554 



Proceedings of the Royal Irish Academy. 



phates of aluminium. "While in vacuo it loses nothing, at 100° C. (in 

 the air-oven), all the water in the mineral is disengaged. In the water- 

 oven the change is incomplete, for at 94°, the maximum temperature 

 reached by my water-oven, rather less than half the total water was 

 removed. It may be noted that the curious pink tint which peganite 

 assumes when heated in a bulb-tube was acquired in the air-oven at 

 100°, but not in the water-oven at a temperature only a few degrees 

 lower. 



My analysis of a sample of this so-called peganite from Striegis, 

 gave the following results : — 



Anal. 6. 



374 gram lost nothing in vacuo, but lost 

 035 ,, HoO in water-oven, and 



047 

 005 

 133 

 243 



HoO in air bath at 100°, and gave 



SiO, 



ALO3 and 



MgoPaO^. 



Translated into percentages and compared with the numbers de- 

 manded by the nearest formula, these results are here shown : — 





Experiment. 



Theory. 

 7Al203,6P20s,24H20. 



Alumina, 



Phosphorus Pentoxide . 



"Water, 



Silica, 



35-30 



41-56 



21-92 



1-33 



35-91 

 42-53 

 21-56 



100-11 



100-00 



"When the correction for intruding silica is made, the correspon- 

 dence between experiment and theory becomes quite satisfactory. 

 Still it would scarcely be justifiable to accord specific rank to this 

 mineral on the strength of a single set of numbers, and I should pre- 

 fer to regard the specimen analyzed as variscite slightly admixed with 

 a more basic aluminium phosphate. 



§ VII. It is, on the whole, evident, from the analyses of gibbsite, 

 variscite, and peganite, which I have now given and discussed, that 

 any native basic hydrate of aluminium, if it were made up, say, of one 

 molecule of the normal native hydrate simply associated with one 

 molecule of the normal native phosphate, might be expected to lose, 

 when heated to 100°, all the water attached to the latter compound, 

 and none of that belonging to the former. 



