650 



HENRY S. WASHINGTON 



lected. The same is true of the pulaskite analyses, these rocks 

 lying quite outside the Magnet Cove area. All these rocks will 

 be discussed subsequently. 



We thus obtain the result shown in Diagram 2, where the 

 points are connected by straight lines. In this, and the follow- 

 ing, the vertical scale for SiOg begins .400 lower than the others, 

 so as to condense the diagram and at the same time preserve 



ALO3 



MgO 

 (FeO) 



Na„0 



SiO, 



Diagram 2. 



the relative forms of all the curves (given later), and not flatten 

 that of silica, as would be done if a smaller vertical scale were 

 used for this than for the other constituents. 



When Diaa^ram 2 is examined it is clear that, with the 

 exception of the values for covite (V), all the points of the 

 respective oxides lie along very smooth curves. For con- 

 venience in further discussion the curves formed by the figures 

 for 11, III, IV and VI are plotted separately in Diagram 3. The 

 values for V (covite) are entirely omitted from this, and the 

 position and relationships of this rock will be discussed later on. 

 The curves marked /^and M will also be explained presently. 

 All the curves, it may be mentioned, were drawn with a spline, 

 so that the personal equation is eliminated as far as possible. 



