Professor Russell, in commenting on this subject, expresses the opinion that the low temperature is due to the convective interchange of air at a low temperature in the upper strata with air of a high temperature in lower strata, such convective interchange tending to make the whole body of air of a temperature coinciding throughout with the adiabatic rate of upward diminution, with the consequent result of rendering the air at the surface of the earth cooler than previously and the upper air warmer. When the upward diminution of temperature is less than the adiabatic rate, in the forced circulation of air crossing a mountain ridge, there occurs the dynamic heating which is observed in the case of the foehn winds. The low temperature near the earth he does not believe could ever be entirely produced by nocturnal radiation from the ground. The high pressure, in his opinion, is largely the result of greater density due to low temperature, as is very clearly indicated by the fact that the temperature is almost inversely proportional to the pressure, and that the places of lower temperature substantially coincide with the places of greatest pressure. In advancing hypotheses and inviting discussion the real object is, or at least should be, to discover the essential cause or causes which determine the initial formation and subsequent maintenance and progress of the cyclone. Some real progress in charting lines of equal density seems to have been made by M. Nils Ekholm following Professor Abbe's system of "isostaths," one using the term density, the other buoyancy. Professor Abbe also introduces the factor of the orographic gradient, but the {44} latter is simply the measure of a resistance. The objection to this form of determination is this, that it is a measure of mass only. The density of two masses of air is determined to be the same; but as the density may result from two entirely different causes, their physical relations cannot be fully expressed in units of gravity. The methods of Professor Abbe and of M. Nils Ekholm undoubtedly give good results, partly from the coincidence that humidity usually varies directly as the temperature. The method proposed by Captain James Allen in 1888, which is briefly described in appendix 24 to the annual report of the Chief Signal Officer for 1890, appears to afford the means of more clearly expressing the relations that exist between the mass of the atmosphere and the forces available for the generation and movement of storms. Its tentative application at the Signal Office has anticipated and explained storm movements not indicated or accounted for by the usual methods. As pertinent to this matter, there is instanced a study of the progress of thunder-storms made by Berg, who observes that the line of storm front in every case investigated made a decidedly conspicuous bend into the densest part of the lines representing the absolute humidity. * * * * *