EVOLUTION OF THE CLUB TYPES. BL] 
shatter under impact; the wood is less dense in structure, but the dis- 
position of its fibrous bundles renders it less liable to shattering; when 
the two are properly adjusted one to the other the tough wood lends 
valuable support to the more fragile stone or shell, and in the case of 
this particular shell the density is so great that we may regard it as 
truly a dense limestone produced by animal rather than geological 
causes. Next we find great significance in the carved ridge across the 
angle between lip and head, which is found in all 5 pieces of this 
species. So long as we confine our investigation to these 5 clubs we 
derive the conclusion that the presence of this constant is indicative of 
a transverse lashing which supported the stone blade in two senses 
equally, one holding it taut against the shaft longitudinally, the other 
establishing its support against the wooden shoulder at its end. We 
shall see still more of this band in the species next to be examined. 
Decorated-panel species—In the former species the surface of the 
panel was regarded as expressive of the difference in finish between the 
blade material and that of the shaft. Here we find the introduction of 
an added unit, which gives us the expression of another motive. Reck- 
oning downward from the top of the head, we have first a surface which 
seems intended to express the continuity of the shaft, then a highly 
decorated panel, last of all along the lower edge in 2 out of the 5 
pieces a blank stripe longitudinally from the face of the head to the 
very end of the panel at the strongly marked shoulder on the shaft. 
This stripe is continuous with the blade on the face of the head and a 
trifle wider. I interpret this combination in the sense of a stone or 
shell, and the dimension of thickness rather suggests the shell source, 
set up against the rectangularly notched shaft with support up and 
down at the proximal shoulder, longitudinally on the upper line by the 
shaft and lip and upper face of the head, and with added support by 
slips of wood protecting its lateral dimension on either side. To this 
we add in 3179 (fig. b) the carved band in the angle of lip and head as 
representing a lashing as explained in the preceding type, and this 
lashing is continued down the face of each of the sculptured side- 
pieces in 3179 by a double tie of band-and-zigzag which we can com- 
prehend only as representative of sennit. The same sort of tie on the 
side-pieces appears in 3186 c, but there lacks the determinant associa- 
tion with a band athwart the lip-head angle. This arrangement looks 
particularly toward the helving type of stone axes in which a subsidiary 
_ socket fixture is employed. 
Rugose-panel species.—In this species we note the following impor- 
tant divergences from the motives of the foregoing species. The band 
in the angle is absent; up the shaft is no shoulder against which the 
panel might end; the panel gradually loses itself indistinguishably in 
the polished tract of the shaft both at the end and along the upper 
edges; the rugosity reaches up practically to the edges of the upper 
