70 QUICKSTLVEU DKrOSITS < H'^ TIIK PACIFIC SLOI'K. 



of which are equal to or less than the distance to wliich the fluid can per- 

 meate will yield A'ery rapidl}", Avhile tliose of less abrupt curvature will be 

 more slowly decomposed or dissolved. There will thus be a constant tend- 

 ency to diminish the curvature of the more salient portions and, if the mass 

 is not too thin, to reduce it to a sphere. 



Cases. — Two special cases need con.sideration: If the action of the fluid 

 were strictly confined to tlie surface (ir if the mass were absolutel}' imper- 

 meable, the radius of curvature would alwa3"s be infinite compared with the 

 distance to which a fluid could penetrate, and, if solution took place, the 

 mass would always retain an angular form, the surfaces of which would be 

 parallel to those whicli it orig-inally presented. On the other hand, if the 

 fluid could permeate to tlie center of tlio Ijody, all portions would be 

 attacked at once and it would disintegrate almost simultaneouslv through- 

 ont its mass. 



Nearly every American has daily opportunities for observing the rela- 

 tions here reduced to exact terms. Clear, solid ice is practically imperme- 

 able by w'ater, and an angular fragment of such ice in a glass of water 

 becomes only sliglitlv i-ounded, wliile the surfaces at all stages of the 

 melting process are nearly parallel to the original ones. On the other 

 hand, a bit of ice which is clouded with small air bubbles is permeable by 

 water to tlie depth of these bubbles, and consequently the edges and cor- 

 ners of an angular mass of such ice are quickly rounded. Again, a lump 

 of cane sugar is very porous and fluids permeate to its center. It tliere- 

 fore disintegrates under tlie action of a solvent fluid almost without pre- 

 liminary rounding of the edges and corners. 



Application to decomposing rocks. — Tlio bcliavior of rocks to dissolviiig or decom- 

 posing agencies is similar. There do not seem to be an}' rocks, excepting 

 perhajis some obsidians, whicli are permeable onlv to an insensible distance 

 by fluids ; but there are many rocks so dense that fluids penetrate them 

 with great ditficulty and very slowly. In such cases the corrosive reagents 

 which waters contain are neutralized by the time the solutions have pene- 

 trated to a very small depth, and corrosive action is limited to this thin outer 

 layer As decomposition is completed in the outer layer, active reagents 

 will of course permeate farther and farther into the rock. The g-eometrical 



