mainly the Greeks and Arabs) had had to depend for fixing relative
positions on the vague statements in the itineraries of merchants and
soldiers; but now, with the aid of the compass, it was not difficult
to determine the relative position of one point to another, while
all the windings of a road could be fixed down on paper without
much difficulty. Consequently, while the learned monks were content
with the mixture of myth and fable which we have seen to have formed
the basis of their maps of the world, the seamen of the Mediterranean
were gradually building up charts of that sea and the neighbouring
lands which varied but little from the true position. A chart of
this kind was called a Portulano, as giving information of the
best routes from port to port, and Baron Nordenskiold has recently
shown how all these _portulani_ are derived from a single Catalan
map which has been lost, but must have been compiled between 1266
and 1291. And yet there were some of the learned who were not above
taking instruction from the practical knowledge of the seamen.
In 1339, one Angelico Dulcert, of Majorca, made an elaborate map
of the world on the principle of the portulano, giving the coast
line--at least of the Mediterranean--with remarkable accuracy. A
little later, in 1375, a Jew of the same island, named Cresquez,
made an improvement on this by introducing into the eastern parts
of the map the recently acquired knowledge of Cathay, or China,
due to the great traveller Marco Polo. His map (generally known as
the Catalan Map, from the language of the inscriptions plentifully
scattered over it) is divided into eight horizontal strips, and on
the preceding page will be found a reduced reproduction, showing how
very accurately the coast line of the Mediterranean was reproduced
in these portulanos.

With the portulanos, geographical knowledge once more came back to