556 Proceedings of the Royal Irish Academy. 



nothing but wavelKte ; another is very nearly related to variscite, if it 

 be not identical with, it.^ The peganite of Arkansas, according to 

 A. H. Chester,® is also identical with variscite. But there is a third 

 mineral known as peganite found also at Striegis in Saxony ; this is 

 very near wavellite, but contains less water. "What is probably the 

 same mineral from Nobrya'' in Portugal has been analyzed by Lichten- 

 berger and Frenzel. As to fischerite, I have not been able to secure 

 an authentic specimen sufficiently ample for analysis. 



§ X. The next group of native aluminium phosphates may now be 

 discussed. Its best known member is wavellite, to which the formula 



2AI0P2O8, Al.HeOg, 9 aq 



is usually assigned. But in spite of the very numerous analyses of 

 wavellites from different localities which have been made, the formula 

 for this species cannot be said to have been ascertained. Two circum- 

 stances make it doubtful. Like many other fibrous minerals, wavellite 

 always retains, under ordinary atmospheric conditions of barometric 

 pressure, moisture, and temperature, about two per cent, of moisture 

 removable in vacuo, or in dry air, or by a sKght increase of temperature; 

 so the question arises "Is this water essential or accidental?" Doubt- 

 less many hydrated minerals, when removed from the natural condi- 

 tions under which they were formed, lose essential or constitutional 

 water very readily, but in such cases the percentage of loss is generally 

 much larger than two. The second circumstance which interferes with 

 the exact determination of the combined water in wavellite is the 

 presence of fluorine in this mineral. This element, which probably 

 occurs to the extent of two per cent,, must be regarded as an acid ele- 

 ment, replacing either the phosphoric constituent or oxygen. If so, 

 it will lower the proportion of phosphorus pentoxide found, and raise 

 that of the water as determined by loss. I have considered this ques- 

 tion in a previous research,® and shown that a formula with llHoO 

 has much to recommend it. It may be added that vacuum-dried 

 wavellite loses no water at 100° C, but 22% at 200° C, and 4% at a 

 low red heat. And if those published analyses of wavellite in which 

 the fluorine has been determined be studied, it will be seen that the 

 water is lower than that commonly assigned to this mineral. But, 

 after all, it is not improbable that we group under wavellite several 

 minerals differing from each other by 1 aq. Possibly there are four 

 such members of the wavellite group : — 



1. Planerite, .... 3ALO3, 2P2O5, 9aq. 



2. CoeruleolacUte, . . SAl.Oa, 2P0O5, 10 aq. 



3. Wavellite, .... SAl-A, 2P20g, llaq. 



4. 8triegisane, . . . SALOa, 2P2O5, 12aq. 



Of Coeruleolactite I shall have something to say further on, but it 



5 See § VI of this Eeport. ^ Jahr. Min.. 1872, 819. 



''An analysis is given in §XII. ^Jour. Chem. Soc, Feb., 1873. 



