22 ON THE PARALLEL HOADS 



cularlj at the cascade of Tivoli, where this substance is con- 

 stantly accumulating *. 



As the ancients mention the existence of the lake, which 

 once filled the valley of Subiaco, but take no notice of the 

 cause by the operation of which it was drained, it must have 

 occurred at some period between the fiftli and the fourteenth 

 centuries. It is as impossible, as it is unnecessary foi* me, to 

 offer even a conjecture about the nature of this ; but I hope 

 enough has been established to prove, that the valley of Subi- 

 aco furnishes a decided and well-authenticated instance of a 

 range of shelf, in every respect like those of Lochaber, and 

 which has been unquestionably formed by the waters of a lake. 

 From the particular description I am to give, it will appear 

 that the resemblance between the Glen of Subiaco and those 

 of Lochaber, is rendered extremely strong by a number of cir- 

 cumstances, common to all of them ; perhaps the most re- 

 markable of these is, that the waterfalls and ravines of Su- 

 biaco, are also found placed in similar situations in the Loch- 

 aber glens. 



Particular Description of the Shelves, as connected with the Geo- 

 graphy of the Country, and the Appearance of the Glens in 

 •which they are found. 



I PKOPOSE to arrange the more particular description of the 

 shelves, which I am now about to give, in the following order, 



which 



* When I was lately at Tivoli, I had the most earnest desire to visit Subiaco 

 in person, not only to make my own observations on the valley, but hkewisc to en- 

 joy a range of scenery, described to me as being of the most romantic chai'acter ; 

 but I regret to say, that a vaiiety of circumstances compelled me to abandon the 

 attempt, with very great reluctance. I had, however, some satisfaction, in avail- 

 ing myself of the oppoi-tunity of corroborating the description in the text, by the 

 tesdmony of one or two of my Roman friends, who had been at Subiaco. 



