T. H. Holland — Constitution of Late rite. 65 



oxidation of the ferrous salts released during the lateritization of 

 basalt. But the important forms to identify will be those which 

 supply the necessary energy for the more complete disintegration of 

 an aluminous silicate than appears to be possible by chemical means 

 alone at the temperature of the atmosphere. Our explanation cannot 

 be accepted as wholly satisfactory until the agents have actually 

 been identified ; but there is enough circumstantial evidence to 

 warrant the suspicion that a special form of organism is the direct 

 cause of lateritization in the tropics. 



III. The Dehydea.tion of Latekite. 



There is another interesting question suggested by laterite. The 

 analysis by Professor Busz of the lateritic sand from the Seychelles 

 shows a lower degree of hydration than that which characterizes the 

 laterite formed in situ on the granite and diorite. This agrees exactly 

 with our experience in India.' ' Low-level,' recemented laterite is 

 always less hydrous than the ' high-level,' freshly-formed material ; 

 and there is in general a correspondence between the development 

 of concretionary structures and the degree of dehydration. Soft, 

 yellow masses, freshly cut out with a spade, gradually become hard 

 and deepen in colour, until they approach a reddish-brown. Every 

 Public Works officer knows this, and feels with satisfaction that his 

 bridges, built with laterite, strengthen with age. 



The drying of laterite is not a case of mere loss of mechanically 

 included water : the ferric hydrate parts with combined water and 

 passes from the condition of limonite, with its yellow colour in 

 powder, to that of turgite, whilst at the same time a concretionary, 

 and in places microcrystalline, structure is developed. The hydrated 

 alumina apparently undergoes a similar change, passing from 

 gibbsite, AlgOg . 3 H2O, to diaspore, AUOg . HjO, and this irregular 

 loss of water is probably the cause of variation in bauxite, which is 

 intermediate between gibbsite and diaspore, but is inconstant in 

 composition, as one would expect in a mixture. 



Our knowledge of the constitution of ferric hydrates is very 

 imperfect, and, of the natural forms, far less precise than one would 

 imagine from the number of specific names employed to distinguish 

 forms which are probably mere mixtures possessing insvifficient 

 individuality to crystallize. Gothite, corresponding to diaspore, is 

 the only hydrate definitely crystallized. But the anhydrous oxide, 

 hematite, on the other hand, shows an eminent tendency to 

 crystallize, and is amongst those minerals which might be well 

 spoken of as crystallographically robust. The tendency to crystallize 

 is thus exhibited by the anhydrous and least hydrated members of 

 this group. 



At the risk of sacrificing precision for brevity we can refer to 

 this tendency for the physical molecules to come together and form 

 crystals as stronger than the chemical affinity between ferric oxide 

 and water. The two manifestations of energy thus come into 



^ Geol. Mag., December, 1899, p. 542, note 1. 



DECADE lY. VOL. X. NO. II. 5 



