SIXTH REPORT OF THE DIRECTOR I909 I95 



to act on its walls, dent-pit action commences, and all other local 

 agencies become more effective. As these caves thus work their 

 way back from the coast line the air in them is at times compressed 

 by storm waves which break against and cover their mouths. A 

 large portion of this compressed air thrusts out a volume of water 

 at the upper portion of the cave mouth with great violence, and 

 the tops of these spouting caves are thus cut into the form of an 

 arch. A portion of this air, however, with its contained hundreds 

 or thousands of dust particles per cubic centimeter 1 must be dis- 

 charged backward into the open joint crevice with considerable in- 

 itial speed and thus serve to widen its proximal portions and allow 

 freer entrance of water to these portions at times of high water. 



Joint channels and joint caves were more or less modified during 

 the last glacial period. With the advent of this period the outer 

 portions of the rock mass were gradually lowered in temperature, 

 and a condition of permanent frost came to occupy the joint chan- 

 nels. These were thus filled with ice before surface movement of 

 the ice sheet became an effective erosive agent. In many locali- 

 ties in the glaciated regions of North America there are deep fis- 

 sures or cavernlike openings which were so filled and which still con- 

 tain glacial ice. The ice filling of these joint channels saved them 

 from being filled with mantle rock when the ice movement began and 

 saved them also from a later filling of till, though they received a 

 till covering. The loss of material from the underside of some 

 of these till coverings is now made manifest by a sinking surface 

 over certain joint channels or by small sink holes like that shown 

 in plate 6. 



At some time during the glacial period subglacial drainage must 

 have been comparatively free on surfaces now some 600 feet below 

 the highest level of Lake Vermont (Glacial Lake Champlain) 2 else 

 it is hard to conceive how giant kettles and rill channels (cut by 

 surface flow of glacial waters) could be now found in this lower 

 region. During such a time of elevation (at least to a degree as 

 high as the present) subglacial waters found their way into wide 

 joint channels through narrow surface openings and these were 

 rapidly cut into circular shape by spirally descending waters. Some 

 of the larger caves were thus occupied for a brief period by rap- 



1 See Aitken, J. On the Number of Dust Particles in the Atmosphere. 

 Roy. Soc. Edin. Trans. 36:1 (1888) ; see also Nature, 37:428 (1S88), 40:394 



(1S90). 



2 See Woodworth, J. B. Ancient Water Levels of the Champlain and Hud- 

 son Valleys. N. Y. State Mus. Bui. 84, p. 195. 



