The Second Lake AUjonqiiin. — TaijJor. 1 17 
actly oil a straight line. A straight line was therefore drawn 
through Mackinac and Houghton and this line was taken as 
the first or fundamental isobase. On a comparison of the re- 
maining points of observation off CC it was found that they 
could not be better represented than by straight lines parallel 
to CC. In this way BB and DD were constructed. The main 
part of the Nipissing plane so determined lies mostly between 
the isobases BB and DD, and it extends from Mackinac to 
Houghton. This area is 250 miles long and about 100 miles 
wide, if we count the distance to Gladstone and Fayette which 
lie south of BB. From the points observed in this principal 
area the place of the node line AA was calculated and its 
place so determined, is approximatel}'^ parallel with the other 
lines. From this principal area, which comprises carefidl}^ 
measured parts of the Nipissing beach in each of the three 
upper lake basins, the Nipissing plane was produced in all di- 
rections and its relation to the various littoral features of the 
remaining parts of the lake basins were noted. Some of these 
will be described in detail later on. The mean rate of rise in 
the Nipissing plane from Petoskej^ to Sault Ste. Marie is little 
more than 6^ inches per mile. It will thus be seen that a differ- 
ence of five feet in altitude is equivalent to nearly ten miles 
difference in the place of an isobasal line. Almost anj'^ of the 
measurements may be in error as much as five feet either way. 
It follows that the isobases may be ten miles in error either 
way at any place. But the extension of the measured plane 
over so wide a space, and especially its very close agreement 
with facts which point to its extension over several times that 
space in other parts, reduces the probable error very much 
and increases the value of the isobases. Facts in widel}^ sep- 
arated places prove that these lines are certainly not far out 
of place. Points of observation not situated on a line gener- 
ally show an altitudie which agrees with an extension of the 
plane between or beyond the lines. There are a few i)oints 
that appear to be exceptions. All l)ut one, however, are within 
the limit of error. Marquette and L'Anse seem a little too 
high and Midland a little low. North Bay alone is wide of 
the mark, being 40 feet higher than the main plane pro- 
duced to that point. The rate of rise from Sault Ste. Marie 
to North Bay (transferred to M on tiie line KK ) is lu-arly 
