194 NEW YORK STATE MUSEUM 



If the temporary base level was determined by a surrounding 

 body of fresh water, as it appears to have been in the case of the 

 Valcour island elevation, the exits of the joint channel system of 

 drainage would be under the influence of a greater number of 

 erosive agencies than the inland portions of the same system. So- 

 lution would be incessant, the solvent would be always in motion, 

 and it never could approach a condition near satu-ration. The pres- 

 ence of ligiht would permit the growth of lowly organized plants 

 and animals, and their metabolic processes while living and 

 their decay at death would both favor the process of chemical so- 

 lution. If we will compare the effects of simple quiescent solution 

 by thin water films approaching saturation with the effects of 

 chemical and mechanical solution by incessantly acting and con- 

 stantly moving waters of great volume, we shall see that this dif- 

 ference in erosive power would of itself turn joint-channel exits 

 into water-level caves. To this means of differentiation we must 

 add the influence of wave motion. This agent does work of very 

 varied eharacter from that accomplished through sliding or roll- 

 ,ing of heavy blocks or the throwing of pebbles taken temporarily 

 into suspension, to the work of broken wave vortexes holding soHd 

 particles so fine as to be almost in a state of permanent suspension. 

 The submerged condition of the cave floors on Valcour island has 

 protected them from the more energetic types of wave erosion, but 

 the effects of vortex work are very manifest. Expansion of freez- 

 ing water in these joint-channel exits is also a factor productive 

 of widening. As a result of these agents of erosion the joint-chan- 

 nel exits become caves which narrow toward their roofs and 

 rapidly narrow to channel dimensions as they enter the rock mass. 



While the foregoing discussion outlines the probable history of 

 formation of the larger shore-line caves of Valcour island, it does 

 not account for the formation of the majority of the smaller open- 

 ings for these do not seem to be connected with joint channels. 



If a narrow joint crevice, so- situated as to receive little water 

 from rainfall in its higher portions, is bathed below by lake waters, 

 such waters will tend to fill the crevice by capillary attraction but 

 will at the same time receive all dissolved material and tend to 

 maintain a fresh-water filling. As the lower portion of the crevice 

 is thus widened the hight to which the water filling is maintained 

 is lessened, but a new factor comes into play, for the effect of 

 wave pressure in filling and emptying the crevice becomes now more 

 pronounced and allows of more rapid changes in the filling. So 

 soon as the breach becomes wide enough to permit water vortexes 



