ON THE ETHNOLOGICAL SURVEY OF CANADA. 443 



but never barbed. A striking feature of some of these slate arrow-points 

 is that the edges are bevelled from different sides, as if designed to give 

 a rotary motion to the projectile ; but these are not common forms. I 

 have seen but one of this type. A few specimens of obsidian have also 

 been found. In form and variety almost every known type of arrow- 

 head will be found represented here. 



Stone swords of several patterns were also used by these midden 

 makers. Some of them resemble in general outline the short double- 

 edged sword of the Roman legionaries ; others resemble, in cross-section, 

 in a genera] way a ship's belaying- pin. The ends of the handles of all 

 these swords were pierced with a counter-sunk hole, the boring being done 

 from both sides and meeting about the middle. Through these holes were 

 doubtless threaded the leathern thongs which bound the weapons to their 

 owners' wrists. Bone needles of various forms and sizes, with the eye- 

 hole sometimes in the centre, sometimes at one end, are quite com- 

 mon. A few specimens of the pestle-hammer have also been recovered 

 from this midden, though I met with no specimen of the kind in my own 

 investigations. These do not differ in any radical feature from the types 

 found in the later middens of this region. Indeed, I am free to confess 

 that although I hold the more ancient of the middens to have been formed 

 by an antecedent non-Salishan race, the specimens recovered from them 

 do not differ in any remarkable degree from those found in later forma- 

 tions or from the utensils and weapons formerly employed by the present 

 Salish tribes when we first came into contact with them. But while this 

 is true it must be borne in mind that we have some types of utensil, 

 notably the several varieties of the pestle-hammer, which are peculiar to, 

 and probably originated in, this region ; and these may very well have 

 been borrowed by the SalLsh from their dolichocephalic predecessors. If 

 this view is not well founded then it would appear that we have in these 

 ancient middens very clear evidence of the antiquity of the Halkome'lEm 

 tribes in their present habitat. But this I seriously doubt. The evidence 

 gathered from a comparison of their tribal customs, beliefs, and speech, 

 especially of the latter, makes it impossible to believe that they could 

 have occupied their present quarters as separate and distinct tribes since 

 the days of the early middens. For by the most conservative calculations 

 the lower strata of these old middens, such as that I am desci-ibing, could 

 not have been formed less than a thousand yeai's ago ; and from what we 

 know of the rate of dialectic change, phonetic decay, and the evolution of 

 new forms in human speech, and particularly in barbarous and unlettered 

 tongues, less than one half of that pei'iod would have brought about such 

 dialectical differences in the language of the outlying and distant tribes 

 as would long ere this have made them mutually unintelligible. Yet 

 such is not the case. The Halkome'lEm tribes from Yale to the Fraser's 

 mouths and down the Sound all speak what is practically a common 

 dialect. This fact makes it impossible to believe that these tribes have 

 occupied the delta for any very considerable period, and yet everywhere 

 throughout the whole area these ancient middens, many of them acres in 

 extent, abound. 



Adjoining the HalkSme'lEm tribes of the Delta and Sound are other 

 Salish tribes, such astlie Sk-qo'mic, Siciatl, StlatluniH, and N'tlaka'pamuQ, 

 whose dialects in some instances differ among themselves as much as 

 Spanish does from Italian or Portuguese. On Vancouver Island it is the 

 same. Tliere we find a branch of the Halkome'lEm division whose speech 



