B. TERRESTRIAL PHYSICS AND METEOROLOGY. 77 



return it is preferable to sail from Australia eastward past 

 Cape Horn. Bulletin Hebdomadaire, XVI., 28. 



OX THE THEORY OF TORNADOES AND WATERSPOUTS. 



In a general investigation into the phenomena of the cy- 

 clonic movements of the atmosphere, Couste states that, start- 

 ing from the general principle that there exists at the centre 

 of the cyclone a column of ascending gyrating air, he deduces 

 logically the following conclusions : First, the whole column 

 must rotate about its geometrical axis, in the opposite direc- 

 tion to the gyration. Secondly, there must be a vertical os- 

 cillating movement by which the column alternately rises 

 above and descends to the ground, carrying devastation be- 

 fore it. Thirdly, there must be a movement of translation, 

 which is accomplished, as shown by observation, Avith a rapid- 

 ity varying between twenty and seventy miles per hour. 

 These three movements are derived from the centripetal 

 forces developed by the gyration which give rise to lateral 

 streams of air, which he calls radiating filaments, in opposi- 

 tion to those interior filaments which gyrate w 7 ithin the he- 

 lix, and which he calls helicoidal filaments. These radiating 

 filaments of air form a nappe which incloses the whole con- 

 vex surface of the tornado, and they constitute the wall of 

 the column, which wall, for a given state of dynamic equilib- 

 rium, is as solid as if it were a solid matter of sheet iron, for 

 example, yet is permeable and indefinitely extensible accord- 

 ing to the conditions of its dynamical equilibrium. These 

 filaments are directed from below upward, following the 

 tangent to the helix farthest from the axis, producing reac- 

 tions similar to those in turbine wheels. Couste has also de- 

 termined the character of the movement alon^ the surface 

 of the earth ; this is, in general, of a spiral nature, at least 

 for waterspouts properly so called, which appear as trunca- 

 ted columns, suspended from a cloud. But for those tor- 

 nadoes whose trajectory is nearly rectilinear, and for the cy- 

 clones and hurricanes whose birth takes place upon the ocean 

 the trajectory takes the form of a parabola, whose summit is 

 always near the side of some large continent. These re- 

 markable peculiarities he explains by the following principles, 

 which he has deduced from his theory : First, if the angular 

 velocity of gyration increases or diminishes that is to say, 



