94 THE DESCENT OF BALLADS

Now there was Thor of Sea-garth,
with the trolls he holds good Thing:

full fifteen trolls and forty-five
they lay there in a ring.

Thor, he tameth his foals on the heath.

The jesting note is present in the Thrymskvida, but it is more dis-
creet. The singer finds these gods, in whom he only half believes,
amusing in their thick-headedness and petty vanities, like the gods
of Homer. The broader humanity and generalized situations make
of the ballad a new creation, different in class, though partly
identical in language and content, from the aristocratic epyllion.
There is no comparison quite so close in Spanish experience.
The epics of Spain consist of the Poem of the Cid, the Rodrigo, and
a fragment called Roncesvalles. The former is extant in a manu-
script executed, without great care, in 1307, but the original must
have been composed about 1140. Only one of the ballads of the
Cid answers to this poem, namely the ballad of Bucar, The resem-
blance is not close; Biicar is killed in the Poem, but survives in the
ballad. In the First General Chronicle Biicar is made to survive,
because in fact he did outlive the Cid; it has been conjectured that
this chronicle used a refundition of the old Poem, and it has been
noted that the ballads correspond generally with late refunditions
of the epic material.1 If that be so, the hope of directly compar-
ing ballad with epic disappears. The Rodrigo has the advantage,
under this hypothesis, of being a late poem; but its state is so de-
plorable, so unlike poetry, that the single instance in which com-
parison is possible becomes fruitless. There are two ballads which
may be conjectured to be part of the Roncesvalles poem, but they
do not correspond with the 100 extant lines. If we augment our
material by admitting that the lines reconstructed by $r. Menendez
Pidal from the Madrid National Library's manuscript chronicle
(FiSz) are genuinely epic—as they may well be—we have still to
face a discrepancy between the circumstantial manner of the epic
lament and the elliptic ballad style.2

1 R. Men&idez Pidal, in his edition of the Cid, his Leyenda de los Infantes de
Lam, and in his note on the Fenian Gonzalez cycle in the Homenaie a Menendez
y Pelayo.

3 The discrepancy has been noted by M. Men£ndez y Pelayo, Tratado de los
romances vigos9AntologiadepoetasUrico$castellanos, Madrid 1934 xi pp 276

?WSee aiSOLS* Griswold Morl^> Spanish Ballad Problems, University of
California Publications, 1925. y