10 October 2012

Almost a year ago I wedged my compass in the corner of my window sill against the aluminium window channel and the wooden casing and left it there for almost two weeks. That is a location where it would not be disturbed and there is nothing close to it that would cause a change in magnetic fields.  I then tried to take azimuth readings every hour.  Unfortunately I had to do some things like sleep and go out to buy groceries so didn't get an unbroken string of hourly readings during the period or indeed during any one day during the period.  But I put all my readings into a Numbers spreadsheet and massaged and graphed them.  Here are the results.

Now I am reading North Pole, South Pole: The Epic Quest to Solve the Great Mystery of Earths Magnetism by Gillian Turner (ISBN 978-1-61519-031-7).  I am just finishing up Chapter 4, The Wandering Compass Needle, which deals with deviation.  (I have used topographic maps since I became a Boy Scout back in the dark ages and never new until tonight that Edmond Halley invented the contour line.)  At the end of the chapter is a section on just the phenomenon I observed.

George Graham was a London-based maker of precision instruments.  He made clocks, watches, quadrants and some of the most stable and sensitive compasses available, so in 1722 he was concerned to find that the needles of his best compasses were actually in constant, seemingly random motion.  He went back and checked meticulously and found that:

. . . all the needles I made use of would not only vary in their direction on different days, bat at different times on the same day . . .

Sometimes the needles fluctuated by more than half a degree in a day.

To be sure he was not imagining things, Graham set up three compasses.  Each had a needle 31 centimeters long, filed to a fine edge at the end.  These needles moved over a scale that was graduated at intervals of one-sixth of a degree, or ten minutes.  subdivisions of two minutes could be estimated with the use of a magnifying glass.

All three compasses behaved the same way.  Graham had discovered the daily variation of Earth's magnetism.  This was quite different from the secular variation--at least ten times smaller in amplitude and something like 10,000 or 100,000 times more rapid.  Further investigation showed that the average amplitude of these daily variations was greater in summer than in winter: in London it was 13 minutes in summer, as opposed to seven in winter.

In Uppsala, Sweden, at a latitude of 60° N and 18 degrees further east than London, the astronomer Anders Celsius and his assistant, Olof Hiorter, had also noticed and become interested in the daily fluctuations of the compass needle.  They coordinated their measurements with Graham, and between them discovered two features.  First, the regular daily variation correlated with local time, beginning shortly after sunrise and quieting down with sunset.  Secondly, on certain days the needle seemed to go crazy.  On April 5, 1741, for example, Hiorter reported that "the needle at 2 PM began to be disturbed, so that at 5 PM it was 1°40' to the west of it's declination at 10 AM," while Graham recorded:

The alterations that day were greater than I have ever met with before . . . The observations on the other days were with the same care, but they were much less and more regular.

These irregular disturbances were truly simultaneous at London and Uppsala, and were not a function of local time.

Following a hint Halley had given twenty years earlier, Hiorter and Celsius hit on the explanation: at the times when the compass needle was most disturbed, the northern lights, or aurora borealis, were putting on their most spectacular show.  The two scientists wrote:

a motion of the magnetic needle has been found that deserves the attention and wonder of everyone.  Who could have thought that the northern lights . . . when they draw southwards . . . could within a few minutes cause considerable oscillations of the magnetic needle through whole degrees?

Gilbert's explanation of declination may have been disproved, but his chief argument--that the major source of Earth's magnetism lay within the Earth itself--had seemed incontrovertible.  However, these new observations seemed to again point to a source outside the Earth: the daily movement of the compass needle was obviously related to local time, beginning at sunrise and ending at sunset, while the bigger disturbances that were simultaneous in London and Uppsala clearly correlated with auroral activity.  Hiorter reasoned that the auroras must occur extremely high in the atmosphere.  Here then, were magnetic variations seemingly equally associated with processes in the heavens.

Was the source of Earth's magnetism internal or external, or were these perhaps two sources?  How could this apparent dilemma be reconciled?

So I was born 290 years late.

My data do not show the increase in variation beginning at sunrise and a slackening at sunset.  But I am sure their equipment and procedures were better than mine.  My compass swings are a degree or more in an hour and through many degrees in a day.  I recognized when I was making my observations that the location of my compass was a little problematic since it was sometimes hard to get my head lined up properly with the north end of the needle.  And I was moving the compass card every hour to align it with the needle and there is some slop in that compass so the card wasn't aligned the same way with the base on any two consecutive observations.  They too I was eyeballing everything and my eyeballs are not what they used to be.  So I decided that I wanted to do the experiment again using an Arduino and a compass module to log the compass heading every minute over a much longer period of time.  This and the availability of a new, inexpensive Arduino give me further impetus to proceed with the project.