9 
ity examining the numbers in the forty-sixth column, it will be seen 
It there are many large differences between the values on consecutive 
If 
j!y laying these down on a diagram all these are shown, and by 
ndng a curved line to pass through or near all these points, giving 
iclal weight to every one, the most probable temperature of every day 
n he year is shown. The numbers in Table XIII. are the best I can 
[dide upon as the nearest approximation to the true temperature 
tonging to every day in the year. 
■he days of the lowest mean temperature (36°*3) are January 7th, 
1, and 9th; it then increases to 39°'3 on the 4th of February, remains 
t ionary at that temperature till the 8th, and then gradually decreases 
c39 o, 0 on the 13th and three following days; from the 16th it 
reases , and continues so to do, gradually for the first few days, but 
awards the end of March at a more rapid rate, the mean temperatures 
f onsecutive days differing as much as two or three-tenths of a degree 
i a each other, and continues increasing at this rate till the 8th of 
i e, when it slackens to a general rate of one-tenth, but still increases 
[ we arrive at July 9th, the mean temperature of this day differing 
la that of the 26th of the same month by one-tenth of a degree only; 
; len begins to decrease , gradually till the 11th of August, and at a 
re rapid rate from that date, decreasing uninterruptedly till the 
lb of November, from which day till the 9th of December, when the 
tlage is very small, the mean temperature being unchanged for three 
i four days together, it then continues to decrease until the end of 
] year. 
1 
ft 
{ 
•j 
<8 
« 
'O 
fl. 
ft 
9 ) 
