loo Charles G. Osgood, 



rilis; nunc folHum flatus, nunc martellorum ictus, nunc cotis et 



ferri sonora fricamina, 



Stridentesque cavernis 

 Stricturas Chalybum, et anhelum f ornacibus ignem.' °^ 



This phenomenon Giraldus explains by the entrance and exit of 

 the sea. 



How Spenser came to confuse the Httle Barry with the big 

 Towy, which really does tumble down among the hills of Dyne- 

 vor, is not clear. Possibly the explanation is found in Holinshed. 

 Harrison, in describing the rivers of Great Britain, follows the 

 coast-line westward, discussing each stream and its tributaries 

 as he passes its mouth. In following the south coast of Wales 

 he has passed the Barry and proceeded westward, taking each 

 stream in order. He has just described the Gwendraeth Tawr 

 and the Gwendraeth Fach, the last before you come to the Towy, 

 and is on the point of proceeding with this stream, when, without 

 clear warning, he suddenly returns to resume his account of the 

 Barry. Perhaps Spenser, expecting next to read of the Towy, 

 with characteristic inadvertence may have taken Barry for an 

 alternative name of that river.®* 



The passage about the Barry, quoted above from Giraldus, 

 Spenser could as well have read in Camden, who quotes it in his 

 account of Glamorganshire (1590, p. 516). He also mentions 

 *Caer-Mardin, which the Britans themselves call Caer Firdhin, 

 Ptolomee, Maridunum' (Carmardenshire, p. 649). On the same 

 page he describes the Towy flowing 'by Dinevor, a princely 

 castle, standing aloft upon the top of a hill, . . . and last of all, 

 by Caer-Marden.' And further: Tn this Citie was borne the 

 Tages of the Britans, I meane Merlin : For like as Tages 

 being the sonne of an evill Angell taught his Countrimen the 

 Tuscans the art of Sooth saying, so this Merlin the sonne of an 

 Incubus Spirit, devised for our Britans prophesies.' Hence it 

 is unnecessary to suppose that Spenser drew any of the material 

 for the passage before us directly from Giraldus. 



To return to one or two considerations concerning the ProfJia- 



""Verg. A en. 8. 420. 



"Giraldus, Itin. 1. 10; Description of Wales, i. 5; and Camden, 

 Carmardenshire, p. 649, as well as the old maps, are all perfectly clear on 

 this point. 



