34 THE CONSTITUTION OF THE SILICATES. [bull. 125. 
only remains to work out the details for each individual mineral. The 
fundamental relations between the two groups have been recognized 
by many writers ; but their interpretations have been diverse. 
In the systematic treatment of the zeolites, the most serious diffi- 
culty is found in the hydration of the several species. To determine in 
any case what part of the water is constitutional and what crystalline 
is not easy, and no fixed criterion exists upon which judgment may be 
based. Water which is retained by a mineral above 300° of tempera- 
ture is almost certainly constitutional; water lost at or below 100° is 
probably w T ater of crystallization ; but between the two there is a range 
of considerable vagueness. There is, however, one datum which has 
claims to high consideration. Certain zeolites, at various tempera- 
tures between the limits indicated, lose water which is regained with- 
out change of crystalline character upon subsequent exposure of the 
minerals to moist air, and this, as has been urged by several authori- 
ties, should be regarded as water of crystallization. By considera- 
tions of this kind we can be guided to some extent, but beyond them 
each zeolite must be studied independently, in the light of such col- 
lateral evidence as may be available. For some species the evidence 
is full, for others it is meager; but such as it is it will appear in the 
systematic discussion following. 
In a recent publication * I have shown how all of the normal zeolites 
maybe formulated as substitution derivatives of the feldspars, although 
in some cases alternative formuhe are readily conceivable. Thus, from 
anorthite several species are directly derivable, as follows : 
Anorthite. Thomsonite. 
y Si0 4 =Al .Si0 4 =H 2 .AlH 2 2 
Al-Si0 4 =Al Al-Si0 4 nAl 
\^iO 4 =0a ^SiO 4 =0a 
I I 
Ca Oa +3H 2 G 
i I 
SiO 4 =0a ,SiO 4 =0a 
Al-SiOl=Al Al-SiG 4 =_EAl 
\siG 4 =Al X S10 4 =H 2 .A1H 2 2 
Gismondite. Foresite. 
.$i0 4 =H 2 .AlH 2 2 ,Si0 4 =Al 
Al— Si0 4 ~H 2 .AlH 2 2 M— Si0 4 =H 3 
\siG 4 =Ca \si0 4 =H 2 
I I 
Oa +4H 2 Oa +H 2 
I I 
SiO 4 =0a .Si0 4 =H 2 
Al— Si0 4 =H 2 .AlH 2 2 Al— Si0 4 =H 3 
^Si0 4 =H 2 .AlH 2 2 ^SiO.EZAl 
-Am. Jour. Sci. (3), XL VIII, p. 187, 1894. 
