account for all the facts that have been noted in a simple and natural manner.
The most important use of the cloud observations is not the study of their constitution, but of their motions relatively to the surface of the earth. A cloud is a meteorological meteor, and moves in the stratum of air at approximately the same velocity as the atmosphere itself, so that a measurement of its direction and velocity gives that of the air current, just as a chip floating on a stream shows how fast the water is running. Repeated measures of this kind, when classified, teach us that the atmosphere flows with certain typical movements, and that by them the laws controlling its average circulation can be determined.
Now it is a fact that because meteorologists could observe the motions of the wind readily at the surface of the earth, but not in the upper strata of the air, they have relied too much upon conjecture in constructing the theories of the constitution of storms. The mathematical analysis has had therefore only an imperfect basis upon which to rest, and consequently it has made slow progress towards a complete solution of the problem. The international observations on clouds, during the year May, 1896-July, 1897, had for their immediate object the accurate determination of the motions of the upper air, with a view of testing the existing theories, and constructing new ones wherever necessary. This period of scientific observation is similar to the Tycho Brahe and Kepler stage of astronomy, when observations of the motions of the planets were accumulated for the use of the coming