[8. B. DAWSON] THE VOYAGES OF THE CABOTS 161 



then, to assume that, wherever the magnetic pole of that period may have 

 been situated, the magnetic curves bore, in a general way over large 

 areas, relations to each other similar to those existing at the present time. 

 The variation now, at the admiral's point of observation in 1492, is 20^ W.^ 

 and the variation at Cape Race is 30° W., and it may therefore be fairly 

 argued a ^)/*/,ori that the relations between the quantities would be the 

 same at that date. We are not, however, shut up to an argument a pr-iori 

 alone. We have actual obsei'vations extending back for 250 years, and 

 it was on these chiefly that I based my opinion that the average variation 

 in Cabot's time over that part of the Atlantic traversed by him was one 

 point and a half, and some details of these observations were given in my 

 first paper. I then pointed out (at p. 69) that Reinel's chart of A. D. 

 1505 showed plainly upon it, by its double scale, a variation on the New- 

 foundland coast of nearly two points. If, in my first paper, I had given 

 Mr. Charles Schott's map of the North Atlantic, drawn to accompany 

 Capt. Fox's paper, this discussion might have been avoided. I give it here- 

 with, and, for the sake of clearness, to avoid wanderi-ng over the whole 

 field of terrestrial magnetism, I have eliminated all magnetic lines, saving^ 

 the line of no magnetic variation of A.D. 1492 to AD. 1500, the period 

 now in question. 



Mr. Schott did not put forth this map as absolutely, but as approxi- 

 matel}^ correct, so far as data existed and the laws of terrestrial magnet- 

 ism were ascertained. For the purposes of a general argument like mine, 

 an apj)eal to the uniformity of nature in connection with the relative- 

 positions of the magnetic curves of the present day would perhaps have 

 been sufiicient ; but I am glad here to bring under the attention of the 

 reader the results reached by a scientific oflScer of high repute in this 

 very difliicult field, particularly as it enables me to show with more pre- 

 cision the place where Cabot crossed into westerly variation. 



And now, if we examine this map, we will see the course of Columbus 

 from Palos to Gomera. There he got upon the parallel indicated by 

 Toscanelli as that of Zipango and Southei'n Cathay. From Gomera he 

 started on his western course as marked out by a due west line dotted on 

 the map. That course, near the meridian of longitude 30° W., crosses a 

 curved line extending northeastwards across the ocean. This is the line 

 of no variation (agonic line), and when Columbus crossed it his needles 

 " wested." Farther on, about longitude 40°, he noticed a " westing " of 

 a full point, and there the straight course on the map stops. The erro- 

 neous datum which vitiates Mr. Harrisse's mathematical formula? is, that 

 he counts the whole course as 3,150 miles from Gomera ; whereas the 

 length should be counted, for the purpose of this argument, from the 

 point where the disturbing influence first began to act. 



From the map it will be seen that the agonic line extends to the 

 northeast, as also do all the magnetic curves upon the present charts ; 



Sec. II., 1897. 8. 



