160 ROYAL SOCIETY OF CANADA 



method. He took the following sentence from my paper of 1894 as his 

 text : " If Columbus, on a direct western couree, dropped 240 miles from 

 " Gromera with a variation of one point west, it is altogether probable 

 " that John Cabot, with a variation of a point and a half, would have 

 '• dropped about 360 miles to the south on his western course across the 

 " Atlantic." To that Mr. Harrisse replies : " Yes, it is probable that then 

 " Cabot would have dropped about 360 miles, provided his course had 

 " been precisely of the same length as the course of Columbus." 



The objection is exceedingly well put, and does not require any 

 mathematical support. The course of Columbus from Gromera to Wat- 

 ling's island is usually estimated at 3,150 miles, and Mr. Harrisse has put 

 that figure in his formulae, but in doing so he vitiated the whole calculation 

 ah initio ; for it is not the total length of the course which is the prime 

 factor here, but the length of the course which was subject to the dis- 

 tui'bing element — to wit, the westerly magnetic variation acting upon 

 and modifying both. 



Perhaps I failed in clearness in not saying expressly that my argu- 

 ment was not a pari but a fortiori ; still, on looking again over the 

 paper, that idea seems to pervade it. At p. 59 I contrasted those very 

 differences of condition which my critics urge. In the one case the steady 

 trade wind astern — the smooth seas and the fair weather ; and, in the 

 other, the variable winds and heavy seas of what I called " the moststorm- 

 " tossed region in the world of ocean." I spoke of fogs, and made express 

 mention of the Arctic current, and estimated its average rate correctly, 

 according to the Admiralty Sailing Directions, atone mile an hour, show- 

 ino- that I had taken these conditions of difference into account and that 



to 



the increment of variation was one only of several influences tending to 

 draw Cabot to the south. The sentence being quoted from the portion 

 of my paper devoted solely to the consideration of variation, hardly ex- 

 presses the full scope of the argument. It was not intended to be, and 

 could not be, an argument in the least degree amenable to mathematical 

 treatment, and when six hypotheses and two erroneous quantities are 

 put by my critic into mathematical costume and treated by mathematical 

 methods, it is no wonder that the results do not tally with the facts. I 

 am glad, however, to have the opportunity of going over the subject 

 again. 



Although the observations of Columbus in 1492 give a firm datum, 

 the argument, in the main, rests iipon the uniformity of the laws of 

 nature, by which we are led to assume that in whatever way the mag- 

 netic pole and curves of variation are shifting now they were shifting 

 then, in that slow change which is still going on from year to year. It 

 has been said that the subject is obscure, and that we do not know these 

 laws in their full extent. That is quite true, and yet we are constantly 

 acting in recognition of them so far as our knowledge extends. It is fair, 



