. Donovan — On a Comparable Hygrometer. 487 
At a subsequent period, I made 77 inspections, with the following 
results, during a week, at unequal intervals, the silk strings being 
equal : — 
In 31 inspections the hj'grometers agreed precisely. 
In 13 they differed one degree. 
In 18 they differed two degrees. 
In 5 they differed four degrees. 
In 6 they differed five degrees. 
In 3 they differed six degrees. 
In 1 they differed ten degrees. 
The instances in which the two hygrometers agreed, or nearly 
agreed, are sufficiently numerous to induce a belief that those in which 
they disagreed were occasioned by the same causes as Regnault found 
to affect his psychrometer momentarily, which he says " must be 
attributed to the successive arrivals of air which contain quantities of 
moisture often very different." Whether this be so or not, a difference 
of three or four degrees in a thousand of the range must be of small 
effect in any inquiry, and is not a greater imperfection than has been 
found in other hygrometers of acknowledged efficiency. Thus, it is 
stated that two of De Luc's whalebone hygrometers, being compared, 
were found to differ five degrees, although the scale of each had been 
recently adjusted ; and that the best hair hygrometers of Saussure, 
when compared, often deviated several degrees in the same medium. 
(Edinb. Encyclop., Art. Hygrom., p. 392). 
From the facts stated, I think it may be admitted that this hygro- 
meter is a comparable instrument, when carefully made and managed. 
There is one other office which a hygrometer has to perform. It is not 
sufficient that it shall inform us of the presence of water in the atmo- 
sphere; we require to know how much — the ponderable quantity in 
grains' weight which a vacuum of a definite measure, such as a cubic 
foot, may at any time contain, without the knowledge of which the 
words " dampness," and ''dryness," give no precise information. 
The hygrometer here described represents the natural unsaturated 
state of the atmosphere: ten rounds or 1000 degrees constitute its 
scale ; all less numbers of degrees constitute its fraction of saturation. 
This fraction is the measurer of the moisture belonging to all existing 
temperatures of the time as they present themselves; but it does not 
give information as to the actual ponderable quantity of water present 
in a certain volume of air or space, such as a cubic foot, may at any 
time contain. 
To arrive at the knowledge of this quantity, we must, in the first 
instance, seek the aid of experiment and computation. The first step 
will be to find the dew-point : the easiest method for which is the cold 
water process of Le Eoy, recommended by Dalton, Berzelius, and 
others. Suppose the dew-point to be GO"*, find the elasticit)^ of that 
temperature from the tables of Regnault, interpreted and adapted to 
