210 A. W. G. WILSON TRAP SHEETS OF LAKE NIPIGON BASIN 



it happens that all traces of the actual character of the upper portion of 

 these sheets has been lost. In the Lake Mpigon basin alone there can be 

 no question but that more than 80 per cent of the original volume of the 

 mass of diabase that once was present has been removed, assuming that it 

 had even a minimum thickness of between 600 and 7U0 feet (600 feet 

 being the actual thickness of the remnants of the sheet in several places). 

 When it is remembered that at the very summit of the thickest sheets 

 known we find no change in texture indicating that an upper limit is 

 being approached, it seems conclusive that before erosion began the sheets 

 were very much thicker than their thickest remnants are today. While 

 it would be rash to give any figure as a possible maximum based on the 

 data now available, still if a minimum of 1,000 feet is allowed it would 

 be found that more than 90 per cent of the volume of the original sheets 

 has been removed from the area of nearly 6,000 square miles here under 

 discussion. 



Whatever may be the accuracy or inaccuracy of the approximations 

 given above, an enormous amount of erosion undoubtedly has taken place, 

 and in view of the fact that no record has been preserved as to the actual 

 character of the upper part of the upper sheets, even if intrusive, the 

 non-occurrence of any phenomena that can be associated with a volcanic 

 flow must be regarded as a negative argument of very slight weight. Tn 

 this case it is an argument of equal weight for both types of invasion, 

 since each type has a characteristic upper surface under normal condi- 

 tions. In fact, the preservation of upper layers that were more or less 

 glassy and porous — the characteristic feature of most surface flows — • 

 would be extraordinary in a locality where the erosion has been so exten- 

 sive, for such a surface would be even more susceptible to the action of 

 erosive agents than either sediments or the upper surface of an intruded 

 sheet. 



It must also be considered that the features characteristic of the upper 

 portion of a flow are confined to a few feet of the upper parts of the 

 sheet. If there were a number of successive flows, following one another 

 at intervals, so that the upper surfaces of the earlier ones had had the 

 opportunity of cooling before the next flow came upon it, it is very prob- 

 able that remnants of these surfaces would have been preserved. 



A comparable instance is that of the Columbia lava fields, which cover 

 an area of 200,000 square miles. The lava fields are built up by a num- 

 ber of separate flows, and the thickness varies from between 300 and 400 

 feet along the edge of the canyon of the Columbia to 3,700 feet, accord- 

 ing to Le Conte, in the Cascade mountains near Dalles. 4 Eussell de- 



American Journal of Science, series iii, vol. 7, 1874, p. 168. 



