
297 
side, which there can be little doubt served for a cord to pass 
through so as so hold the blade back to the handle.”* 
In order to utilize the whole length of the cutting edge at 
each blow, it seems as though this Celt had been fixed on the 
shaft of the handle with the loop away from the user. Socketed 
Celts are usually represented as fixed with the loop towards 
the user,t but if the one now described had been fixed in 
that manner, only a very small portion of the cutting edge 
would have been utilized at each blow. It is difficult now to 
determine exactly the method of fixing Celts of this shape at 
the cutting edge, as it does not seem ever to have been used, 
From my vertical section, figure 3, it will be seen that the 
width at the bottom of the socket is } of an inch greater than 
it is about half way down. This seems to show that the men 
who cast these socketed _ Celts were acquainted with the 
principle of foxtail wedging, a mode now used by joiners 
of spreading the end of a tenon, in a mortise, so as to give it 
a dovetail character to resist withdrawal.” 
DETAILS OF MEASUREMENTS. 



Narrow Way. INSIDE. Broap Way. 
12” wide At the mouth 14” wide 
1” wide 13” below the mouth 1}” wide 
2” wide 3” below the mouth 12” wide 
4” wide At bottom of socket 14” wide 
Narrow Way. OUTSIDE. Broap Way. 
18” wide At the mouth 1%” wide 
1}” wide In hollow under flange 14" wide 
—1}’ wide Under the loop 12” wide 
£” wide 3” below the mouth 18” wide 
Sharp edge At the bottom 25" wide 

* “ Ancient Bronze Implements,” p. 160. 
+ “Idem,” fig. 186, 
