PERMIT me to observe that the integrating anemometer devised by Mr. Shaw and Dr. Wilson, an abstract of whose paper, read before the British Association (Section A), appeared in your issue of September 15 (p. 467), is in principle and in several of its details identical with a machine intended for the mechanical reduction of anemograms of the Kew pattern adopted by the Meteorological Office, a description of which, with drawings, was placed by me in the hands of Mr. R. H. Scott, and by him transmitted to Prof. Stokes in February last. It is however to be noted that there is a fundamental objection to the mode in which such machines deal with the data submitted to them, namely this, that the air does not, in fact, move parallel to itself, as these integrators and Lambert's well-known expression assume that it does. In other words, the integrator should concern itself only with those particles of air which are passing the anemometer at each instant, i.e. with the directions and velocities of successive elements of the wind at a fixed point. Dr. von Oetlingen (Wild's “Repertorium für Meteorologie”, Band v.) has shown this.