48 THE RUTHWELL CROSS. 
The one with its head down we have already noticed in con- 
nection with the Cundall-Aldborough shaft. This is obvi- 
ously a post-Anglian but pre-Viking Age work, and of the 
late gth century. 
At Kirkby Wharfe (Fig. 27) is a cross with late, expanded 
arms to the head; this expansion went on till at Whithorn 
we find the arm-ends nearly meeting, and thus forming a 
wheel. The arrangement of plaits in the centre of the head to 
side a was repeated through south-west Yorkshire, and car- 
ried to the Isle of Man, where it was adopted by that remark- 
able artist, Gaut Bjarnarson. The step-pattern and TILT, 
the ring in the plait of side c, the basket-plait and joined 
triquetrae, are all what became common in the roth century ; 
but the figures of SS. Mary and John beside the cross are of 
Anglian tradition. |The cross is transitional between late 
Anglian and Viking, and of about goo a.p. 
The same mixture of forms is seen in the shaft at 
Urswick-in-Furness (Fig. 28), which ‘‘ Tunwini set up in 
memory of Toroeotred (? Torhtred),’’ as the inscription in early 
runes states. Across the late rude figures is written “‘ Lyl 
this wo(rhte?) ’’—Lyl wrought this (?)—and the scroll, with 
its grotesque figures, birds and beasts, shows the Anglian 
tradition far gone in decay. It cannot be 12th century, for 
we know the history of Urswick after Domesday Book, and 
no such names occur. It must be of about goo A.D. or a little 
later. 
The stone found by Mr George Benson at St. Mary 
Bishophill Junior, York, built into the early Norman fabric 
(Fig. 29), shows the scroll turning into the ‘‘ snake-sling ’’ of 
the Viking Age; the berries dropped off and treated as pellets, 
and the leaves becoming snakes’ heads. In the 1oth century 
basket-plait a snake is inserted. The well-known Lancaster 
cross-head in the British Museum is another example, a little 
further developed, of the same transition, though its early 
runes record a purely Anglian name, Cynibald, son of Cuth- 
bert. The change in style was conditioned by period, not by 
race. 
In the Leeds cross (i'ig. 30) we have the debased Anglian 
scroll (6, 8), the plaits of the 10th century (16, 20), grotesque 
