166 KOYAL SOCIETY OF CANADA 



is of very doubtful composition and is, indeed, regarded by many as a 

 mixture of an aluminium silicate with calcium bicarbonate. 



In the case of dawsonite, Harrington brought forward several satis- 

 factory arguments in support of the view that the aluminium is present 

 in tlie fomi of carbonate. He pointed out that the excess of COo shown 

 in his analyses over that required to form normal carbonates with the 

 bases is about 11 per cent, which must either be in combination with the 

 aluminium, or else the bases must be partly present as bicarbonates and 

 the aluminium as hydrate. On tlie latter hypothesis, there is just about 

 sufficient water present in the mineral to form the dihydrate, A1(0H)2; 

 but this is a substance which can be prepared in the laboratory and is 

 found to be insoluble in the stronger acids, whereas dawsonite is easily 

 and completely dissolved by them. Further, the monohydrate (diaspore) 

 is only soluble after ignition, and the trihydrate (gibbsite) is difficultly 

 soluble in acids, both the latter being well known as minerals. On 

 these grounds Harrington felt justified in concluding that the aluminium 

 is present in the mineral in the form of carbonate, and even to-day 

 dawsonite remains unique among minerals as being the only one of any- 

 thing like definite chemical composition which contains aluminiurn in 

 this form. 



Some years afterwards, in 1881, dawsonite was reported as having 

 been found in Tuscany^ (Pian Castagnaio, Province of Seina, etc). 

 Here it occurs in the form of extremely fine *white needles, arranged in 

 radiated groups and tufts (closely resembling some wavellite), and 

 having a silky lustre. These form thin coatings along crevices in the 

 marls and sandstones of the district, the associated minerals being cal- 

 cite, dolomite, pyrite, cinnabar, bitumen and fluor. These small tufts 

 of white crystals had long been noticed by the miners, who are said to 

 have regarded their presence as a favourable indication in searching for 

 cinnabar. 



Analyses of this material by Friedel proved it to be dawsonite; it 

 was evidently possible to obtain it in a purer condition for analysis than 

 was the case with the Montreal specimens, and although it yielded 

 similar percentages for the main constituents, lime was found to be 

 absent altogether; from the results of his analyses, Friedel calculated 

 the composition to be Al2O3.Na2O.2CO2.2H2O, which he also wrote in 

 another form as Al2(C02N"a)2(0H),.* 



' ('. Friedel. Bull. Soc. Min. 4. 28 (Jan.), 1881. M. Chaper. Ibid. 4. 155 

 (June) 1881. 



* It will be noticed that these two formulze are not identical. A con- 

 sideration of the chemical composition is left till p. 173. 



