374 RECORD OF SCIENCE FOR 1887 AND 1888 



page 282 Ferrel shows that the formula expressing the relatiou between 

 the barometric gradient (G) and the velocity of the winds is very much 

 the same for the pressures and winds around a cyclone center as for 

 the pressures and winds about the polar axis of the earth, and is ex- 

 pressed by a quadratic equation s^ + as = 6G. As the cyclone moves 

 from the land to the ocean the so-called friction term would, a ijriorl 

 be expected to sufi'er a decided change, and the actual auiount of such 

 change is indicated by the data collected by Loomis. Adopting his re- 

 sults Ferrel deduces a correction to his formula based upon tbe principle 

 " that the frictional resistance of any stratum of air moving over the 

 earth's surface comes both from the earth's surface and from the stratum 

 above it," and again, "that the direction of motion of the air of the 

 stratum above differs considerably in a cyclone from that at the surface 

 of the earth." To me it seems that there is here not a sufficient dis- 

 tinction between the small and negligible friction called viscosity, which 

 acts both from above and below upon any intermediate stratum, and the 

 resistances due to impact and convection, both which operate princi- 

 pally from below upward ; still the general effect is undoubtedly, as 

 Ferrel says, to make the gradient that accompanies a given velocity of 

 the wind at the earth's surface greater than that accompanying the 

 same velocity of wind over the sea or at higher altitudes. Similarly 

 the gradient is less in summer than in winter. 



The chapter on tornadoes deals in a very interesting manner with the 

 different types of cyclones in which the horizontal movement is less 

 conspicuous than the vertical movement ; the formation of water-spouts 

 is explained as a special case of tornado action, and examples are com- 

 puted showing the dimensions of the spout as depending on the hu- 

 midity of the air. In a section on the force of the wind and supporting 

 power of ascending currents, Ferrel gives merely the old approximate 

 formuhie for the resistance of the air varying as the square of the velocity 

 and the square of the cosine of the angle of incidence, and applies the 

 resulting numerical resistances to the explanation of the formation of 

 large drops of water, cloud-bursts, hail-stones, and the destructive ef- 

 fects of tornado winds. Stokes's explanation of the effect of viscosity 

 seems to have been overlooked. /)n page 314 he explains the effective 

 force of the wind against an obstacle, or in the production of drafts up 

 chimneys, as due " not simply to the inertia of the air but to the drag- 

 ging effect of the air through friction upon the columns of air in the 

 front and rear of the obstacle." 1 presume that this "drag" is inten- 

 ded to refer to viscosity or so-called internal friction of gases, and appar- 

 ently the same use of the word is made by Hagen in his exi)lanation of 

 the fact that the pressure against a thin plate depends upon the size and 

 shape and especially upon the sharp angles of the plate. It is a suffi- 

 cient answer to this introduction of viscosity to state that the value of 

 the viscous resistance can be easily computed with sufficient accuracy to 

 show that it is not an important factor in these experiments on large 



