THE DATE OF THE GLACIAL PERIOD. 583 



solid: those in the west more oxidized and rotten, but the 

 critical condition in the west as in eastern Pennsylvania 

 is the constant presence and mixture of fresh rolled material. 

 As the age of the mixture is the age of the freshest part, 

 there is finally no difference between the ages of the eastern 

 and western Kansan drift. , 



"The rustiness of the western gravels shows that they were 

 the rolled and weathered surface fragments picked up by 

 the ice, and modified by its action. With the crystallines 

 which ara thoroughly oxidized to their center, we find a 

 few specimens which look on one side like a piece of rusty 

 gravel. The black bisilicates have entirely disappeared, 

 and have left pits and a rusty staining. The feldspar also 

 has kaolinized. In these very old ones, the glaciated sides 

 were never scraped down to the fresh interior, but uniformly 

 show a rusty though solid exterior, quite smooth and firm 

 to the hammer. 



"By breaking these, we would find that the solid nucleus 

 might be one-eighth of an inch from one side which had been 

 glaciated, and three inches or more from the other side which 

 had remained unglaciated." 



I am not aware that adequate attention has been paid to 

 this class of facts over the Illinoisan and Kansan areas in the 

 Mississippi Valley, but I was much impressed with the fresh- 

 ness of the Canadian bowlders which were found at Tus- 

 cumbia, on the Osage River, in central Missouri, and by the 

 freshness of many of the pebbles in a great gravel-pit at Hol- 

 liday on the Kansas River a few miles above Kansas City to 

 which the railroads have resorted for a long time for ballast 

 and which contains much material from the far north. 



In this connection it is proper to call special attention to 

 the accumulating evidence going to show that the glacial 

 movement from the Keewatin center was not strictly con- 

 temporaneous with that from the Labradorian center but 

 preceded it by a longer or shorter interval. 



In the first edition it was suggested that the remarkable 

 re-entrant angle in the glacial border at Salamanca, N. Y., 

 indicated the junction of two ice-movements from widely 



