Vol. 60.] OF THE EAST-CEXTRAL HIGHLANDS. 443 



the felspar with the dark dust is still seen to he present. A good 

 illustration of the failure of the newer intrusions to affect the 

 Highland metamorphism is afforded by the two specimens (11,137 

 and 11,136) selected to fix the upper limit of the lloine Gneiss. 

 The first is a small infold on the north side of the Tarf, and a con- 

 siderable distance from any granite. The second was taken near 

 the margin of the Glen-Tilt diorite, and forming really part of its 

 thin roof. The metamorphism of the two is substantially identical ; 

 indeed, it is not easy to obtain two rocks, so far apart, which have 

 so nearly the same composition, and show so exactly the same 

 metamorphism. 



The published Geological Survey-Maps of Scotland (sheets 66 & 

 67) equally show that the course of the great ' sillimanite-aureole ' 

 is entirely unaffected by the Kincardineshire granite, for the aureole 

 meets the margin of the intrusion at right angles on its eastern 

 side. 



Explanation of Maps and Section. 



In order to understand the meaning of the maps and section 

 that accompany this paper, it is necessary to realize that the out- 

 crops here shown of such a rock as the Central-Highland Quartzite 

 are not the outcrops of an ordinary bed. They are really the 

 outcrop of a great sheet formed by the repeated folding of a bed on 

 itself, after the manner of the bellows of a concertina when shut up 

 (concertina-structure). 



This concertina-structure was produced by the first and greatest 

 folding of the Highland rocks, and to it is due the erroneous idea 

 that the latter were of great thickness originally. A section drawn 

 across the country, after this folding was completed, would closely 

 resemble that drawn through a comparatively-undisturbed area, 

 except that the original beds have to be replaced b}~ these horizontal 

 sheets. The structure has been considerably blurred, in many cases 

 by later movements ; but over large portions of the typical ' lloine- 

 Gneiss areas,' this sheet-structure must be still retained, for these 

 gneisses cover an area of several thousand square miles, and must 

 obviously, when viewed on a large scale, be still roughly a horizontal 

 sheet. To the south-east of Glen Tilt these sheets have lost this 

 horizontality, and been thrown into anticlines and synclines that 

 give rise to the ridge-and-valley scenery referred to in the section 

 on the ; Succession in the Braemar Area ' (p. 423). 



This type of folding, however, attains its full development only 

 in the harder bands, which must, moreover, have a certain thickness 

 before its development is possible. A perfect illustration of these 

 principles is afforded by the little sill of hornblende-schist shown in 

 the section across Glen Tilt (fig. 0, p. 444), the thickness of which 

 has to be greatly exaggerated to enable it to be shown. But in one 

 place, owing to a sudden increase in its original thickness, it was 

 able to fold on itself, and form a homogeneous mass \\ miles 

 long, and 300 yards broad at the observed outcrop, having a 



