662 ON CONCRETIONS, SPICULES, ETC. L^^h. 



unstratified structure all the colours of the opal and the iridescence 

 of ancient glass are alike due. 



Besides the tendency to rhythmic action, as manifested in 

 Liesegang's rings, the association of colloid matter with a crystalloid 

 in solution may lead to other well-marked effects. These include*: 

 (1) the total prevention of crystallisation; (2) suppression of certain 

 of the lines of crystal growth; (3) extension of the crystal to 

 abnormal proportions, with a tendency to become compound; 

 (4) a curving or gyrating of the crystal or its parts. 



It would seem that, if the supply of material to the growing 

 crystal begin to run short (as may well happen in a colloid medium 

 for lack of convection-currents), then growth will follow only the 

 strongest lines of crystallising force, and will be suppressed or 

 partially suppressed along other axes. The crystal will have a 

 tendency to become filiform, or ''fibrous"; and the raphides of our 

 plant-cells, and the needle-hke "oxyotes" of sponges, are cases in 

 point. Again, the long slender crystal so formed, pushing its way 

 into new material, may start a new centre of crystallisation: 

 whereby we get the phenomenon known as a "relay," along the 

 principal lines of force and sometimes along subordinate axes as 

 well. This phenomenon is illustrated in the accompanying figure 

 of common salt crystaUising in a colloid medium; and it may be 

 that we have here an explanation, or part of an explanation, of 

 the compound siliceous spicules of the Hexactinellid sponges. 

 Lastly, when the crystaUising force is nearly equalled by the 

 resistance of the viscous medium, the crystal takes the fine of least 

 resistance, with very various results. One of these results would 

 seem to be a gyratory course, giving to the crystal a curious wheel- 

 like shape, as in Fig. 306; and other results are the feathery, 

 fern-like or arborescent shapes so frequently seen in microscopic 

 crystallisation. 



To return to Liesegang's rings, the typical appearance of con- 

 centric rings upon a plate of gelatine may be modified in various 

 experimental ways. For instance, if our gelatinous medium be placed 

 in a capillary tube immersed in a solution of the precipitating salt, 

 we obtain (Fig. 304) a vertical succession of bands or zones regularly 



* Cf. J. H. Bowman, A study in crystallisation, Journ. Soc. of Chem. Industry, 

 XXV, p. 143, 1906. 



