JUNE 25, 1914] 
NATURE 
441 
THE GULF STREAM.!} 
ANY theories have been advanced to account for 
4 ocean currents in general and for the Gulf 
Stream in particular, ‘heir causation has been 
attributed by various writers to:—(1) Differences in 
the temperature and density of the sea in widely 
separated geographical positions. (2) Differences in 
level due to inequalities in different regions of 
evaporation and precipitation; and to the outflow of 
great rivers. (3) To convection currents. (4) To the 
rotation of the earth on its axis. (5) To the direct 
action of persistent winds. 
Wind is the prime cause of all currents; persistent 
winds the motive power to which all the great ocean 
streams may be assigned. If anyone be in doubt as 
to the fact, let him place tracings of maps on which 
the direction of the principal currents ot tne globe in 
the different months or seasons of the year are in- 
dicated, over maps on the same scale on which wind 
distribution, referable to the same months or seasons, 
is shown ;and it will be seen how closely the cur- 
rents follow the direction of the wind, and how 
quickly the former respond to changes in the direction 
of the latter. 
In this connection the course of the equatorial cur- 
rent of the Indian Ocean, on the western side of 
the Arabian Sea, may be cited as a striking example. 
During those months when the north-east monsoon 
prevails the current in that region turns to the south- 
ward and joins the Mozambigue current, but as soon 
as the change in the direction of the wind occurs, 
and even before the south-west monsoon is estab- 
lished, the current swings round and flows in the 
new direction of the wind, tc the northward and 
eastward. 
All winds by friction cause some movement of the 
water surfaces over which they blow, while the 
Waves, and even the wavelets they raise, add im- 
pulse to the motion; the stronger the wind the greater 
being its effect at the time. This surface movement 
caused by wind is gradually imparted to the water 
layer below it, and when the wind persists in the 
same direction for long, the motion is transmitted 
from layer to layer to a considerable depth. 
Under the influence of the trade winds, the currents 
when nearing equatorial regions probably extend to 
a depth of from 200 to 4oo ft. 
Although the principal currents are produced and 
maintained by the action of persistent winds, their 
direction is largely controlled by the rotation of the 
earth on its axis and by variation in temperature and 
in density, also in evaporation and precipitation in 
different geographical positions, but these exert only 
slight local modifying effects. Moreover, as regards 
the Gulf Stream and its causation, it was found by 
the officers of the United States Coast Survey that 
the Atlantic Ocean at Sandy Hook was 3 to 4 ft. 
lower than the waters of the Guif of Mexico at the 
mouth of the Mississippi. This difference of level, 
which is said to have been ascertained by accurate 
measurements, doubtless is caused by the heaping 
up of water in the gulf by the equatorial current; 
and the power requisite for maintaining the constant 
flow of the Gulf Stream through the Strait of Florida 
must in a large measure be attributed to this agency. 
The warm, relatively high salinity water which 
undoubtedly exercises an ameliorating effect upon 
the climate of our islands and upon that of north- 
western Europe generally is mainly of equatorial 
origin, and is directly attributable to the agency of 
the Gulf Stream. 
In support of this belief, let me refer you, in the 
1 From a lecture delivered before the Royal Geographical Society on 
May 21, 1914, by Commander M. W. Campbell Hepworth, C.B. 
NO. 2330, VOL. 93] 
| 
first place, to a chart of surtace temperature of the 
North Atlantic in order to show what evidence the 
distribution of mean annual surface temperature will 
reveal. 
The effect of the collision between the Gulf Stream 
and the cold Labrador current is boldly marked by 
the steep temperature gradient from 40° to 46° N. 
Now trace the course of the isotherms onward. The 
isotherm of 50°, which on the s5oth meridian is in 
43° N. lat., on the 28th meridian is in 60° N.; the 
isotherm of 52°, which on the s5oth meridian is 
situated only a few miles south of the 50° isotherm, 
on the roth meridian is in 57° N.; but the isotherm 
of 60°, which on the 5oth meridian is in about 
413° N., reaches the coast of Portugal, after making 
a curve northward, in about the same latitude. In 
other words, the surface temperature of the Atlantic 
between the 43rd and 6oth parallels, and the 4th 
and 32nd meridians is the same as that which is 
found on the 5oth meridian between the qgist and 
43rd parallels, where the Gulf Stream and Labrador 
current meet. 
Now let us see what corroborative evidence a chart 
of average salinity will afford. 
The northern portion of the North Atlantic, the 
southern portion of the Greenland sea, and the part of 
the Barents Sea which are enclosed by the 35 and 36. 
isohalines, are filled with water of the same salinity 
as that which we find in the Gulf Stream between 
Cape Hatteras and its place of meeting withthe 
current from the north. 
Whether the relatively warm saline stream or any 
part of its waters which flows north-eastward from. 
ithe region south of the Great Bank is derived from 
that stream which issues for the most part from the 
Gulf of Mexico, or, as some aver, is an independent 
stream which takes its origin in the former locality, 
is a question which must remain unsettled until the 
results of further investigations are available. This, 
at the least, we know, that from the Strait of Florida 
northward and north-north-eastward to the edge of. 
the Bank; thence north-eastward, as well as east- 
ward, across the ocean, aided, no doubt, by the pre- 
vailing westerly and south-westerly winds, there 
exists throughout the year a continuous flow of warm 
saline or relatively warm saline water to the north- 
easterly branch of which these islands owe much of 
their salubrity. 
The salubrity of our climate is, of course, largely 
due to its comparatively mild and even temperature. 
The relatively small annual range of temperature that 
obtains normally results from our insular position; 
the warmth we owe also in a large measure to the 
surrounding sea, which receives much of its heat 
from that ocean stream, the course of which we have 
been following. 
I will endeavour to show you by means of diagrams 
the somewhat frequent correlation of sea temperature 
with the air temperature over our islands during the 
decade 1903-1912. In order to confine within manage- 
able limits that portion of the inquiry which relates to 
sea-surface temperature, the North Atlantic is repre- 
sented by a broad zone situated between Florida Strait 
and Valencia in the south-west of Ireland. It hap- 
pens that the changes in surface temperature, which 
may be regarded as of premier importance in this 
connection, occur in this zone. 
In the diagrams relating to sea temperature in this 
Florida-Valencia zone, the excess or defect in the 
surface temperature is expressed by the number of 
degrees of longitude in which the 70°, 60°, and 
55° isotherms are east or west of their average limit 
for the month. 
Air temperature over our islands is represented, 
roughly it must be admitted, by the temperature 
