112 B. A. Gould—Algebraie Expression of the 
Temperature. | Temperature. 
Hour, He Poa = per | Cees Hour. 55) Pee aro —C. 
Observed. | Computed. Observed. | Computed. 
jh 13°°36 13°-41 | —0°-05!| 134 15°°85 18°79 +0°°06 
2 13°28 13:3 —0°08 14 16°02 16°02 0 
5. 13°18 13°25 —0°07 15 15°97 16°01 —0°04 
4 13°15 13°12 + 0°03 16 15°90 15°80 +0°10 
5 13°10 13°00 +0°10 17 15°69 15°42 +0°27 
6 13°06 12°96 +010 || 18 15°13 14°94 +0°19 
7 13°05 13°05 0:00 19 14°48 14°47 +0°01 
8 13°26 £330 —0°04 20 13°95 14°04 —0°0 
9 13°68 13°72 —0°04 21 13°24 13°74 0°90 
10 14°26 14°24 +0°02 22 13°61 13°56 + 0°05 
ll 14°83 14°83 0°00 23 13°52 13°47 +0°05 
12 16°35 15°38 —0°03 24 13°45 +0°01 
So far as it can be deduced from the hourly observations, the 
true formula to the 2¢ variable term inclusive (i. e. with four 
constants) is : 
T= 14-20 +1°415 sin (A+ 223° 48’) + 0°500 sin (2h + 20° 50’) 
and the true daily maximum is at 14" 38™. 
9. Let us next consider the observations made at Hobart 
Town in Tasmania during the month of January in eight suc- 
cessive years, also published by Dove and expressed in degrees 
of Réaumur. Here T,=11°'90, Ty=17-29, Ty=12°07, the 
mean of these three temperatures being 18°75. 
Beginning with this for the value of M we find the second-— 
ary maximum to be very pronounced; and assuming succes- 
sive values, inferior to this, we find that the highest which 
gives, for H,=14", a line without contrary flexure is 13°°45, 
and tnat none above 13°-40 fails to show decided indications of 
a secondary maximum between 0" and 4", For M=13°40 we 
have a=322, A=238°'46, b=0°68, B=82°-56, which repre- 
sent a curve with rounded and flowing outlines. Yet even 
this not only exhibits a very marked want of symmetry be- 
M is still smaller. 
80’, 
I 
b=059, B=39° 12’ these representing a curve which manifests 
the same peculiarity of a too early minimum. But by vary- 
