104 
Proceedings of the Royal Society of Edinburgh. [Sess. 
air velocities, and like those just mentioned be so designed, because of the 
lightness of its moving parts, efficient damping, etc., as to have no error of 
lag ; but what is required is not a record of the velocity of the wind at 
any moment, but the time integral of that velocity. In theory, of course, 
this might be obtained by the subsequent use of a planimeter ; but owing 
to the gustiness of the wind, either the record strip would have to be of 
prodigious length, or else the fluctuations would be crowded together upon 
it so closely as to put the use of a planimeter out of the question. For 
this reason it is necessary to give up the idea of using an instrument to 
record the velocity from moment to moment, and to turn rather to the 
consideration of totalling instruments such as the Robinson cup anemometer 
already spoken of. In considering the effects of lag upon it, and how they 
may be overcome, but little help can be got from the analogy of other 
totalling instruments. A novel form of ballistic galvanometer has recently 
been introduced possessed of the property that its indications are determined 
by the total quantity of electricity passed through it, and not at all by the 
duration or mode of fluctuation of the current.* The principle relied on 
here, however, has no analogy in air currents. Nor again, of course, has 
the principle so much used in photometry, according to which the eye and 
the photographic plate are both regarded as instruments capable, within 
a considerable range, of accurately integrating the total effect of a number 
of successive flashes of light. f As none of those cases, nor any others that 
could be thought of, seemed to be at all similar to that of the anemometer, 
and as the latter is of some consequence, it was decided to carry out a 
series of experiments with a view to ascertaining the extent of the error, 
and possibly of learning how it might be minimised. 
Apparatus and Experiments. 
A special air meter of the usual fan type was constructed, the velocity 
of the fan being recorded at every instant on a moving strip of paper. 
Difficulties, however, in connection with the production of wind currents 
possessing accurately known fluctuations caused this method to be tempor- 
arily abandoned ; and it was decided to employ currents of water, which 
are susceptible of much easier control and measurement, and which would 
probably give rise to analogous results. A small impact water turbine was 
* The mathematical theory of the Grassot fluxmeter is to be found in the catalogue of 
the Cambridge Scientific Instrument Co. 
+ As regards cases in which this no longer holds true, see, e.g., E. F. Ferry, Phys. Rev., 
vol. i., No. 5, p. 338 ; and for the contrary see, e.g., Investigations on the Theory of the 
Photographic Process , Sheppard and Mees, p. 18, 1907. 
