MICHIGAN ACADEMY OF SCIENCE. 
89 
The elevation of the gages above the ground is not given, but that 
on the City Hall is doubtless on the roof, perhaps 100 feet from the 
street. That at the Park is doubtless near the ground. The figures, 
both average and individual, agree well with this supposition, the roof 
station having the usual defect of rainfall. To judge from the figures, 
Dublin and London are also surface stations, but 1 it is observed that 
while Dublin, Park and London show close agreement in the ten-year 
average, in the single years they differ widely and in different senses 
in different years. We need not be surprised then to find an inch or 
two of difference even in the means of observations at stations a hun- 
Fig. 7.—Rainfall map by C. A. Schneider, 1900. 
dred or more miles apart and having very different exposure to the 
winds, I believe interesting results will come of a better examination 
of the winds prevailing during actual rainfall at various stations than 
has yet been made. The increment of rainfall on northwesterly slopes 
can hardly be due to other than westerly or northwesterly winds. I 
have some evidence of this at Saugeen in an unusual proportion of 
heavy rains with northwest winds. In any year when the cyclonic paths 
run south, northwest winds should be frequent in this region. All the 
slopes to lee of Lakes are likely to get larger rain under these circum¬ 
stances, but the slopes to windward of the Lakes as at Detroit should 
suffer a distinct defect. Or perhaps there will be a defect at all stations 
of the region in such years mitigated on leeward shores by some com¬ 
pensating increase in precipitation due to northwest winds. A compari- 
12 
