10 ME. T. r. JAMIESON- SUPPLEMENTARY 



The Ordnance Survey had not then extended to the district, and no 

 good map existed to show the topographical features with accuracy. 

 Agassiz insisted that all the three Glen l^oy lines were marked on 

 a certain part of the south side of Glen Spean. " These three 

 terraces," he says, *' though in a less perfect state of preservation, 

 are repeated for a short distance at exactly the same level on the 

 southern wall of the valley of Glen Spean, just opposite the opening 

 of Glen Roy." ^ This, however, appears to be a mistake. At least 

 I could find no trace of the two upper lines in any part of Glen 

 Spean, neither have the Ordnance Surveyors or any other person 

 been able to discover them, and I suggested that during the period 

 to which they belong all the lower part of Glen Spean might have 

 been filled with ice, and hence the absence of the two upper lines 

 there. I thought, however, I could perceive a faint trace of the 

 uppermost line in Glen Glaster, even when checking my observation 

 by the spirit-level ; but the Ordnance map makes it stop just above 

 the mouth of that ravine and not enter it at all. Nevertheless- 

 Mr. Milne-Home ^ in his last paper maintains that it does occur 

 there, so that the matter is not altogether free from doubt. 

 If it does, it is at all events very obscurely marked. Although 

 this may be thought rather a matter of minor detail, yet the 

 point is one of some importance in the explanation of the phe- 

 nomena, and a good deal turns upon it, because if the water had 

 stood for any length of time in Glen Glaster at the level of this 

 uppermost line we would require to have something blocking the 

 col at the top, otherwise the water would have escaped over it, and 

 could never have stood above the level of the middle line which 

 corresponds with the Glen Glaster col. In my former paper I 

 assumed that such a block was necessary, and showed that it could 

 be accounted for by the Treig glacier, which during a certain stage 

 of its existence protruded right across Glen Spean and rose to a 

 great height upon the hills on the opposite side of the valley, thus 

 barring the passage to the eastward. But if it turns out that the 

 highest line is really quite absent in Glen Glaster, then this blocking 

 of its col is not required. Depending on the accuracy of the Ord- 

 nance Survey I shall assume such to be the case. I shall also- 

 assume that the middle line does not extend farther down Glen 

 Roy than where the Ordnance map shows it, namely opposite the 

 middle of Bohuntine, and consequently that Agassiz was wrong in 

 supposing that the two upper lines go down to Glen Spean. These 

 assumptions relieve us of certain diflficulties and enable us to sim- 

 plify the explanation, as we shall afterwards see. 



1 have already mentioned that the era of the lakes must belong 

 to the decay of the last sheet of ice if there was more than one. A 

 change of climate was taking place, the long secular winter had 

 begun to pass away, and when the thaw set in the thinner ice away 

 to the north-east would of course disappear first. The col or water- 



^ 'Atlantic Monthly* for June, 1864, p. 730. See also the map in his- 

 original paper in the Edin. New Phil. Journ., October, 1842, p. 237. 



2 Trans. Roy. Soc. Edin. vol. xxriii. (1879) p. 97. 



