138 Pke-Histoeic Forts. 



young' larches and firs with their thick growth have quite con- 

 cealed them again. It is difficult, indeed, to make one's way 

 among them, and quite impossible to realise thoroughly how well 

 this lofty platform has been defended by nature and by the hand 

 of man. But it is easy to see that if this hill-top was chosen as a 

 post of observation or as a signalling station, no place better 

 adapted for the purpose exists in all the country round. The 

 whole of Lower Nithsdale lies beneath the eye. The Tinwald and 

 Torthorwald Hills, with the square top of Birrenswark appearing 

 above them ; Glencairn and Glenesslin, and the mazes of mountain 

 land beyond them on the border of Galloway ; the Lowthers and 

 Queensberry, Criffel, and, if it be clear, Skiddaw on the other side 

 of the Solway, are all within sight. 



In striking contrast with the " camp " on Springfield Hill is 

 the Earthwork at Snade, about as far from the village of Dunscore 

 in the opposite direction. Instead of being planted on a hill-top it 

 lies low in the valley of the Cairn. Here no natural points of 

 strength have been seized and skilfully adapted for defence, but 

 everything has been laboriously constructed by the hand of man. 

 Though numerous steep places, which would readily lend them- 

 selves to fortification, are to be found in the neighbourhood, a site 

 has been chosen in the level holms beside the river. The 

 entrenchment is not readily seen, for it is hidden in a grove of 

 oaks like that which surrounds Dardarroch on the opposite bank 

 of the Cairn. These trees would seem to be the descendants of 

 the primeval forest which once filled all the valley and extended 

 up through Glenesslin, giving its name to Holywood (Sacrum 

 Nemus — Sacrobosco), and, like Sherwood Forest, sheltering 

 outlaws and robbers. The earthwork lies on the western bank of 

 the Cairn, about 300 yards from the river, and a short distance 

 above the ford which crosses from Cambuscairn to Snade. Two 

 deep trenches surrounding a level platform are the most striking 

 features, and, from the pi'oximity of a stream which passes fifty 

 yards away, it is probable that they were kept constantly filled 

 with water. There are indications, indeed, that the course of this 

 rivulet has been altered for agricultural reasons, and that at one 

 time it flowed close past the outermost trench. 



I visited this earthwork several times during the past few 

 months, the most recent occasion being on Saturday last, the 7th 

 of December. I was accompanied then by Mr Aitken, school- 



