THE LAW OF EVOLUTION CONTINUED. S77 



transition to its existing state has been through stages in 

 which the characters became more determinate. Besides 

 being comparatively unstable in surface and contour, a 

 liquid spheroid is less specific than a solid spheroid in having 

 no fixed distribution of parts. Currents of molten matter, 

 though kept to certain general circuits by the conditions of 

 equilibrium, cannot, in the absence of solid boundaries, be 

 precise or permanent in their directions: all parts must be 

 in motion with respect to other parts. But a superficial 

 solidification, even though partial, is manifestly a step 

 towards the establishment of definite relations of position. 

 In a thin crust, however, frequently ruptured by disturbing 

 forces, and moved by every tidal undulation, fixity of rela- 

 tive position can be but temporary. Only as the crust 

 thickens, can there arise distinct and settled geographical 

 relations. Observe, too, that when, on a surface 



that has cooled to the requisite degree, there begins to pre- 

 cipitate the water floating above as vapour, the deposits 

 cannot maintain any definiteness either of state or place. 

 Falling on a solid envelope not thick enough to preserve 

 anything beyond slight variations of level, the water must 

 form shallow pools over areas sufficiently cool to permit con- 

 densation ; which areas must pass insensibly into others that 

 are too hot for this, and must themselves from time to time 

 be so raised in temperature as to drive off the water lying 

 on them. With progressing refrigeration, however, — with 

 a growing thickness of crust, a consequent formation of 

 larger elevations and depressions, and the precipitation of 

 more atmospheric water, there comes an arrangement of 

 parts that is comparatively fixed in both time and space; 

 and the definiteness of state and position increases, until 

 there results such a distribution of continents and oceans 

 as we now see — a distribution that is not only topographic- 

 ally precise, but also in its cliff-marked coast-lines presents 

 divisions of land from water more definite than could have 

 existed when all the uncovered areas were low islands with 

 2G 



