ON THE ANEMOMETERS OF PLYMOUTH. 245 



certain extent empirical is still not without claims to consideration, and it is 

 moreover the only one apparently within our reach. This method consists 

 in determining experimentally, the actual space passed over by the pencil in 

 a given time corresponding to certain velocities of the wind, and finding from 

 this how these spaces vary with the velocities. 



To this effect I examined the best tables of the relative force and velocity 

 of the wind laid down by Rous, Lind, Hutton and others, and by some fur- 

 ther experiments enlarged and extended them. I found that observations 

 with a gauge on Lind's principle were sufficiently accurate for my purpose, 

 and accordingly noted the indications of this instrument simultaneously with 

 the anemometer now under consideration. 



In a former but rather limited series of experiments, the spaces passed over 

 by the pencil came in several instances near the proportion of the square of 

 velocity of the wind. In very strong and steady breezes, however, the spaces 

 passed over by the pencil came nearer the simple ratio of the wind's velocity. 

 When the obstacles arising from the friction and resistance of the machine 

 bear a high proportion to the force which sets it in motion, then, as may 

 be readily conceived, the space passed over by the pencil is less than it would 

 be if no such retarding force existed. As the velocity of the wind increases 

 the resistance is comparatively less, until it may be at last so small as not to 

 interfere considerably with the revolutions of the fly; accordingly we find 

 that in strong winds the ratio between the revolutions of the fly and the ve- 

 locity of the wind give a constant quantity, the spaces described by the pen- 

 cil being taken as proportionate to the revolutions of the fly. 



If we consider attentively the nature of the mechanism in this machine, it 

 will be seen that the friction is very considerable. We have for example 

 (see figure in pageS-iS) perpetual screws working in toothed wheels, so as to 

 convert the rapid motion of the fly into a slow descending vertical motion, 

 again carried out by a thread turning within a moveable nut. We have fur- 

 ther the friction of a pencil attached to this nut, against a fixed cylinder so 

 as to leave a trace on it. This instrument therefore involves the greatest 

 amount of friction incidental to any mechanical machine. We may there- 

 fore conceive that with gentle winds and light breezes the motion of the fly 

 would experience a greater amount of comparative retardation than with 

 strong gales ; hence in the construction of instruments on this principle we 

 should employ a fly of considerable power as compared with the work the 

 machine has to perform. 



In Table XI. will be found the mean results of a series of experiments on 

 the indications of this instrument, as compared with the velocity of the wind 

 deduced by Lind's gauge. 



We may observe in this table, that when the breeze is strong the velocity 

 of the wind (taken in feet per second ) divided by the space passed over by the 

 pencil gives in several consecutive experiments a constant quantity, or nearly 

 so. This is seen in observations 4>, 5, 6, and G, 7, '6, and 9, 10, 11, 12. In 

 these instances therefore, as compared with each other, the velocity of the fly 

 is nearly in proportion to the velocity of the wind. This however is not the 

 case in comparing more distant observations. Thus in observations 1 and 7, 

 the velocity of the wind is as 1 : 2, whilst the space passed over by the pen- 

 cil is as 1 : .^. In the first and tenth series of observations it approaches 

 the square of the velocity of the wind. In fact it is evident by column/J 

 that the ratio of the velocity of the wind to the velocity of the fly is con- 

 tinually decreasing by a variable quantity. In looking over some of Cou- 

 lomb's experiments on the working of windmills, we find a similar result. 

 Thus when velocity of wind was 7 feet per second, the sails made 3 revolu- 



