EVAPORATION 
INSTRUMENTS AND METHODS 
Evaporimeter 
The rate of evaporation from a pan of sea water was 
calculated at intervals on board the Carnegie from meas- 
urements of the changes in salinity of a sample of sea 
water. The evaporimeter, a thermometer-right-angle 
type (Richter and Wiese No. 16), consisted of a copper 
vessel (fig. 37) within which was fitted a glass container 
capable of holding 2000 cc of sea water and exposing a 
surface of 263 cm2. The whole apparatus was set in 
gimbals to offset the effects of the natural rolling of the 
ship. A set of gimbal stands was mounted on each side 
of the stern near the rail, and by exchanging the evapori- 
meter and rain gage it was possible to keep the evapo- 
rimeter always on the windward side of the vessel. 
Supplementary Instruments 
and Observations 
The Assmann psychrometer was used for obtaining 
wet- and dry-bulb temperatures at the evaporimeter, 
and a standard Tycos anemometer (DTM No. 4), record- 
ing wind speed in feet per minute, was used to determine 
total wind movement. The thermometers used in the 
psychrometer were standard instruments, as previously 
described. 
Throughout the duration of each series of measure- 
ments the following observations were made at each hour: 
(1) ship’s course and speed; (2) wind speed and direction 
at the evaporimeter and at the rail; (3) wet- anddry-bulb 
temperatures at the evaporimeter and at the rail; (4) 
sea-surface temperature; (5) temperature of the water in 
the evaporimeter; (6) state of the sea; (7) state of the 
weather; (8) pitch and roll of the vessel; (9) amount and 
type of clouds; (10) atmospheric pressure; and (11) the 
amount of precipitation during the hour. Precipitation 
was measured in inches by a standard rain gage. 
Methods of Determining Salinities 
The salinity of the sea-water samples in the evapo- 
rimeter was determined by means of the salinity bridge 
used in oceanographic work, except when the salinity was 
greater than could be recorded on the scale of the bridge. 
In such cases, the titration method was used. 
The depth (h) of evaporation ir. millimeters was de- 
termined by the following formula from Wist [37] 
h = Cs6/(S2 - S1) / S2 
where 6 is the specific volume of distilled water at the 
mean temperature of the water in the evaporimeter; C 
is the constant of the vessel, and is equal to the quotient 
of the volume and the evaporating surface (for the Car- 
negie data this is 76.05 mm); S1 and Sg represent the sa- 
linities in parts per mille at the beginning and end of the 
run; and s is the density of sea water at Sj and ty (begin- 
ning temperature). The values of s were obtained from 
Knudsen’s Hydrographical Tables®. 
Actually, for salinities between 30 parts per mille 
and 40 parts per mille, and for temperatures between 

4 English edition, Copenhagen (1901) 
52 
-2°0 and 30°0, the formula can be simplified by using 
the value 1.027 for the product of s and 6. The error 
resulting from the use of this mean value does not 
amount to more than 2 per cent of the actual value. 
The amount of precipitation during the run has been 
added to the resulting depth of evaporation as deter- 
mined from these measurements. 
Evaluation of Data 
Although the observer, Dr. J. H. Paul, chose calm, 
fair weather in which to carry out the measurements, it 
was often necessary to discontinue runs when rolling or 
pitching of the vessel, or vibrations caused by running 
the main engine, brought about the possibility of water's 
being splashed from the evaporimeter. Even though ev- 
ery precaution was taken to insure the accuracy of the 
results, there remains the possibility that, unknown to 
the observer, water was splashed out of the container or 
added to that already present by salt spray, dew or spray 
from the water used in washing down the decks. These 
seem to be the chief sources of observational errors. 
It is possible, however, to calculate the probable 
inaccuracy of the results due to the accidental addition 
or subtraction of sea water from the evaporimeter dur- 
ing the run. If it is assumed that the change in height of 
the water in the vessel owing to the above causes amounts 
to +5.0 mm (a generous allowance), then by taking the 
mean values of Sj and S2 as 36 and 40 parts per mille 
respectively, it is found that 
h = (263 x 0.5 cc/263 cm2) 1.027 
(0.040 - 0.036) /0.040 = +0.5 mm 
This is 6 per cent of the mean value of h (7.8 mm) fora 
vessel containing exactly 2000 cc of water at the above 
mean salinities. 
Because of these uncertainties, as well as the pres- 
sure of carrying out other programs of the Carnegie’s 
work, it became necessary to discontinue the evaporation 
observations after January 9, 1929. 
DISCUSSION 
One of the major problems of marine meteorology is 
that concerning the quantity of water evaporating from 
the surface of the sea. Consequently, preparations were 
made before the Carnegie left Washington for determin- 
ing evaporation rates of sea water from a pan on board 
the vessel. It was not until July 19, 1928, however, that 
conditions were obtained which were favorable for be- 
ginning these measurements. Between this date and 
January 9, 1929, a total of twenty-three successful evap- 
oration series were made. Nine of these were made in 
the North Atlantic Ocean and the remaining number in 
the southeastern Pacific Ocean. Most of the series were 
carried through 24 hours; five through 48 hours. 
The results of the*twenty-three series are presented 
in table 63. The uncorrected evaporation values range 
from 2 mm to 10 mm with a mean of 6.22 mm. Accord- 
ing to Wiist [38], these values must be reduced by multi- 
plication with the factor 0.53 to represent fairly actual 
evaporation from the surface of the sea. The mean of 
the twenty-three Carnegie series corrected in this man- 
