Correspondence—C. E. Tilley. 559 
Staffordshire ; (2) the incoming of marine conditions in the Millstone 
Grits as they are traced northwards ; (3) the possible correlation of 
the Alton (Ganister) Coal and its marine roof with the Slatyband 
Ironstone of Scotland, a correlation adopted by Hull; (4) the 
probable Lanarkian age of the measures below the Kilburn Coal 
of North Derbyshire; (5) the many marine bands containing 
Productids in the Westphalian (Middle) Coal Measures of the Midland 
Province; (6) a considerable thickness of barren measures not 
infrequently of a red colour, high up in the Westphalian Coal 
Measures, but below the Staffordian Coal Measures, and their possible 
correlation with the barren red measures of Fife and Midlothian ; 
(7) the not infrequent occurrence of a marine bed resting directly 
on a seam of coal and the persistence of such seams of coal even 
when a few inches in thickness. 
CORRESPONDENCE. 
THE MAGNESIAN GROUP OF IGNEOUS ROCKS. 
Str,—The interesting paper appearing in your November number 
under the title: ‘‘ The Magnesian Group of Igneous Rocks,” raises 
a number of points which seem to require some further discussion. 
It is to a few of these/only that I wish to refer here. The application 
of the term “‘ magnesian ” to a group of igneous rocks, the ultrabasic 
members of which alone show notable content of MgO, is an 
innovation that appears scarcely justified. In the intermediate and 
acid types it is iron and not magnesium which calls for note. The 
peculiar properties of this group of rocks are of a provincial order 
in a petrographic sense, and serve to rank the group as a distinct 
unit of the calc-alkaline suite, rather than a group co-ordinate with 
this suite. Amongst the “ magnesian ’’-provinces listed by Mr. Dixey 
there are some whose inclusion requires some further word of 
explanation. Number 7, in particular, finds itself among strange 
companions. Some of the characteristics of these rocks are explicable 
on the assumption of a low active mass of water in the magma, 
combined in certain cases with high pressure during consolidation. 
A low magmatic water content may lead to an early precipitation 
of potash felspar, and we may note that orthoclase or microperthite 
is a not uncommon feature in the basic types, as in the N orwegian 
norites and the basic members of the Hyre Peninsula group. It 
is to this early precipitation of potash felspar that the occurrence 
of oligoclase or andesine as the plagioclase constituent of the 
mucroperthite in the basic types is due. This feature is best 
exemplified in the rocks of Western Norway. The occurrence of 
hyperstbene in the intermediate and acid types is probably explicable 
on these limes. In the absence of abundant water the degradation of 
the metasilicate molecule to the orthosilicate type is inhibited, so 
