January 12, 1900.] 



SCIENCE. 



49 



On the other hand, the general character 

 of the circulation of the atmosphere and the 

 meteorological consequences thereof, have 

 been brought within the domain of mathe- 

 matical research, if they have not yet been 

 wholly reduced to quantitative precision. 

 The pioneer in this work was a fellow-coun- 

 tryman, William Ferrel (1817-1891),* who, 

 like Green, came near being lost to science 

 thi-ough the obscurity of his early environ- 

 ment. It is a curious though lamentable 

 circumstance, illustrating at once the pe- 

 culiar shyness of Ferrel and the proverbial 

 popular indifference to discoveries which 

 cannot be patented, that a man who had 

 mastered the Principia and the M^canique 

 C61este and who had laid the foundation of 

 our theory of the circulation of the atmos- 

 phere, should have found no better medium 

 for the publication of his researches than 

 the semi-popular columns of a journal de- 

 voted to medicine and surgery. But such 

 was the medium through which Ferrel 's 

 ' Essay on the Winds and Currents of the 

 Ocean ' appeared f in 1856. Since that time 

 notable progress has been made at the hands 

 of Ferrel, Helmholtz (1821-1894), Ober- 

 beck, Bezold and others ;% so that we may 

 entertain the hope that the apparently 

 erratic phenomena of the weather will 

 presently yield to mathematical expression, 

 just as the similar phenomena of oceanic 

 tides and terrestrial magnetism have al- 

 ready yielded to the power of harmonic 



* For a biography and autobiographical slsetoh of 

 Ferrel, and a list of his publications, see Biographical 

 Memoirs of the National Academy of Sciences, Vol. 

 III., pp. 265-309. Washington, 1895. 



fin Nashville Journal of Ifedioine and Surgery, Oct. 

 and Nov., 1856. 



J Some of the most important papers and memoirs 

 on this subject, collected and translated by Professor 

 Cleveland Abbe, have been published by the Smith- 

 sonian Institution under the title ' The Mechanics of 

 the Atmosphere.' Smithsonian Miscellaneous Col- 

 lections, No. 843, Washington, 1891. 



When we pass from the atmosphere to 

 the hydrosphere, several questions concern- 

 ing the nature and properties of their com- 

 mon surface, or what is usually called the 

 sea surface, immediately demand attention. 

 The most important of these are what may 

 be distinguished as the static and the kinetic 

 phenomena of the sea surface. Since tidal 

 oscillations belong more properly to hydro- 

 kinetics, we may here confine attention to 

 the static phenomena. 



■ Starting from the datum plane fixed by 

 Laplace, the most important contribution 

 to the theory of physical geodesy since 

 his time is the remarkable memoir of Sir 

 George Gabriel Stokes ' On the Variation 

 of Gravity at the Surface of the Earth.'* 

 Adopting the hypothesis of original fluidity, 

 or the more general hypothesis of a sym- 

 metrical arrangement of the strata of the 

 earth, with increasing density towards the 

 center, Laplace had shown that the acceler- 

 ation of gravity in passing from the equator 

 to the poles should increase as the square 

 of the sine of the latitude. f This conclu- 

 sion agreed well with the facts of observa- 

 tion ; and Laplace rested content in the 

 opinion that his hypothesis was verified. 

 But Stokes showed that the law of variation 

 of the acceleration of gravity at the surface 

 of the sea is wholly determined by that 

 surface, regardless of the mode of distri- 

 bution of the earth's mass. This, as we now 

 see, of course, is a direct result of the theory 

 of the potential function ; for the sea sur- 

 face is an equipotential surface, and since 

 it is observed to be closely spheroidal, the 

 formula of Laplace follows independently 

 of all hypothesis save that of the law of 

 gravitation. But while Laplace's formula 



* Read April, 1849. See Mathematical and Phys- 

 cal Papers by G. 6. Stokes, Cambridge University 

 Press, 1883, Vol. II. 



t Laplace's formula is g = a -\- fi sinV, where a is 

 the value of g at the equator, /3 is a constant, and iji is 

 the latitude of the place. 



