F. W. Clarke — Minerals of Litchfield, Maine. 269 



In attempting to discuss the formulae of cancrinite, socialite, 

 and hydronephelite, certain points should be carefully borne 

 in mind. First, the three species must be considered, not inde- 

 pendently, but relatively to each other ; for all the evidence 

 indicates for them a common origin. That origin is from the 

 first member of the group, elaeolite or nephelite; the empirical 

 formula for which has been finally fixed by Kauff.* In par- 

 tially rational form it may be written JSra 8 A] 8 (Si0 4 ) 7 (Si0 3 ) Q ; 

 ignoring the small replacement of sodium by potassium which 

 has been shown by synthetic investigations to be non-essential. 

 Not only does the mode of occurrence and association of the 

 minerals point to community of origin, but the same conclu- 

 sion is emphasized by the experiments of Lembergf upon the 

 artificial alteration of silicates. When elseolite from Fredriks- 

 varn was digested 180 hours with a solution of sodium car- 

 bonate, a partial transformation into a soda cancrinite was 

 effected; while a digestion of six months with a caustic soda 

 solution containing sodium chloride gave a product identical 

 in composition with sodalite. Many such experiments were 

 tried by Lemberg, yielding a large class of similar results. 

 His method of procedure probably did not give absolutely pure 

 or definite compounds, and yet his researches furnish evidence 

 of great value in discussing the chemical structure of many 

 minerals. 



If we compare the published analyses of cancrinite from 

 different localites, we shall find that they vary in two ways. 

 First, there are variations which are probably due to small 

 admixtures of eleeolite, such as I have shown to occur at Litch- 

 field ; and secondly, the ratio between the lime and the car- 

 bonic acid ranges between rather wide limits. In the cancrin- 

 ite from Miask, the two are about equivalent; while the Litch- 

 field mineral contains only half enough calcium to saturate the 

 carbonic acid. The lime and soda, however, vary reciprocally; 

 so that when one is high, the other is low ; and, furthermore, 

 the experiment quoted from Lemberg goes to show that a can- 

 crinite may exist containing no lime whatever. If this conclu- 

 sion be correct, then the carbonic acid of the mineral must be 

 represented as linked with aluminum; a supposition which 

 finds some justification in the existence of the rare species daw- 

 sonite. The function of water in cancrinite remains doubtful; 

 if it be regarded as water of crystallization, the formula of the 

 residue becomes less easy to write intelligibly; but if it forms 

 a part of the atomic structure, it is almost necessary to repre- 

 sent the carbonic acid as orthocarbonic, in the group C0 4 . 

 This mode of consideration, as will appear later, leads to a 



* Zeitschr. fur Kryst., ii, 445. 



f Zeitschr. der Deutsch. Geol. Ges., xxxv, 557, 1883. 



