34 MAPS AND THEIR MAKERS 



confined waters and spread across the great oceans, observa- 

 tions for latitude served as a check on dead reckoning. As a 

 further consequence, the sphericity of the earth's surface 

 could no longer be overlooked, and the problem of choosing 

 a projection which would allow a line of constant bearing to be 

 represented on the chart by a straight line became acute. The 

 solution finally achieved is the projection which bears the 

 name of Mercator. 



For the reasons set out above, it seems clear that the 

 portolan charts were from the beginning closely related to the 

 compass, and that it was the introduction of this instrument 

 which made their construction possible. Some students, 

 however, have held contrary views. Professor Wagner, in his 

 study of the scales, equated the shorter mile used in the 

 Mediterranean to an ancient Italian-Greek unit, and that used 

 on the Atlantic coasts with the later Roman mile; from this 

 he argued that the Mediterranean portion of the chart must 

 date back to a period long before the introduction of the 

 compass. Though no chart has survived from Roman or 

 Greek times, a few books, or portions of books, of sailing 

 directions are extant, and it is probable that details from these 

 were incorporated in the portolan charts. 



As for the portolano or book of sailing directions, known 

 later in England as a 'rutter of the sea', it is true that there is 

 evidence for its existence in medieval times, before the 

 introduction of the chart. The oldest known example is con- 

 tained in Adam of Bremen's 'Ecclesiastical History', written in 

 the twelfth century. This has the appearance of being a very 

 summary version of a more detailed document. In a few lines 

 it sets out the stages on a voyage from the mouth of the river 

 Maas to Acre in Palestine. It gives the distances from point to 

 point by the number of days' sailing, with an approximate 

 indication of the direction to be followed. The only point 

 mentioned on the English coast is Tral', perhaps Prawle 

 Point, or Portland Bill. Unless much more detailed instructions 

 existed, it is difficult to imagine the portolan charts being 

 produced from material of this type. At the very least, the use 

 of the compass would simplify the work greatly and give 

 much improved results. Others have argued that the method 



