44.0 Dr. A. Holmes and Dr. S. Smith— 
The number of analyses taken for each individual type is, we 
believe, in most cases larger than has hitherto been used in dis- 
cussions of this nature. It will be seen that the value given by the 
apex of each curve varies slightly from the average silica percentage. 
We suggest the former value as more appropriate, since rocks 
obviously wrongly named have been eliminated in making the smooth 
curves. 
DiscussIoN OF THE CURVES. 
A few points of interest are evident in the graphs. 
The considerable lateral range of a curve may indicate either 
looseness in naming, or lack of definition as regards silica-percentage. 
The absence of sharp demarcation between types, which we find 
in nature, is well brought out. The overlapping of the curves is 
largely due to the influence of those factors, other than silica- 
percentage, which enter into the present system of classification. 
The relative basicity of the felspathoid branch of the igneous rocks 
is well shown. 
The diorite curve, plotted on the basis of all recorded analyses, has 
two well-marked apices, the one at 51, the other at 57. As this feature 
is too pronounced to be smoothed over, a closer examination of the 
data has been made. It is found, on considering separately such 
analyses of diorites as are recorded in Huropean and American 
publications respectively, that these give two totally distinct curves 
(Fig. VII). It will be seen that the “ Huropean ” curve is much the 
more basic of the two. In fact, in Europe, diorite, as opposed to 
quartz-diorite, generally means a basic, rather than an intermediate, 
tock. This is all the more remarkable since rocks named diorite- 
gabbros and gabbro-diorites have been excluded in the drawing of 
the curves, Hven taking all the diorites together it is clear that their 
mean silica-percentage is markedly lower than what is usually given 
as typical for a diorite analysis. 
It is not our intention to dogmatize on the curves, however, but 
to make them available to those who may be interested in particular 
points. . 
In conclusion we would like to thank our colleague, Mr. H. B. Bailey 
for his help and advice in this compilation. 
The Wackerfield Dyke, County Durham. 
By Arraur Houmes, A.R.C.S., D.Sc., F.G.S., & STaNLEY SMITH, 
MAY DSc. 0.G.8. 
(With analysis by H. F. Harwood, M.Sc., Ph.D.) 
(PLATE VIII.) 
fete dyke to be described in this paper has not hitherto been 
noticed in petrological literature, although the western part 
has been recorded on the map of the area by the Geological Survey. 
It exhibits, nevertheless, some interesting features and adds data 
to our knowledge of the minor intrusions of north-east England. 
