Department of
Science & Technology Studies
University College London

Nicholas Kollerstrom's
Newton's 1702 Lunar Theory


Halley's Error Values (1732)

Though Halley believed his version of TMM to be more accurate that Newton's, my reconstruction shows it to be LESS accurate. This graph shows data compiled by Halley as Astronomer Royal while routinely estimating the TMM error-values in the summer of 1732. The separate points on the graph are the values he recorded, once per two or three days, for each of his lunar meridian readings. One notes the baseline as having drifted by two arcminutes, as a cumulative error in the Newtonian values.

Halley's Data

In Halley's table, the far right-handed column shows "Halley's Computed Errors from Observation" beginning June 21st." Use the Flamsteed replica to check the results.

The 1733 graph shows how, a year later, maximal errors of eight arcminutes started to turn up. Alas we lack any comment from Halley as he faithfully recorded these errors. Actually, Halley's Saros method could have coped quite well with this, as very similar error-values appeared eighteen years and eleven days earlier. 


The contents of this page remain the copyrighted, intellectual property of Nicholas Kollerstrom.  Details. rev: May 1998