T. H. Holland — Constitution of Later ite. Q7 



of water must consequently be a chemical change of the endo- 

 thermic kind. But where, seeing the compound is not heated from 

 ■extraneous sources, dues the energy come from ? I believe the 

 answer to this question is given by the fact that the dehydration 

 is accompanied by the crystallization of the molecule of ferric oxide 

 and of the aluminic hydrate with the lowest possible molecular 

 volumes. 



The concretionary structures v^hich characterize laterite and 

 bauxite are but the foreshado wings of crystallization — a preliminary 

 segregation of the molecules, accompanied by expression of the 

 water — to which all well-defined chemical compounds aspire, though 

 molecular movement may not always be sufficiently free to produce 

 noticeable results at ordinary temperatures. 



The comparative opacity of the iron-oxides prevents the use of the 

 microscope as a means of tracing the passage from the most hydrated 

 to the anhydrous condition; but it seems very likely, judging by the 

 analogy of the aluminous hydrates, that some of the intermediate 

 stages are mere mixtures, and that there is the usual tendency to 

 pass into the forms with smaller molecular volumes. The tendency 

 to crystallize — the grouping of physical molecules — is analogous to 

 the manifestation of chemical affinity amongst chemical molecules, 

 and there must, in the same way, be differences of intensity. I am 

 indebted to Professor A. H. Church for references to the case of 

 arsenious oxide, As^Og, which, on passing from the amorphous to 

 the octahedral form, gives out 5,300 gram-units of heat, whilst the 

 passage from the amorphous to the monoclinic form is attended by 

 the production of 24,950 heat-units per As^ Oe grams. In these 

 cases the octahedral form has a lower molecular volume than the 

 amorphous, and the monoclinic form lower than either. Thus, the 

 change of physical state amongst solids may result in a change of 

 entropy as great as that due to chemical change ; that, in fact, 

 crystallization may cause a measurable dissipation of energy. 



Thomseii gives the heat of formation of aluminic hydrate 

 [AP, 0^ 3 H-0] as 388,920 calories, and of ferric hydrate [Fe-, 0^ 

 3 H^ 0] as 191,150 calories, both values being less than the heat of 

 formation of the respective anhydrous oxides.^ Without a direct 

 determination it would be unsafe to regard the dehydration of these 

 compounds as an exothermic change ; but with the simultaneous 

 crystallization of diaspore in one case and of gothite or even 

 hematite in the other, one is safe in assuming that the changes 

 manifested by the development of concretionary structures in laterite 

 are in the total exothermic, and that the ' tendency ' to crystallize 

 is a condition worthy of serious consideration and of comparison 

 w^ith chemical changes. 



The expression crysfaJline affinity might be employed conveniently 

 as an analogue of chemical affinity for the tendency exhibited by 

 substances to form crystals by the regular disposition, as we 



1 J. Thomsen : " Thermocliemische Untersuchungen, " 1882, vol. iii, pp. 240, 294. 

 Roberts - Austen : Nature, August 8th, 1901, p. 360. Naumanu : "Thermochimie," 

 1882, p. 514. 



