part 3] THE GIACTATION OF NORTH-EASTERN IRELAND. 417 



from east to west. Kilroe's paper is illustrated by two maps 

 {op. cit. pp. 828 & 831) showing his interpretation of the ice- 

 movement during the Scottish and Irish glaciations : these maps 

 are reproduced in figs. 10 & 11, for purposes of comparison with 

 my own interpretation. 



Kilroe appears to have made the assumption that an ice- 

 sheet flowing across the grain of a country would produce only 

 striae parallel to its general direction of movement, or, in other 

 words, that the movement of the lower layers of the ice, entangled 

 amid the irregularities of hills and valleys, would conform to those 

 of the main mass above. 



We need not be surprised that the conclusions arrived at in the 

 paper under consideration should prove to require some modifi- 

 cation, seeing that they are based upon the study of the striations 

 alone, withoirt due regard to the nature of the drift deposits or to 

 the transport of erratics. 



The drift-deposits and erratics have been studied by several 

 observers, and many descriptions of local sections have been 

 published. The most important of these are to be found in the 

 Memoirs of the Geological Survey of Ireland, also in the reports 

 issued from time to time by the Geological Section of the Belfast 

 Naturalists' Field-Club, and published in the Transactions of that 

 Society, to which I am indebted for many records of erratics. 



The first references to the overflow-channels or ' dry gaps,' as 

 being connected with glaciation, are to be found in the Drift Map. 

 of the Belfast District, published by the Geological Survey in 1904,. 

 and in the accompanying Memoir. 



That there was something abnormal about sOme of the valleys 

 in the North of Ireland appears to have been noticed by Joseph 

 Nolan, of the Geological Survey, although he did not connect 

 them with the action of ice. In the explanatory memoir to. 

 Sheet 34 (1878), on p. 8, Nolan writes: 



' These tablelands [near Pomeroy] are intersected in every direction by 

 deep winding- valleys and ravines, which sometimes present very bold and 

 striking characters. Since the formation of the older of these valleys and 

 ravines the physical geography of the district appears to have undergone 

 considerable alteration, so that it is not unusual to find a ridge of hill cut 

 through by a deep ravine, the denuding agents having operated in a direction 

 at right angles to that of the original valley. Bernisk Glen, some 4 miles 

 south of Carrickmore, is a remarkable illustration of this.' 



By studying and combining these three types of evidence it is 

 possible to arrive at fairly definite conclusions as to the sequence 

 of events, although in a district so complicated in structure and 

 relief as that now under consideration, much in the nature of 

 minor detail must remain doubtful. 



From the careful study of a great mass of observed detail, only 

 the leading features of which are described in the foregoing pages, 

 there emerge certain main conclusions which will now be stated. 



During the early stages of the glaciation the dominant agent in 



