BY THE BOILING POINT OF WATER. 241 
Upon carefully projecting Dr Hooxer’s results in Plate ITL., fig. 2, in the same 
manner as my own, that is, by exhibiting the logarithms of the pressure (or the 
heights) in terms of the temperatures of boiling water, I found, in the first place, 
large deviations among the results, increasing also at great elevations. The 
breadth of the space is so considerable over which the individual observations 
are distributed, that it seems impossible, from the observations only, to assign 
any one curve as particularly indicated by them; and for the most part they are 
as well represented by a straight line as by any curve not absolutely sinuous. I 
must, however, note that below the temperature of 187°, or at heights above 13,600 
feet, something like a dislocation occurs in the continuity of the observations. Dr 
Hooxer was aware of this circumstance, and ascribes it to “the metal of the 
kettle, and consequently of the thermometer, getting heated above the tempera- 
ture of the boiling water.” Whatever may have been the cause, this part of the 
series, the most important for testing a formula, can hardly be relied on for that 
purpose. — 
There is no doubt that M. Reanavtt’s numbers represent, as well as any num- 
bers can be expected to do, the main features of Dr Hooxer’s observations.* But 
the differences between M. ReaNav.t’s numbers and my approximate formula are 
trifling, compared to the latitude of error which the projection of the observations 
themselves discloses. Balancing the errors as nearly as possible, the observations 
between 212° and 190° are well represented by a line which gives 538 feet of ascent 
for a fall of one degree in the boiling point, which it will be seen differs only by 
qth part from the corrected result of my Alpine observations.} I beg to observe, 
that this coincidence is the more striking, because, from the method of projection 
used, it was impossible to guess at the numerical result until the interpolating 
line had been fixed upon. 
In criticising Dr Hooxer’s results, I do so with every feeling of courtesy and 
respect, in the same spirit, in short, in which I am sure he found it necessary to 
state his objection to my formula. The whole of his barometrical observations 
appear to have been made with the greatest care and fidelity, and, judging by the 
results, with great success. From not knowing his thermometric apparatus, I am 
_ unable to determine why these observations are of less value. I should attribute 
it rather to the boiler, or to the mode of using it, than to the thermometers; for 
Dr Hooker speaks of a coincidence in the readings of different thermometers so 
exact as to be unusual. Dr Hooxer states that he finds the errors by actual cal- 
culation considerably less, if reduced by M. Recnauut’s numbers than by mine; 
* A few points marked by the letter 7, calculated from Reenavxt’s formula, are inserted in the 
figure for the sake of comparison. 
} The entire series of Dr Hooxer’s observations is best represented by 548 feet for 1°, when 
we include the (somewhat doubtful) highest observations. This agrees almost exactly with my 
earlier determination. 
