342 REPORT— 1846. 
In this calculation a is taken at 1:0 and 4 at 14°75. The small differences 
are quite within the range of errors of observation. 
Mr. Harris has furnished us with experiments* in which the revolutions 
of the vane of Whewell’s anemometer were compared directly with the 
wind-pressure on Lind’s well-known instrument, and to these the same form 
of calculation is equally applicable. In the subjoined Table the observed 
values of v! and V2 are first given, and then a column of values of V®, cal- 
culated from the formula P = av'?+ 6v', the values of a and 6 being taken 
at 1 and 10. 
v. V2, v2. 
Observation.| Calculation. | Difference. 
| 
* . i“ . “ut . t 
Se ah mts 7 These differences of V? are 
2-0 -080 -081 +001 much within the possible 
2-5 -100 “106 +006 errors of observations. Those 
3-0 “130 132 4-002 marked ” examined by differ- 
355 “160 “160 asta ences, appear to be the least 
4-0 ‘190 -190 i in harmony with the rest, 
aoe -290 297 +007 and it is from these that the 
5:0 “270” O54 _ 016 greatest deviations occur in 
Be -290 -288 _-002 the calculation. Similarly 
6:0 330 396 —-004 compared, V as deduced from 
65 +350” 360 4-010 Lind and V as deduced from 
7-0 “400 +1) hee hae age Whewell,are almost identical. 
By this very simple calculation, then, any one rate of revolution of the vanes 
of Whewell’s anemometer may be made to indicate the corresponding velo- 
city of the wind. But we cannot from the sum of these rates, obtain by this 
calculation the corresponding sum of the velocities of the wind; since the 
relation of these sums to each other depends on the individual values of v’, 
and these are not recorded. They may be recorded, in a form fit for the cal- 
culation, by adding a clock-movement, which shall cause the instrument to 
register the series of values of v', but the machine then loses its simplicity. 
An approximation to the individual values of v' may be had by a process 
suggested by the inventor, but not (it is believed) put into constant use by 
any observer. It consists in simply turning round the cylinder, on which the 
wind register is written, after regular intervals of time by hand. The shorter 
these intervals, the nearer the approximation. 
If we knew precisely the Jaw according to which the wind’s velocity rises 
and falls with the lapse of time, the correction of the record in Whewell’s 
anemometer might become more complete ; and it seems no small recommen- 
dation to observers for practising the hourly rotation of the instrument, that 
this process would speedily furnish data for the determination of that law. 
The conversion of the register effected by Whewell’s anemometer into 
pounds of pressure or miles of air-movement may perhaps be sufficiently 
easy and accurate, if only two things are attended to :—first, the values of 
the constants a and 6 in the previous formula must be determined by obser- 
vationt; and secondly, the register scale should be read and the results re- 
corded sufficiently often to obtain an adequate number of values of 2’. 
* Reports of the British Association, 1844, p. 263. 
+ The observation must be a comparison of Whewell’s registration with some experimental 
contemporaneous determination ; the simple pressure of wind may be obtained by Lind’s 
