Coffey — Irish Copper Halberds. 107 



Cavan, in having two large rivet-holes, and also two notches in the margin 

 at the back, and has likewise a sort of treble mid-rib ; otherwise, it is of 

 the same form as that from Cavan, which is also of bronze, and both agree 

 in being somewhat broader at the base than the length. These two 

 appear to be the only examples of that type of halberd-blade which are 

 known. 



COXCLUSIOXS AND DaTE. 



Of the thirteen copper celts, analyses of which were published in my pre- 

 vious paper, in one case only was the tin returned as reaching 1 per cent. This 

 was the specimen analysed by Mallet, who returned the tin as r09; and 

 it was the only Irish copper celt analysed previous to that time. As 

 Mallet's analysis has been shown to be erroneous in the case of the copper 

 halberd, I am inclined to think that the percentage of tin in this celt 

 may likewise have been stated too high; and it will be best to rule this 

 case out in any discussion of the subject. 



Of the remaining twelve specimens, in eight cases the antimony was not 

 separated from the tin ; and in three of the eight the conjoined tin and 

 antimony reached 0*8 ; in the other five of the eight the conjoined tin and 

 antimony varied from a trace to 0"6. In the remaining four cases out of the 

 twelve, in all of which the tin and antimony were separated, the highest tin 

 reached was 0"12. 



In the five analyses of copper halberds, in all of which the tin and 

 antimony were separately determined, it will be seen that the tin varies 

 from 0"18 to 0"31 per cent. ; and that antimony was present in two cases, 

 amounting to 0*27 in one specimen ; in one of the copper celts, in which 

 the antimony was separately determined, it rose as high as 0"6 per cent. 



We may therefore conclude that the copper halberds are simply 

 coarse or unrefined coppers from similar ores to the copper celts, and that 

 the copper implements found in considerable numbers in Ireland may contain 

 from a trace up to about O'o of tin — rarely, if ever, exceeding that per- 

 centage. 



This small percentage of tin has been shown in my previous paper to 

 be derived from the ore and not intentionally added, and may occur in 

 the copper ores of even a conspicuously non-tin district, as shown by 

 Siret's investigations in the south-east of Spain. It is not necessary to 

 press this point further. 



An increasing percentage of tin was not found in any of the copper 

 celts, or, contrary to expectation, in the copper halberds. Whether a 

 gradual increase of tin would be found in the early bronze celts, showing 



