76 
tiire to be 50° F,, gives us 29*912 inches as the mean yearly 
value reduced to sea level. 
If we take the complete period of the last eleven years 
we have a mean reading of 29*838 inches, which is 0*074 
inches below the value given for the entire period 1849 to 
1872 . 
In the earlier series the month of August was generally 
wanting in the observations except for a few years, but this 
month having a reading generally in excess of the yearly 
value would tend if anything to make the annual value 
somewhat higher, and the result would tend to show that 
during recent years the pressure has been lower than during 
the earlier part of the series. 
The variation between the highest and lowest mean 
yearly readings during the last eleven years is 29*828 inches 
(1870) and 29*624 inches (1872), or equal to 0*204 inches: 
this would appear greater than probability would point out, 
1872 having had an exceptionally low mean reading ; leav- 
ing 1872 out, we have a difference of 0118 inches between 
1870 and 1866. 
Looking at the mean monthly values, we find that the 
maximum occurs in J une and the minimum in January ; the 
order beginning with the minimum is January, October, 
March, December, November, September, February, July 
and April equal, also May and August equal. If fine 
v/eather depends upon barometric pressure, the above shows 
that May and August should be the finest months as a rule, 
and January and October the worst: May is next to the 
driest month, April, and October bears off the palm for 
being the wettest, at least in this district. 
The remarkable year for rain, 1872, would appear to have 
been also remarkable for deficient atmospheric pressure, as 
the table annexed will show, every month except August 
having had a pressure below the average, and August was 
only a few thousandths of an inch above the average. 
