NIAGARA FALLS AND VICINITY 79 



hundred feet in bight, with such a pronounced ahernation of hard 

 and soft layers. We must rather assume that a separate fall existed 

 over each hard layer, and that, as in the other streams flowing north- 

 ward over these same strata, these falls were separated from one 

 another by considerable distances. If then, as is clearly indicated 

 by the quartzose sandstone ledge at the inlet to the whirlpool, the 

 lowest of these falls was at that place, the other two or three must 

 have been at some distances up stream, and in thai case it is not too 

 much to assume with Pohlman, that the upper old falls over the 

 Lockport limestone were somewhere near where the gorge is now 

 spanned by the railway bridges. Taylor, however, does not en- 

 counter this difficulty, for he assumes that the St Davids gorge was 

 formed by an interglacial Niagara, the great cataract of which, just 

 before its cessation (probably through a southward diversion of the 

 drainage) plunged as a single fall over the clifY into the basin now 

 holding the whirlpool. To this view it may be objected that the 

 old St Davids gorge is not such as would be formed by a single 

 great cataract, since it flares out northward, having a width at St 

 Davids of perhaps twO' miles. Such a form is more readily ac- 

 counted for if one assumes that the valley was made by the headward 

 gnawing of an ohsequent stream and its various branches. Taylor 

 meets this objection by invoking the action of readvancing ice to 

 broaden the gorge, but, unless the last ice advance was from a very 

 different direction from that indicated by the striae of this region, 

 this hypothesis will scarcely hold. That direction, as already noted, 

 is 30° west of south, while the direction of the old gorge is almost 

 due northwest. Why may we not assume that only a portion, the 

 southern one of the gorge of the whirlpool rapids, was carved by 

 the' Niagara during the time that its volume was diminished, and 

 that the greater portion of this gorge was preglacial? This would 

 greatly reduce the length of time during which the upper lakes dis- 

 charged by way of the Nipissing-Mattawa river, though probably 

 leaving time enough for the waters from these lakes to produce all 

 the erosion features found in this ancient stream channel. This 

 would still leave the Eddy basin to be accounted for, a difficulty 

 which may perhaps be diminished by assuming that the second of 



