February 16, 1894. 
SCIENCE: 
PUBLISHED BY N. D. C. HODGES, 874 BROADWAY, NEW YORK. 
SUBSCRIPTIONS TO ANY PART-OF THE WORLD, $3.50 A YEAR. 
To any Contributor, on request in advance, one hundred copies of the issue con- 
taining his article will be sent without charge. More copies will be supplied at about 
cost, also if ordered in advance. Reprints are not supplied, as for obvious reasons 
we desire to circulate as many copies of SciENCE as possible. Authors are, however, 
at perfect liberty to have their articles reprinted elsewhere. For illustrations, 
drawings in black and white suitable for photo-engraving should be supplied by the 
Rejected manuscripts will be returned to the authors only when the 
Whatever is intended for 
contributor. 
requisite amount of postage accompanies the manuscript. 
insertion must be authenticated by the name and address of the writer; not necessa- 
rily for publication, but as a guaranty of good faith. We do not hold ourselves 
responsible for any view or opinions expressed in the communications of our corres- 
pondents. 
Attention is called to the ‘‘ Wants”? column. It is invaluable to those who use it 
in soliciting information or seeking new positions. The name and address of 
applicants should be given in full, so that answers may go directly to them. The 
“ Exchange ’* column is likewise open. 
PRESSURE OF THE VAPOR OF WATER. 
BY H. A. HAZEN. 
One of the more important facts needed in meteorology 
is the pressure of the vapor of water, commonly called 
‘vapor pressure”’ for short. Until very recently Reg- 
nault had obtained the best values for this element. His 
method was to introduce a capsule of distilled water, 
from which all air had been expelled, into the vacuum of 
a barometer and there liberate the water. It is well 
known that this vapor will diffuse itself and absolutely 
saturate the space above the mercury. Its pressure will 
be exactly dependent upon the temperature and can be 
ascertained approximately by comparing the barometer 
reading with a perfect barometer. It is easy to see that 
many elements of inaccuracy are introduced in such an 
apparatus. Most of these have been eliminated by a most 
beautiful apparatus, designed and constructed by Prof. 
C. F. Marvin, of the Weather Bureau. I have made 
thousands of readings with it, and it is one of the most 
satisfactory instruments to manipulate I have ever seen. 
In this, the two barometers are dispensed with, but there 
are two vertical tubes connected at the bottom and partly 
filled with mercury. A bulb is attached to one of the 
tubes, and afterward the air is exhausted and vapor 
liberated by breaking-a capsule of water previously in- 
serted in the bulb. ‘The heights of the mercury columns 
are read by means of a vernier. I found no difficulty i in 
repeating again and again, and day after day, readings 
within .o3 to .o4 of a millimetre (.oor2 to .oor6 in.). 
‘A full description of a perfected apparatus will be found 
in the Annual Report of the Chief Signal Officer for 1891, 
Pp- 351-383. 
In January, 1890, I carried the apparatus with great 
pains to Northfield, Minn., and there made a series of 
readings, the results of which were published in the 
Annual Report of the Chief Signal Officer, 1890, pp. 658— 
662. Perhaps the most interesting result obtained was a 
marked difference in vapor pressure when the temperature 
of the water was different from that of the vapor. ‘This 
is best shown in a series of comparisons of water at 
freezing and a portion of the vapor above it at different 
temperatures, as follows 
Water | Vapor Temperature Fahr. 32 42 52 G2g7 2 2 LOZ et 
32°. | Vapor pressure 4.00 4.57 4-55 4:53 4-50 4-48 4.44 4.39 4.33 
Most of the vapor in these experiments was at the 
temperature of the liquid. 
. 
SCLIN: 87 
It should be noted, in passing, that Professor Marvin 
did not obtain any effect of this kind, and in his results it 
is ignored. 
There has just come to hand a very interesting paper, 
by Prof. Geo. W. A. Kahlbaum, of Basel, Switzerland, 
in ‘Archives des Sciences Physiques et Naturelles,” 
Geneva, vol. 31, p. 49. In this paper the author 
shows very clearly that there isa marked effect depending 
upon the difference in temperature between the liquid 
and the vapor above it. The only portion of this in- 
vestigation needed for our purpose is that relating to the 
vapor of water. To present the facts in the best possible 
shape for comparison I have placed in the following 
table values of vapor pressure at different temperatures 
by various experimenters. In the first column is the 
temperature of the vapor in degrees Fahr., and in the 
succeeding columns the pressure in millimetres of mercury, 
as observed by Regnault, computed by Broch from 
Regnault’s observations, observed by Kahlbaum, by 
Professor Marvin, by myself, and as determined by the 
Royal Society of England, probably from the results of 
various observers, but. this is a mere inference. 
VAPOR PRESSURE IN MILLIMETRES AT VARIOUS 
TEMPERATURES. 
(x) (2) (3) (4) (5) (6) (7) 
Fahr. Reg. Broch. Kahl. Marvin. Hazen. Roy. Soc. 
° I.Or I.I4 97 1.1L 1.30 
5 1.32 1.44 1.2 1.38 
Io 1.72 1.81 1.60 1.72 1.98 
15 2-18 2.25 2.06 2.18 ‘ 
20 2.78 2.79 2.61 2.75 2.05 
2 3-45 3:4. 3:35 3:44 
30 4.25 122 4-17, 4:25 4.37 
35 5-17 15 [5.24] 5-17 
40 6.20 26 6.28 6.38 
5 7-60 58 7-62 
5° 9.16 14 9.36 9.15 9-17 
55 11.09 10.97 10.97 10.91 
60 13.15 13.22 13-13 13.05 13.11 
65 15.68 15.05 15.67 15-50) 
Jo 18.62 18.59 18.47 18.53 18.36 
75 22.04 22.01 21.76 21.06 
80 25.99 25.90 25.07 25.88 25.53 
85 30.50 30.52 30.51 30.30 
90 35-81 35-75 36.12 35-41 35.05 
A comparison of (2) and (3) shows a tendency to error 
at very low temperatures in the mathematically computed 
results. I do not see how we can go back of the original 
record in such case. ‘The values in (4) agree fairly well 
with those in (2), except at Ome 7°, 75 and 80. It is 
possible there is a misprint, otherwise there would seem 
to be some error in the values. It would be very gratify- 
ing if we had the results at the lower temperatures. 
Column (5) gives results which are lower than those in 
(2), and as the observations were all at temperatures of 70° 
for some of the vapor, it would seem, according to the law 
determined by two independent observers, that these 
values are all too small. Perhaps Regnault’s work was 
done at temperatures of water, or ice, and vapor approxi- 
mately the same, and, if so, his results may be very nearly 
correct, so far as this point is concerned. In column (6) 
all the values are reduced to a common temperature of 
water and vapor. ‘These agree remarkably with column 
(2) till we reach 85°, when there is a falling off. It was 
found rather difficult to manipulate the apparatus at this 
high temperature, and it is also probable that whatever 
errors existed in the apparatus were largely increased at 
these higher temperatures, so that I do not insist upon 
the absolute accuracy of the results in (6) above 85°. It 
is a little singular that there should be this rather rapid 
fall in my values as compared with Regnault’s, and I am 
confident they are not due wholly to errors in (6). There 
is another point of great interest in this connection. 
Column (4) supposed to have this effect entirely 
eliminated, but that has a very sharp fall as compared 
with (2) at 7o° —. .28™m, 80° —.32™™ then 
arise at 90° -+.31™™. It seems very difficult to account for 
these jumps in (4), and it may be that there is some error 
15 mm) 5° — 
