494 Proceedings of the Royal Irish Academy. 



has been noticed by Sir (tben Mr.) A. W. Franks, in bis description of 

 tbe plates to Kemble's "Horse Ferales." 



Referring to tbe Arreton Down examples be says : — " At first sigbt 

 tbese would appear to be dagger-blades ; but with tbem was found a 

 socket of bronze fitting over a similar blade, and giving it tbe appear- 

 ance of tbe small spear-bead (fig. 19), wbicb seems copied from sucb a 

 design, andbas imitative rivets" (J. c. p. 152). Tbe spear-bead referred 

 to appears to be tbat in tbe Academy's collection (fig. 2) ; but tbe descrip- 

 tion does not agree witb tbe drawing given in tbe plate. 



Sir Jobn Evans bas also noticed tbe imitative rivets of tbese spear- 

 beads, and tbe resemblance to tbe Arreton Down type. He observes 

 tbat " tbe socket part is made to appear somewbat like a baft to tbe 

 blade, as in tbe Arreton Down specimen" (I. c. p. 325). 



But tbougb it appears from tbe preceding extracts tbat tbe survival 

 of tbe rivet-beads in imitative form, and in general tbe bafted character 

 of tbese spear-heads has been already noticed. Tbe general question of 

 tbe relation of certain spear- bead types to tbe dagger-blades has not 

 hitherto been taken up, or a classification based on such relations of 

 types attempted. 



The evolution of tbe true spear form appears to have proceeded by 

 the following stages : — 



1 . Tbe dagger-blade is mounted on a socket or hollow hafting which 

 does not enter the blade but is stopped at the base. 



2. The midrib which in tbe earlier forms is flat and rounded, as in 

 the daggers, become accentuated, and takes the form of a stiffening 

 ridge from socket to point. Tbe transition between tbe socket and the 

 blade is thus rendered less abrupt, and tbe baft-form of the socket is 

 less apparent (figs. 5 and 6). At this stage the central ridge of tbe 

 midrib tends to be prolonged along tbe socket, marking tbe unity of 

 tbe socket and midrib. Tbe central rib so frequently found along tbe 

 sockets of tbe looped class appears to be a survival of the ridge of the 

 angular midrib of tbe earlier forms. It was probably retained as 

 giving a cutting edge or bayonet-Hke section to the socket, and survives 

 on several of the leaf-shaped class. 



3. Tbe socket is now prolonged into the midrib, and the final spear- 

 head form determined (fig. 7). 



4. Tbe evolution of the socket being now completed, the tradition 

 of the dagger -form is broken, and the blade which bas hitherto pre- 

 served a dagger-like outline, develops in harmony with the conditions 

 of tbe improved form. In the earlier examples the expanded sides of 



