Church — On Native Phosphates of Aluminmm. 553 



An inspection of these figures shows tliat variscite loses nearly all 

 its water at 100°C., but practically none, or at least not a whole mole- 

 cule, in vacuo over sulphuric acid. If we reject, as non-essential, the 

 2-64% H2O lost in vacuo, then the percentages deduced from the 

 analyses given above will stand thus : — 





Experiment. 



Theory. 

 AI2O3, P2O5, 4aq. 



Alumina, 



Ferric Oxide, 



Phosphorus Pentoxide, . . 

 Water, 



31-04 



1-56 



4.5-15 



• 23-27 



32-45 



44-82 

 22-73 



§ V. According then to the above-given analyses of normal native 

 hydrate and phosphate of aluminium, the former loses no water at 

 100°, the latter all. The results of R. Helmhacker's experiments 

 with the supposed variscite of Ereienstein, near Leoben,^ are not in 

 accordance with this view so far as regards the normal phosphate. 

 He obtained 16-11% wa,ter lost at 100°, and 17-57% on ignition. 

 The alumina amounted to 34-46%, while the phosphorus pentoxide 

 was only 25-69. These figures are allowed by Helmhacker to point 

 to an admixture of diaspore with his variscite in the ratio of 4:5; 

 now, as diaspore resembles gibbsite in its total retention of water at 

 100°C., we possess, in these apparently anomalous results, an actual 

 confirmation of the conclusion to which my own experiments had led. 

 Possibly Helmhacker's 100°C. may have been the conventional expres- 

 sion for the temperature of the water-oven, which would fall consider- 

 ably short of that figure. With an air- oven at 100° C. his mineral 

 would probably have lost more than 16-11%. 



§ VI. Here perhaps it may be well to introduce the analysis of a 

 mineral from Langen Striegis which, though in physical characters 

 resembling the peganite of Breithaupt, yet gave very different results 

 from those obtained with that mineral (or what we must assume to 

 have been that mineral) in the hands of Hermann. He obtained 

 numbers corresponding to those of a member of the calaite group 

 (2AI2O3, P2O5, 6H2O); but I find that my specimen is chemically and 

 physically much nearer variscite than calaite. Anyhow, the assump- 

 tion that this mineral is nothing but wavellite cannot be maintained. 

 If not a perfectly definite species, yet its behaviour when heated to 

 100° marks it out from wavelUte and from nearly all the other phos- 



2 Tchermak's Min. % Petr. Mltth. ii., p. 229, et seq. 



