EXCURSION VIII. 143 



The botany of this district is meagre, as is usually the case 

 with open uniform granite tracts; the only good plant is the 

 rarer species of Pyrus, which is found in rocky places on the 

 stream and at the base of the castles. The junction of the 

 fine granite with the coarse generally takes place where the 

 gradual slope ends, and the steeper climb to the mountain 

 tops begins. An important service would be rendered by 

 tracing this line along both sides and observing the appear- 

 ances whether they are similar to those already noticed by 

 iis on the south side of the district (Art. 14); whether veins 

 everywhere pass from the fine into the coarse. It would be 

 almost necessary to " camp out" two nights running, and it 

 is not easy to accomplish this in a climate so uncertain. On 

 the hill-sides N. of the watershed, Mr. "Wiinsch has recently 

 noticed that there are two terraces corresponding on opposite 

 sides of the glen, strewed with blocks of fine granite resting 

 on coarse, which have a striking resemblance to lateral 

 moraines. 



Having reached the summit-level on our ascent from the 

 north, we are at the curious little tarn Loch-au-Davie, which 

 discharges both ways, as already noticed (Art. 2.) 



In descending the lorsa about a mile, we pass off the fine 

 granite and come upon the coarse variety, at first in the bed 

 of the stream and then in both banks. Near the waterfall 

 a basaltic dike, twelve feet wide, crosses the bed of the stream 

 nearly at right angles, and intersects both granites. It is 

 indicated on our section (fig. 3), and was traced by us a long 

 way in both directions ; of many others there are indications 

 along the hill-sides below Loch Tanna, and on the walk to the 

 north-east of Sail-Chalmadale ; but the disintegration of the 

 dike and granite alike is so rapid that the width cannot be 

 ascertained with any exactness. In tracing these continua- 

 tions we must mainly look to the direction. In later visits 

 to the lorsa valley I felt satisfied that this range is from side 

 to side of the valley, and that the dikes which ascend from 

 the depths of Glen Rosa and cross the Ceims, notching their 



