172 The Rev. T. R. Robinson's Description of an improved Anemometer. 



plane which holds the paper move by means of two slides at right angles to 

 each other. Its motion should be given by a rack, travelling proportionally 

 to the space, but revolving so as to be always in the direction of the wind. 

 A pencil placed over the centre of its revolution would describe on the 

 paper a track perfectly similar to that of the wind. At each hour and quarter 

 hour, a clock was to print by punches a series of marks, which would repre- 

 sent the time.* The mechanical arrangements were planned, and certainly this 

 method would have the great advantage of showing at one view the three vari- 

 ables, and speaking most distinctly to the eye; but I gave it up, from a convic- 

 tion that it is much less adapted to give periodical means than the method 

 of co-ordinates. 



Of these I prefer the polar to the rectangular, for the following reasons. 

 In the first place, the direction, being an angle, is at once recorded ; secondly, 

 a movement of rotation can be given to the paper-holder with much less fric- 

 tion than a rectilinear one ; thirdly, this movement may be continued through 

 many cii'cumferences without inconvenience, while the other is limited by the 

 length of the rack, or other contrivance for producing it. In the rectilinear 

 direction-register there is also the inconvenience that, if the wind veer several 

 times in the same direction round the horizon, a new series of graduations must 

 commence. I may add, that there is, perhaps, a want of graphic propriety in re- 

 presenting angular veering by a right line, but none in measuring miles by a 

 graduated arc. Fourthly, one form of printed paper serves for both. The only 

 objection to the polar form, of which I am aware, is, that the scale is less near 

 the centre than at the circumference; this, however, may be obviated in any case, 

 when it is desired, by winding up the apparatus at shorter intervals, so as 

 to keep the pencil near the latter. 



First, then, as to the space: the dimensions which I have adopted for the 

 windmill are such that, in 440 revolutions, the hemispheres travel one mile, 



* August I6th. This, I find, has been applied by Mr. Ossler, who showed me at the late meeting 

 of the British Association, some beautiful wind-curves, where the time is thus expressed. He checks 

 the excursions in direction by using a windmill, and with great success. The x of his curves is 

 the space, the y the direction. The time is shown by dots, single and multiple, struck in pairs at 

 each side of the paper, and its record is very complete. 



