ught that they always pointed to the North Star under the influence
of some stellar compulsion. But even in the fifteenth century it was
noticed independently by Columbus and by German experimenters that the
needle did not point true north. As the amount of its declination
varies at {615} different places on the earth and at different times,
this was one of the most puzzling facts to explain. One man believed
that the change depended on climate, another that it was an individual
property of each needle. About 1581 Robert Norman discovered the
inclination, or dip of the compass. These and other observations were
summed up by William Gilbert [Sidenote: Gilbert] in his work on _The
Magnet, Magnetic Bodies and the Earth as a great Magnet_. [Sidenote:
1600] A great deal of his space was taken in that valuable destructive
criticism that refutes prevalent errors. His greatest discovery was
that the earth itself is a large magnet. He thought of magnetism as "a
soul, or like a soul, which is in many things superior to the human
soul as long as this is bound by our bodily organs." It was therefore
an appetite that compelled the magnet to point north and south.
Similar explanations of physical and chemical properties are found in
the earliest and in some of the most recent philosophers.

[Sidenote: Geography]

As might be expected, the science of geography, nourished by the
discoveries of new lands, grew mightily. Even the size of the earth
could only be guessed at until it had been encircled. Columbus
believed that its circumference at the equator was 8000 miles. The
stories of its size that circulated after Magellan were exaggerated by
the people. Thus Sir David Lyndsay in his poem _The Dreme_ [Sidenote:
1528] quotes "the author of the sphere" as saying that the earth was
101,750 miles in circumference, each mile being 5000 feet. The author
referred to was the thirteenth century Johannes de Sacro Bosco (John
Holywood). Two editions of his work, _De Sphaera_, that I have seen,
one of Venice, 1499, and one of Paris, 1527, give the circumference of
the earth as 20,428 miles, but an edition published at Wittenberg in
1550 gives it as 5,400, probably an {616} attempt to reduce the
author's English miles to German ones. [Sidenote: 1551] Robert
Recorde calculated the earth's circumference at 21,300 miles.[2]

Rough maps of the new lands were drawn by the companions of the
discoverers. Martin Waldseemuller [Sidenote