ON THE ANEMOMETERS OP PLYMOUTH. 253 



estimated by the motion communicated to some other secondary machinery, 

 termed the gauge, as we find on Mr. Cavendish's manuscript the following 

 remarks : — " Rad. circle which it revolves in := 10 feet and circumf. = 62*83 

 feet, and by mean gauge makes 3*73 rev. for 1 rev. arm, and therefore makes 

 1 rev. for 16"85 feet, or jy^of a mile, therefore, as in former trials, it seemed 

 to make 360 rev. per mile." He also says, " Diameter of vanes is 14*2 inches, 

 therefore vanes make one revolution whilst wind moves over 4^ times the 

 circumference, and consequently if wind is 4^ slower at leading edge of vanes 

 than at the other, they should stand still." 



This instrument worked very freely with light breezes, and when the wind 

 did not exceed 1 mile per hour, or l'43feet in a second, the fly appears to have 

 made "2*62 rev. in ^min." Such a wind as this, according to Rous, is scarcely 

 perceptible, and would not exert more than '005 lb. pressure on a square foot ; 

 it is not measurable by any anemometer at present employed in meteorology. 

 One revolution of this instrument per second = 9*955 miles per hour. 



I have thought it ilesirable to notice these remarks in Mr. Cavendish's 

 papers with a view of showing how possible it is to measure the velocity of 

 very slow currents, and furnish some hints at least relative to the method 

 which this profound and accomplished scholar and philosopher resorted to 

 for the purpose. 



Osier's Anemometer. — A full description of this instrument has been given 

 by the inventor in a quarto pamphlet printed at Birmingham about five years 

 since ; no particular notice or figured description of it however has yet 

 appeared in the Reports of the British Association, notwithstanding that 

 several sums of money and much labour has been devoted to it. I have 

 thought it therefore desirable to preface what I have at present to state re- 

 lative to this anemometer, with a very brief notice and drawing of its prin- 

 cipal parts, in order the more effectually to carry out the object of this Re- 

 port, viz. a full account of the results of the experiments undertaken by the 

 Physical Section of the Association for the purpose of advancing that depart- 

 ment of meteorology relating to the phaenomena of wind. 



Osier's anemometer traces the direction and pressure of the wind on a 

 given area, together with the amount of rain, on a longitudinal register di- 

 vided into twenty-four portions, corresponding to the twenty-four hours of the 

 day, as shown by fig. 1 , Plate XXXVII. The central portion of this paper is 

 devoted to the register of the direction, and has a series of longitudinal lines 

 on it corresponding to the cardinal points. The lower portion,^ g?, is devoted 

 to the register of pressure, and is graduated also by a series of longitudinal 

 lines corresponding to lbs. pressure on the square foot. The upper portion, 

 a c, is devoted to the register of rain, and is graduated in a similar way by a 

 series of lines corresponding to given quantities. Finally, the whole length 

 is divided by verticals or lines perpendicular to the former, as 1, 2, 3, 4, &c., 

 corresponding to the twenty-four hours of the day. This register-paper 

 being placed on a board M, fig. 2, and accurately set every day, is carried 

 along by means of a clock, C, under three pencils, l',2', 3', which may be con- 

 sidered as the fingers or indexes of the machine. The board M moves on 

 friction rollers, and is hence easily drawn along as the clock, and conse- 

 quently the time, advances. 



The pencil 1 ' is the index of direction ; this pencil is operated on by a vane 

 V, turning a vertical hollow shaft, W p. There is a pinion at p which, as the 

 vane turns in the direction of the wind, acts on the rack-work of a transverse 

 bar ef, and so causes it to move either to one side or the other. Now the 

 pencil 1' is attached to this bar, and hence is caused to leave a trace on one 

 of the longitudinal lines of the register corresponding to the cardinal points. 



