68 LijV (111(1 L<((( IS oj Francis (iaffon 



machine for plotting isochrones ; the individuaUtv of the ship is represented 

 by zinc teniphites cut to its sjiiling qualities unaer each force of wind, and 

 each t*?niplute corre*iH>nding to the number of |X)ints the course of the sliip 

 lies off the wind. Galton considered that such templates could be cut for some 

 moderate number of classes of ships, and charts of isochrones for such classes 

 hv then issued for the principal oceans, (ialton remarks in liis Memories' : 



"I wu rather scandalised by finding how little whs known to nautical men of the sailing 

 qualities of their own ships, along v»ch of tlie sixt'Oen points of the conii>as.s, a.ssuming a 

 modernto soa and a moderate wind blowing stoAdily from one direction. I think, if I had 

 a yacht, that this would be the first point I should wish to ascertain in respect to her 

 perfonnances." 



In his Royal Society paper he states that no human brain from a mere 

 inspection of the crude data of winds, currents, etc. can deduce a correct 

 result as to the distances likely to be run by a given ship on various 

 courses. 



"As an example, I may be allowed to mention*, that I asked a naval officer of unusually 

 large experience in the construction of weather charts, and who was familiar with the sailing 

 quslities of a ' l^aufort staml'ird ship,' to csliniatp portions of isochrones in certain cases; and 

 I found the mean error of his estiinat^^^ to exceed 15 |)er cent. The guesses of ordinary navi- 

 gators would necessarily be much wider of the truth. Now we must rrcollect that a very small 

 saving on the average length of voyau<^ would amount to an enormous aggreuate of commercial 

 gain, and that, where precision i.s practicable, we should never re-sl satisfied with rule of thumb. 

 Our meteorological statistias afTord the best information attainablti at the present moment, and 

 they exceed by some hundredfold the experiences of any one navigator; their probable errors 

 may be large but that is no reason for needlessly astiociating them with additional subji'cts of 

 doubt. The probable error of a navigator's estimate of an isochrone, and consequently of the 

 data which he must use whether ccn.sciously or not, whenever he att(>mpts to calculate his best 

 track, is due at the present time to no less than tliree distinct sets of uncertainties: (a) the 

 average weather; (6) the performance of his ship on different courses with winds of difTerent 

 force (which I undci-stand to be hardly ever ascertained with much precision); (c) the computa- 

 tion of the isochrone." 



(ialton propo.sed to reduce the uncertainty to (a). 



There is much of interest in this series of papers which are very charac- 

 teristic of the author's originality in idea and in method, but alas ! the 

 papers ought to have appeared 20 years earlier. The modern reader hardly 

 realises that the bulk of our stores were carried in 1857 to the Crimea in 

 sailing ships; that even at the time Galton wrote these pa|)ers a considerable 

 proportion of trade was still carried by sail. Published in 1850, these 

 papers would probably have been followed by the universal construction and 

 use of isochronic charts, and Galton 's name would have been honoured in 

 the history of navigation. But in the 'seventies steam wiis rapidly super- 

 seding sails, and sails were practically discarded before the Meteorological 

 Office had time to collect the more ample and trustworthy data of ocean 

 statistics, on the publication of which Galton's charts depended. Each 

 mfxle of transit is succeeded by another, the railways killed «inals, as motor 

 traffic is killing the railways. It is hard on the discoverer and inventor to 

 be working at a period of transition on a method of transit which has not 



' p. 240. « Nin,. Soc. Proc. Vol. xxi, p. 267. 



