68 T. H. Holland — Constitution of Laterite. 



imagine, of physical molecules. That this tendency is variable is 

 evident from the fact that some substances readily form well- 

 developed crystals, whilst others require a prolongation of the 

 physical conditions which are favourable. The display of energy 

 due to chemical union is in general sufficiently pronounced to be 

 noticeable ; but there must also be a display of energy when 

 physical molecules have the privilege of exhibiting crystalline 

 affinity, and it is not unlikely that in some cases the heat of formation 

 of a crystal may be greater than that of a chemical compound, in 

 which case, as the hydrates of iron and alumina appear to teach, 

 a chemical compound may be broken up to permit some constituent 

 to crystallize. We shall probably be near the truth if we regard 

 the crystalline affinity as the heat of formation of a double molecule. 

 Thus the thermal value of [Fe- 0^ Fe^ 0^] may exceed [Fe^ 0^ 

 3H^0], in which case, under suitable physical conditions, the 

 water would be driven out of limonite and a crystal of hematite 

 formed, because the system is one containing potential energy due 

 to the proximity of molecules of Fe, O3. Similarly, the potential 

 energy of a system in which Alj O3 . 3 Hj exists contains sufficient 

 potential energy to displace the two molecules of water having 

 a high specific volume to form a crystal of the compound AI2O3 . 

 H2 0. In other words, the thermal value of [Al H 0^ Al H 0^] is 

 greater than [ Al H 0", H- 0] , and the formation of diaspore at the 

 expense of gibbsite is an exothermic change capable of spontaneous 

 development under suitable conditions. 



At a high temperature diaspore would change further to corundum, 

 a change equivalent to the formation of hematite at lower, probably 

 at ordinary, temperatures. Aluminic oxide thus shows a greater 

 affinity for water than ferric oxide does, which is shown by the 

 facts, (1) that limonite is not found crystallized, whilst gibbsite 

 forms definite crystals, and (2) that gibbsite at ordinary temperatures 

 is reduced to diaspore as an exothermic change, whilst the anhydrous 

 compound, hematite, is formed from ferric hydrate under similar 

 circumstances. 



One is aware, of course, that we have but a vague idea of what 

 the amorphous state means. But the passage to the crystalline 

 condition is in general one accompanied by a decrease in molecular 

 volume ; and with the striking case of arsenious oxide before us, 

 it would be well to examine more critically cases in which the 

 manifestation of energy is less pronounced, though necessarily 

 existent and possibly measurable. 



IV. Summary. 



The objects of this note are three : (1) to call attention to the 

 essential chemical similarity between laterite and bauxite ; (2) to 

 offer a theory for the formation of laterite in the moist tropics ; 

 (3) to suggest an explanation for its spontaneous dehj'dration. 



(1) Laterite has generally been referred to as a ferruginous clay ; 

 but if the term clay is restricted to substances having a basis of 

 hydrous silicate of alumina, this definition is incorrect. The 



