324 On the Distribution of Temperature in 



44-6 F., while that of the surface of the lake in its neighbour- 

 hood was 49-0. On I3th November the temperature of the 

 stream at Inversnaid was 40, and that of the lake surface 46. 

 From its rise and course the Falloch is more likely to be colder 

 than warmer than these streams, so that even in October it 

 must have begun to spread its cooling influence over the lower 

 waters of the Ardlui basin. When the water of the stream is 

 warmer than that of the lake surface, it passes away with the 

 drainage, and imparts as much of its heat to the atmosphere 

 above it as to the water below. When its temperature is lower 

 than that of the lake surface, and in all probability it is so for 

 more than half the year, it sinks into the body of the lake, and 

 imparts its cold entirely to its deeper waters. It is obvious 

 then that during the time that it is colder than the lake, the 

 water of the Falloch must produce a much greater effect on 

 it than during the opposite season; hence the position of the 

 Ardlui station, with respect to the principal tributary of the 

 lake, renders it natural to expect that its waters would be 

 colder than they are found to be in the Luss basin, which, from 

 its size and position, is comparatively exempt from the direct 

 influence of tributary waters. 



Both in September and in October the temperature of all 

 these bodies of water is lower at Ardlui than at Luss, but the 

 contrast is much greater in the colder than in the hotter month. 



In the Tarbet basin a salient feature is the greater mixture 

 of waters at Inversnaid than at either Rowardennan or Stuck- 

 gowan, the curves at the latter localities being much steeper 

 than at Inversnaid. The same phenomenon was observed on 

 the i8th August. But, perhaps, the principal feature in the 

 Tarbet basin is that both at Inversnaid and at Stuckgowan the 

 water at and near the bottom has risen in temperature by 

 0-2 F., or from 41-8 to 42-0 since the 22nd September. As 

 the temperature of the bottom water at these localities was 

 41-8 on the 5th September, it is probable that it had been so 

 during the summer, and it is only by the end of September or 

 beginning of October that the summer heat begins to have any 

 effect on the water near the bottom. The steepest gradients 

 are at Rowardennan and Stuckgowan between 20 and 30 



