DRUM LIN, OSAR AND KAME FORMATION. 267 



otherwise occupied the channel previously cut, and that the 

 stream gradually built up its gravels within the tunnel so formed, 

 essentially as indicated by Professor Russell in the case of 

 tunnels under the Malaspina glacier. While the inferences 

 drawn from this peculiar association of the osar ridges with 

 river -like channels cannot be urged with the same force as the 

 preceding considerations, they seem to support. them in some 

 degree. The constitution of these osar ridges is of the same 

 local character as that of the kames above discussed except that 

 perhaps it is less narrowly local and less intimately related to the 

 underlying formations. The difference, however, is not marked. 

 From the foregoing evidences, the inference is drawn that 

 the osars and kames of the plain region of the interior are basal 

 phenomena in a degree almost as complete as the drumlins or 

 the ground moraine. Inferences from such evidences as have 

 been cited cannot, however, be applied with so much rigor in 

 the case of osars and kames as in the case of drumlins, for the 

 subglacial streams, that are held to have formed them, cannot be 

 assumed to have always pursued strictl}^ basal courses. Con- 

 ditions may be supposed to have arisen which would have forced 

 the streams into channels above the base of the ice, or even up 

 over the ice in the thin marginal portion, so that accumulations 

 may have taken place that were less strictly basal than those of 

 the drumlins, and it is of course possible that kames and osars 

 may have been formed, in particular instances, out of the 

 englacial and superglacial material of the ice ; but, following 

 what seems to me the legitimate teachings of the foregoing lines 

 of evidence and of observation, there seems warrant for conclud- 

 ing that such instances, though theoreticall}' possible, are practi- 

 cally rare. I beg that it may be observed that these conclusions 

 are drawn from the phenomena of the plain region of the interior 

 and are applied to it, with the full recognition of the possibility 

 that in hilly and mountainous regions modifications of the con- 

 clusions may be necessary. 



T. C. Chamberlin. 



