POLAR ICE-CAPS AND SEA LEVELS, 321 



" function to a constant. This function, when ex- 

 " panded in series of spherical harmonics, has for 

 " its first term the potential due to the whole mass 

 " supposed collected at its altered centre of gravity. 

 " Hence a spherical surface round the altered centre 

 " of gravity is i\\e first approximation in Laplace's 

 " method of solution for the altered level surface. 

 " Mr. Croll has with admirable tact chosen, of all 

 " the arbitrary suppositions that may be made 

 " foundations for rough estimates of the change of 

 " sea level due to variations in the polar ice-caps, 

 " the one which reduces to zero all terms after the 

 " first in the harmonic series, and renders that first 

 " approximation (which expresses the essence of the 

 "result) undisturbed by terms irrelevant to the 

 " great physical question. 



" Mr. Croll, in the preceding paper, has alluded 

 "with remarkable clearness to the effect of the 

 " change in the distribution of the water in in- 

 " creasing, by its own attraction, the deviation of 

 " the level surface above that which is due to the 

 "given change in the distribution of solid matter. 

 " The remark he makes, that it is round the centre 



VOL. II. Y 



