TRANSACTIONS OF THE SECTIONS. 39 



ment, has now been so altered as not to cause any injury when made use of, as it 

 sometimes did previously. 



It was found in a late tour on the continent that no part of the instrument is 

 liable to injury from the ordinary concussions of travelling ; and its use was ascer- 

 tained to be as well adapted to continental climates, so far as was tried, as to that of 

 Great Britain. 



Wind-charts of the Atlantic, compiled from Maury's Pilot Charts. 

 By Captain FitzRoy, R.N., F.R.S. 



These diagrams are intended to show what winds prevail, at the four quarters of 

 the year, in the Atlantic. 



Each figure should be considered by itself alone, as the scales are generally very 

 different, depending on the number of observations from which the respective 

 diagrams are constructed. 



Relative prevalence of wind (or calm) is shown in each square of ten degrees ; but 

 in no case is absolute amount given ; nor is strength of wind exhibited, as it may be 

 hereafter. 



The navigator may be influenced, in shaping a course, by the probabiUty of find- 

 ing certain winds more or less favourable in certain localities. 



To a sailing ship such considerations are most important ; and a glance at these 

 charts shows a seaman how the wind blows (usually) during a season, as readily as 

 his " dog-vane " indicates the (apparent) direction at any moment of observation. 



The diagrams illustrate Maury's Pilot Charts, in which similar information is 

 offered by numbers, which require more mental operation in their use than these 

 graphical figures. 



In each square the numerical data contained in four of Lieut. Maury's five-degree 

 squares are combined in the following manner. 



Of a circle inscribed in any such square, the radius is taken as a measure of the 

 sum of the greatest number of observations of the most prevalent wind ; and other 

 lines, likewise drawn (to leeward respectively) from the centre, and on the same scale, 

 indicate the relative duration or prevalence of other winds (each observation 

 referring to a period of eight hours), and through the extremities of these lines a 

 boundary is traced. 



As a circle is said to be generated by the revolution of a line around a point, so 

 the figure representing successive directions of wind may be supposed to be generated 

 by the motions of a wind-vane, and the lines or points may extend from the centre 

 (like the growth of crystals) in proportion to the persistence (or continuance) of the 

 vane in their respective directions. 



The relative amount or duration of calms is shown by a circle, of which the 

 radius equals (according to the scale of the diagram) the number of (eight-hour) 

 periods in which there was little or no wind. 



The direction of wind is corrected, approximately, for variation of the compass. 



The larger area of each figure is to leeward of the centre of the square (or inscribed 

 circle) . 



The calendar quarters of the year are adopted advisedly, because the considera- 

 tion of seasons in all quarters of the globe, and the examination of Maury's charts 

 (including those of the trade-winds), induce the belief that extreme periodical changes 

 of wind follow at a certain interval, rather than accompany the extremes of tempe- 

 rature or climate. 



The small figures at the lower left-hand corner indicate the total number of (eight- 

 hour) observations of calm, as well as of wind, recorded as having been made in 

 that square ; and the figures at the lower right-hand corner show, in decimals of an 

 inch, the unit of scale employed in constructing the diagram in that square. 



The force of wind is not shown, because it was not noted in the records from 

 which these charts were compiled ; but at a future time it may be given so com- 

 bined and arranged as to indicate average strength as well as direction. 



Nothing more is thus shown, in a graphical manner, than has been exhibited 

 numerically in Maury's original Pilot Charts, whence solely the data for these were 

 obtained. 



For the few squares still blank, sufficient data have not yet been collected. 



The number of observations used in constructing each diagram aflfords a scale of 



