HYDROLOGY OF NEW YORK 



169 



from north to south, while the stream, which is tributary to the 

 Seneca river, the main westerly branch of the Oswego, flows from 

 south to north, or in the right direction to realize the maximum 

 possible leakage, or percolation, through the strata. Inasmuch as 

 no such leakage is mentioned, it may be reasonably concluded that 

 none occurred. 



Cazenovia lake and Erieville reservoirs are also mentioned, and 

 considerable loss of water is given, which when analyzed is found 

 to be loss of water in the canal, and hence not in any degree 

 attributable to leakage through strata. Cazenovia lake and Erie- 

 ville reservoirs both lie south of the Erie canal, and flow across 

 substantially the same strata as the outlet of Skaneateles lake. 1 



Such facts as these, while lacking the proof of a scientific 

 demonstration, are still very interesting and indicate that we- 

 have yet much to learn of the peculiarities of stream flow. On 

 the whole, Avhile they point to a moderate loss from percolation, 

 so far as the writer can see they do not indicate any great prob- 

 ability of very large loss from this cause. They do emphasize the 

 fact that every catchment area will have its own formula. 



By way of showing that the theory of large evaporation on 

 deforested catchment areas is broadly more reasonable that the 

 theory that there is any great loss of water by seepage owing to 

 inclination of the strata, we may consider the Croton record as 

 given by the appended table, where it will be noted that the 

 evaporation from this area is substantially the same as that from 

 Muskingum and Genesee rivers; that is to say. it is the evapora- 

 tion of a deforested area — the area in forest on this catchment 

 does not exceed 10 per cent. In placing it at 10 per cent the 

 writer means the equivalent in actual effect of dense forest. As 

 regards geologic formation this catchment lies almost entirely 

 in granites and gneisses, in which, from their homogeneous char- 

 acter, it is difficult to assume any loss by percolation through 

 strata. There is, however, a small area of metamorphic Hudson 

 formation, consisting of slate, schist and quartzite. and also a 



^nn. Rept State Engineer and Surveyor for 1862, pp. 403-404. 



