ITALIAN PETROLOGICAL SKETCHES 369 



tain alkali feldspar, oligoclase, labradorite or anorthite, and these 

 groups are subdivided according to their ferromagnesian min- 

 erals, biotite, hornblende, or pyroxene. Neither trachyandesites 

 nor trachydolerites are mentioned. 



It must also be noted that a somewhat similar scheme was 

 proposed by H. O. Lang^ in 1891 and carried out by him in 

 great detail. He bases his purely chemical classification entirely 

 on the relative proportions of the percentages of lime, soda, and 

 potash, neglecting the other components and the mineralogical 

 composition, so that the present grouping and his differ dis- 

 tinctly from each other. 



LEUCITIC ROCKS. 



Comparatively little can be said of these rocks here, since 

 detailed descriptions of manv of them have been already given, 

 and since a proper treatment calls for more space than is avail- 

 able. As regards classification, while they are on the whole 

 sharply separated from the trachydolerites, yet inter se they 

 grade into each other to such an extent that the discrimination 

 of definite types is more than usually difficult and subjective in 

 character. 



It will also have been evident from the descriptions given 

 that many of the varieties described differ from the types gen- 

 erally accepted. Thus those rocks which have been called leu- 

 cite-phonolite for want of a better name contain only a small 

 amount of nepheline and are quite distinct from the t3'pical 

 German leucite-phonolites. The unfortunate difference of 

 opinion as to the use of the terms leucite-trachyte and leucite- 

 phonolite has been noticed, and in other ways the need is made 

 evident of some general revision of all this group of important 

 rocks. At present, however, the material is not at hand for such 

 an attempt, which it is hoped will be undertaken later ; and the 

 names used are to be regarded as provisional only, and the 

 whole group of leucitic rocks as the subject of future investiga- 

 tion. In Table V, only a few of the many analyses are given, 



'H. O. Lang, Min. Petr. Mitth., XII, 199, 1891. 



