On the Drying-up of Bivers. 201 



attributed it to the high twwrf obstructing the flowing of the current ; 

 others, to X\\g frost in forming barriers of ice on the caulds or dam- 

 heads ; others, again, had suggested that the phenomenon might be 

 connected with an earthquake. In support of this last theory, it 

 was mentioned, that Professor Phillips had, in a recent work on 

 geology, attributed to this cause the drying-up of the English rivers 

 Trent and Medway, in the 12th century. 



Mr IMilne stated that he adopted none of these views, and that 

 he thougiit tiie phenomenon might be accounted for by the united 

 action of tiie frost and wind which prevailed during the night of 

 the 26th November. After four o'clock that afternoon, the thermo- 

 meter all over the south of Scotland sunk to 2G°, at which point 

 it remained for several iiours. Accompanying this frost, there 

 was a gale of wind from the east, wliich had the effect of very 

 rapidly reducing the temperature of exposed and unsheltered spots. 

 In this way, tlie small and shallow streams flowing in open dnuns 

 and rivulets, or oozing through mosses and marshes in the hills, 

 were soon frozen and arrested. But, on the other hand, larger 

 bodies of Avater flowing rapidly in the main channels, at a lower 

 level, and sheltered by high or wooded banks, could not in the same 

 space of time lose enough of their temperature to be frozen. The 

 waters thus ran off, without the usual renewal of supplies from the 

 sources, so that the channel or bed of the river became speedily 

 drained. 



The reason of this phenomenon not happening more frequently 

 appears to be, that there is very seldom a gale of wind in this coun- 

 try accompanied by a severe frost ; and even on this occasion, the 

 frost was not equally intense over the whole island. When a se- 

 vere frost sets in, there is usually but little wind, so that the water 

 in the upper parts of the river, is not liable to be cooled more ra- 

 pidly than in the lower and more sheltered parts of its course. 

 Though the sources will, in that case, to a certain degree, be frozen, 

 and so, part of the usual supply cut off, the main body of the stream 

 is frozen likewise, whereby the velocity of its current is diminish- 

 ed, by the obstruction of the ice at the bottom and at the surface 

 of the current. So that if only half the usual supply is furnished 

 to the river from its partially frozen sources, there will be no di- 

 minution in the quantity of water flowing in the main bed of the 

 river, if it flows off with half its usual rapidity. This is the ordi- 

 nary way in whicli frost acts on the rivers, in this country. But 

 when, as on the night of the 26tli November, the frost is accom- 

 panied with a strong and keen wind, whicli lasts for only a few 

 hours, it freezes the water in the small rivulets near the sources of 

 the rivers in high and exposed situations, whilst it has not time to 

 freeze even the surface of the deeper and more rapid cmrents flow- 

 ing in the lower parts of the rivers. 



The easterly gale whicli, by its low temperature, produced this 

 phenomenon, continued to blow until about 7 or 8 a. i\i. on the 

 morning of the 27th November. The temperature of the at- 

 mosphere then underwent a sudden change, as indicated both by 



