128 
NATURE 
| Fune 11, 1885 
sea-water the observations were continued with it. A 
few comparative observations were made in order to 
determine the effect of replacing the sea-water by fresh 
water. On February 2, after a shower, the tempeiature 
of the air was 25°o0 C. When wet with sea-water the 
temperature of the thermometer was 2375 C., and with 
rain-water 23°1 C. Similarly, at noon on the same day, 
the following temperatures were observed: dry bulb, 
26™1 C.; wet bulb (sea), 24°5 C. ; wet bulb (rain), 24°°2 C. 
The air, at this time, appeared, to the sensation, to be 
damper than at any other time, and yet, when suitably 
exposed, there was a difference of nearly 2° C. between 
the wet and dry thermometers. 
There is an advantage in having the bulb of the ther- 
mometer wet with a continuous film of water, instead of 
being surrounded with damp muslin, namely, that it more 
nearly resembles the surface of the sea, which is exposed 
to the influence of the atmosphere. Observations with 
the wet thermometer were not made as regularly as those 
with the dry instrument, and no observations were made 
with either of them after dark, owing to the difficulty of 
securing proper exposure and reading the instrument with 
a lantern, without heating it. 
The temperature of the air and of the water were taken 
generally every two hours from 6 a.m. to 6 p.m., but the 
intervals between the observations were not always the 
same. These observations showed that only on two days, 
January 31 and February 1, between lat. 6° S. and 12’ S., 
did the mean day temperature of the air exceed that of 
the surface-water. On these days the temperatures were 
taken every two hours from 6 a.m. to 6 p.m., and the 
means of the groups of seven observations gave, on 
January 31: air, 27°13 C.; sea, 2690 C.; difference, 
0°23 C. ; and on February 1, air, 27°"26C. ; sea, 26°96 C.; 
difference, 0°30 C. These differences would have been 
reduced in amount if the observations had been carried 
on through the night, though, from the very high tempera- 
ture of the air just before sunrise on February 1 and 2, 
they would not have been reduced to zero. 
In the table (p. 129) all the simultaneous observations of 
temperature of air and water made during the voyage, ex- 
cept those of the last day, when approaching the mouth of 
the River Plate, are collected in small tables for each day. 
The time of day is given in hours, from o to 24; the 
temperatures are in Centigrade degrees ; ¢ denotes the 
temperature of the sea-surface,7— TZ the difference be- 
tween that of the temperature of the air, and 7—7 the 
difference between the readings of the thermometer in 
air with its bulb dry and when it is wet with sea-water. 
At the head of each table is given the meteorological dis- 
trict of the ocean through which the ship was passing, as 
“north-east trade-winds,” “equatorial calms,” and the 
like ; also the day of the month (1885) and the latitude 
and longitude at noonof the day. The means at the foot 
of each table are simply the arithmetical means of the 
numbers in each column; and their meaning and value 
are at once apparent on inspecting the column. 
With the two exceptions above-named, the temperature 
of the sea was always found higher than that of the air, over 
the day, and only very seldom was it exceeded by that of the 
air at the hottest time of the day. Had the observations 
been carried on through the night, the contrast between 
the two temperatures would have been much greater. On 
January 31 and February 1 the conditions were somewhat 
exceptional. On the former of these days the ship passed 
into the northerly monsoon, which prevails all down the 
Brazilian coast during the southern summer. Like the 
similar monsoons in the northern hemisphere, it is caused 
by the proximity of a large mass of land, which gets 
intensely heated by the vertical rays of the sun. On 
January 30 the wind had been light south-easterly ; during 
the night it fell calm, and at sunrise a light easterly wind 
sprang up, which gradually drew around towards the north 
and blew all day, with just sufficient force to travel exactly 
at the same rate as the ship (114 knots); consequently, 
during the whole of the day the atmosphere on the deck 
was motionless, with a very powerful sun beating on it 
and heating up every thing, so that it was impossible to 
find any place where the air could be got, coming fresh 
on board, without having been exposed to the influence 
of the highly-heated deck and fittings. It is therefore 
certain that the air-temperatures are somewhat above the 
water. 
It is probable that, when the true temperature of the 
air can be ascertained, it will be found to be usually below 
that of the sea-surface. The cause of this is, I think, to 
be found in the relative dryness of the atmosphere over 
the ocean. If the observations with the wet-bulb thermo- 
meter be considered, it will be seen that the least differ- 
ence of reading between the dry- and the wet-bulb 
thermometers was 1°o C. on January 28, when the ship 
was in the middle of the equatorial belt of calms and 
rains. In this region perfectly saturated air might be 
expected, and with instruments exposed in the usual form 
of box I have no doubt that here, and in the very 
oppressive weather of the northerly monsoon, the two 
instruments would have given identical readings. The 
readings of the air-temperature on January 28 were per- 
fectly trustworthy, as the sky was thickly overcast with 
dense rain-clouds all day ; there was thus no risk of over- 
heating ; the readings with the wetted bulb were equally 
satisfactory, so that the results of the observations on that 
day may be taken to represent fairly the normal state of 
things in the “ Doldrums.” The temperature of the sea 
varied from 26°°3 to 26°6 C., the mean of five observa- 
tions during the day being 26°-42 C. The mean tempera- 
ture of the air during the day was o°'92 lower than that of 
the sea, or 25°°5 C., and the temperature of the wet-bulb 
thermometer 1°°3 lower still, or 24°°2 C. It will be seen 
that, on the two exceptional days, January 31 and 
February 1, the difference between the wet- and the dry- 
bulb thermometer is greater than would be expected from 
the oppressive damp feeling of the air; it is therefore all 
the more likely that the dry-bulb readings are too high 
as indicated above. However, it is important to observe 
that in all the regions passed through, whether in the 
westerly winds of the North Atlantic or the equatorial 
calms, or the monsoon of the South Atlantic, the tem- 
perature of the wet-bulb thermometer is always very 
markedly below that of the dry-bulb thermometer. In 
fact, such is the mobility of the atmosphere that it rarely 
has the opportunity of saturating itself; and if the effect 
which must be produced when this air meets the surface 
of the water be considered, it will, I think, afford some 
explanation of why at sea the temperature of the air, even 
by day alone, is usually markedly below that of the sea- 
surface. 
If we consider the film of water immediately at the 
surface of the sea, having the atmosphere on the one 
side of it and the bulk of the water on the other, it is 
strictly comparable with the film of water surrounding the 
bulb of the thermometer, when exposed to the atmosphere 
in the way described above ; and the air playing upon it 
must produce exactly the same effect in the one case as 
in the other. The evaporation lowers the temperature of 
the aqueous film, which proceeds to extract heat from the 
neighbouring bodies—namely, in the one case the air and 
the bulb of the thermometer and in the other case the air 
and the layer of water immediately below the surface film. 
lf we imagine for a moment the surface film separated 
from the bulk of the water below it by a diaphragm im- 
pervious to heat, then exposed to the atmosphere so as to 
suffer evaporation and lowering of temperature, then on the 
removal of the diaphragm it would immediately sink 
away from the surface and its place would be taken by 
warmer, and therefore less dense, water from below. In 
the case of sea-water this effect would be slightly intensi- 
fied by the concentration produced by evaporation. But 
a a on 
a 
