SHELL-MOUND BUILDERS NOT MERE SUMMER VISITORS. 185 



the ancient fishermen resided on these spots for at least 

 two- thirds, , if not the whole of the year. . This we learn 

 from an examination of, the bones of the wild animals, as it 

 is often possible to determine, within very narrow limits, the 

 time of year at which they were killed. Por instance, the 

 remains of the wild swan (Cygnus musicus) are very common, 

 and this- bird is only a winter, visitor, leaving the Danish 

 coasts in March, and returning in November. It might 

 naturally have been hoped that the remains of young birds 

 would have supplied evidence as to the spring and early 

 summer, but unfortunately, as has been already explained,, 

 no such bones are to be found. It is, therefore, fortunate 

 that among the mammalia two periodical phenomena occur ; 

 namely,, the shedding and reproduction of stags' antlers, 

 which,, wiffi slight variations according to age, have a fixed 

 season; and, secondly, the birth and growth of the young.. 

 These, , and similar phenomena render it highly probable 

 that the " mound-builders" resided on the Danish, coast all 

 the year round, though I am disposed to think that, like 

 the Fuegians, who lead, even now, a very similar life, they 

 frequently moved from spot to spot. This appears to me 

 to be indicated not only by the condition of the deserted 

 hearths, but by the color of the flint flakes,, etc. ; for while 

 many of these retains the usual dull bluish black color which 

 is characteristic of newly-broken flints, and which remains 

 unaltered as long as they are surrounded by carbonate of 

 lime, .others are whitened,, as is usual with those which have 

 been exposed for any length of, time. Perhaps, therefore, 

 these were lying on the surface during some period of deser- 

 tion, ,nd covered over<only when the place was again inhabited. 

 The flint implements found in the Kjokkemnoddings re- 

 semble those which are characteristic of the " Coastfinds." 

 They may be classed as flakes (figs.. 62- 69); "heU-mound "* 

 axes, which, as we have already observed, present a peculiar 



