KiNAHAN — A New Reading of the Donegal Rocks. 21 



Rocks, as represented in the accompanying diagrammatic section 

 {see fig. 3, Plate IV.). ^ Moreover, in places in the Great quartzite 

 we find litoral deposits on different levels that must have ac- 

 cumulated in the vicinity of land, while the Grreat quartzite seems 

 to he of different thickness, as it is followed from the north-east 

 into the country to the south-west. 



To the east of Mulroy Bay, at Ballyhork and Croaghan, in 

 Fanad, there are massive Boulder-heds and finer conglomerates 

 apparently under the Grreat quartzite. There is no junction be- 

 tween them and the Older Period Rocks exposed ; but they evidently 

 are made up of detritus from the latter. The matrix is always 

 more or less greenish, and in general it is brittle, weathering 

 freely; but sometimes it is very quartzose and hard, with the 

 boulders few and far between. The latter are for the most part 

 coarse granites similar to those found in the neighbourhood of 

 Ballyhorrisky, to the north-west, and Glen Lough, to the east of 

 Mulroy Bay. There are also finer granites, granitic gneiss, 

 micaceous or gneissose quartzite, like that in the hill north of Bally- 

 hork, hornblende rock (diorite), dolomite, and vein quartz. In 

 general the largest and more abundant blocks and fragments are 

 of granite ; but in some places they are principally of the gneissose 

 quartzite and vein quartz. In Crockmore, south of Croaghan, 

 and in Ballyhork, there are, in the Boulder-beds, interbedded 

 quartzites ; while to the south of Crockmore the massive Boulder- 

 bed seems to suddenly thin out and be replaced by quartzite. 

 Under the Boulder-bed, both at Croaghan and Ballyhork, there are 

 dolomites and limestones, some of which appear to belong to it 

 while others may possibly belong to the Older Period Rocks. But 

 the sections are so obscure that, from my brief examination, it is 

 impossible to say anything positive as to their age.^ 



Other Boulder-beds occur inter-stratified with the Great 

 quartzite of the Knockanteenbeg outlier, which lies about eight 



1 This explanation of mine has been said to he a "physical impossibility," yet 

 exactly a similar overlap, but, of course, ou a much grander scale, occurs in Central 

 Wisconsin, U. S. A., the Potsdam sandstone lying on the Huronians at Baraboo, while 

 due north, near the Grand Eapids, they lie direct on the Laurentians ; the absent strata 

 between these two localities being, in other places, thousands of feet in thickness. 



^ In general limestone does not occur at the base of the Great quartzite, but near 

 Lag, N. N. "W. of Carndonagh, in Inishowen, it does occur. Here it may also 



