A place five degrees east of the meridian its clock is set to is twenty minutes ahead of that clock as far as the sun is concerned.

Because the step moves the instant itself, and not just the hour column, it can reach any of the four columns. Most of the time only the hour column is close enough to a boundary for a step of that size to show, which is why the correction is usually talked about as an hour pillar question.

Why this site leaves it switched off

The site default does not apply the correction. That is a data problem and not a judgement about the practice: applying it needs a longitude for the place of birth, and the table of representative longitudes this site would read from has not been collected. Nothing is filled in to stand in for it.

The screen says the same thing where the setting is shown, so a reader does not have to come to this page to find it out. What this page adds is the reason the table is missing rather than wrong: it is an open item with a named source to collect it from, and the setting flips over once that is done.

While it is off, the result table still carries two rows for it. The correction in minutes comes out as zero, and the reference meridian it would have measured from comes out as null, marked the way every other undecided value on this site is marked.

Switching it on, and the two rows to read afterwards

The correction can be switched on from the calculator. There is an optional longitude field above the settings, and the third of the three settings chooses whether a correction is applied at all. Switching it on without a longitude stops the calculation and says so; no longitude is estimated on a reader's behalf.

With a longitude in place, two rows in the result table fill in. One gives the correction that was applied, in minutes. The other gives the reference meridian it was measured from, in degrees east. The subtraction between the two is the one described above, so the number can be checked against the screen rather than taken on trust.

A worked case: a birth recorded at 11:30 in New York on June 15, 1990, with a longitude of 74 degrees west, comes out as a correction of minus fifty-six minutes measured from a reference meridian of 60 degrees west. Moving the instant back by that much lands the hour column in the neighbouring one of the twelve periods.

The same place on two dates gives two different numbers

The correction is not a fixed figure attached to a place. It is measured against the meridian the clock at that place was actually set to on that date, and in a country that moves its clocks twice a year that meridian moves with them.

The same New York birthplace and the same longitude give a correction of minus fifty-six minutes for a June 1990 birth and plus four minutes for a January 1990 birth, measured from 60 degrees west and 75 degrees west in turn. An hour of daylight saving is the whole of the difference, and the result table prints a row saying whether it was in force at that date and place.

That is also why the longitude field alone is not enough to fix the figure. The time zone chosen for the birth is what supplies the meridian, so the two fields work as a pair, and both of them are read off the screen afterwards rather than assumed.