728 LOUIS AGASSIZ, 



in five hours. One such isolated observation 

 is of course of little comparative value. For 

 himself, Agassiz reserved the study of the 

 bay, the ancient bed of the glacier in its for- 

 mer extension. He spent the day in cruising 

 about the bay in the steam-launch, landing at 

 every point he wished to investigate. His 

 first care was to examine minutely the valley 

 walls over which the glacier must once have 

 moved. Every characteristic feature, known 

 in the Alps as the work of the glaciers, was 

 not only easily recognizable here, but as per- 

 fectly preserved as anywhere in Switzerland. 

 The rounded knolls to which De Saussure first 

 gave the name of roches ')noutonn6es were 

 smoothed, polished, scratched, and grooved 

 in the direction of the ice movement, the 

 marks running mostly from south to north, or 

 nearly so. The general trend of the scratches 

 and furrows showed them to have been con- 

 tinuous from one knoll to another. The fur- 

 rows were of various dimensions, sometimes 

 shallow and several inches broad, sometimes 

 narrow with more defined limits, gradually 

 passing into mere lines on a very smoothly- 

 polished surface. Even the curious notches 

 scooped out of the even surfaces, and tech- 

 nically called " coups de gouge," were not 



