NATURE OF THE WATER PARTICLE 207 



Now scrutiny of streams under the variable conditions of nature 

 reveals striking diversities, presumably due to variability in the nature 

 and dimensions of the constituent particles : The water of a limestone or 

 granite spring is peculiarly transparent yet highly refractive and spark- 

 ingly brilliant, notably free and rippling in movement, and manifestly 

 tends vigorously and persistently to exckide extraneous matter by prompt 

 grounding or above-wash floating ; the same water, when confined in pools 

 or mixed with surface-flow, soon loses brilliancy and develops an evident 

 tendency to retain and diffuse extraneous matter, which may remain in 

 suspension for hours, days, or even weeks, the difference being so impres- 

 sive that the million casual observers habitually describe one aspect as 

 living and the other as dead; yet the muddied water (in which the extra- 

 neous matter may conveniently be conceived as largely lodged between or 

 among rather than within the water particles) may be induced to clarify 

 itself and at least partially resume its original condition by merely intro- 

 di;cing a clear stream such as that of the original spring. The ready 

 inference is that the particles of the spring water are relatively large and 

 free moving, those of the pool relatively small and constrained. This 

 inference is consistent with the conditions under which ground waters 

 gather — that is, in trickles or series of drops each constrained by surface 

 tension and hence of considerable size. It is consistent also with the pre- 

 sumable mechanism of clarification of the miiddy pool by a clear stream : 

 When the flocculation of matter in suspension was investigated by 

 Hilgard, and later by "Whitney and Cameron, it was found to be affected 

 by chemical conditions, but no adequate explanation was found for the 

 flocculent sedimentation due to the introduction of clear and pure moving 

 water — a phenomenon akin to that of mere introduction of movement, 

 which is well known not only to experimentalists but to every observer of 

 streams (including the boy and even the horse who take supplies from 

 the riffle rather than the pool) ; yet it is manifest that if the clean particles 

 be larger and freer moving, they will on mingling with the stagnant parti- 

 cles of the pool enlarge the interstices or thicken relatively the films of 

 sediment, and so tend to aggregate the extraneous matter in floccules 

 sufficiently large to escape entanglement and gravitate to the bottom by 

 a sort of mechanical precipitation somewhat like that in which hailstones 

 are formed in the air. 



The inference that the tendency toward flocculation inheres in the 

 water rather than the sediment in turn accords with or explains other 

 phenomena: A concurrent inference is that the formation of crystals on 

 freezing is primarily an attribute of the liquid itself, and that the hexa- 

 gonal forms represent lines of structure or tension at or near the surface 

 of the liquid, which persistently tends to extrude these solids much as it 



