The Deep Currents of the North Sea. 7 
remaining two as shewn on the chart, near Christiansand. By a 
curious coincidence, these two bottles were found on the same day 
30 miles apart from each other, having covered respective distances 
of 350 and 380 miles in 294 days. 
Seventy-three was also a group of ten, one of them being trawled 
up three days after it was put away, the only others recovered being 
the two plotted on the chart, which were found on the Norwegian coast. 
Bottles 43 and 51 were put away at the same place,.the former in 
December, 1910, the latter in May, 1911, and both have been 
carried to the eastward in opposition to the very decided inset of 
the bottom eurrents towards the Firth of Forth. So, again, it is 
highly probable that these long voyagers may have drifted on the 
surface, and a glance at the daily average velocity attained by each 
tends to confirm this view. 
Reference No. of Bottle, . . | 6 | 62 | 67 | 67 | 73] 73 | 43 | 51 
Average Velocity per day, . . | 1:1| 0-4} 1:2] 1:3] 1:5] 1:2] 05} 1:0 
These velocities are, of course, based on the assumption that the 
bottles have been found on the day they were washed up on the 
shore, but the time they may have lain on the beach before being 
discovered must for ever remain unknown, so the actual speed at 
which these bottles have been carried along will be somewhat 
greater, im a more or less degree depending on the period elapsed 
between the time of their stranding and the discovery of the bottle. 
A record of these long-distance drifts is given in Table V.(s). 
Table IV.(s) contains a record of the bottles found on the Scottish 
coasts, all of which shew that the bottom curves calculated from the 
trawled bottles may be continued right into the coast line. 
SUMMARY. 
A summary of the results obtained from the present experiment 
is given in Table If. under subhead (c), and in order to obtain 
curves from the maximum number of observations available we 
have combined the resultants of subheads (4) and (c) and shewn 
the result in the same Table under subhead (p). These results are 
also shewn diagrammatically on Chart III., and, as might be 
expected from the greater number of observations extending over a 
longer period of time, the local resultants are found to arrange 
themselves with greater uniformity, and the curved resultants appear 
more symmetrical; nevertheless, the combination demonstrates 
a modification in the shape of the cyclonic system discovered from 
our last experiment. The system is more elongated than was 
shewn by the result of our first investigation, the longer axis lying 
in a north and south direction along the meridian of Greenwich, 
and forms a neutral lane about 30 miles broad between a southerly 
and a northerly going stream. With this exception, however, the 
present investigation confirms the deductions that were made in 
1909 regarding the direction and velocity of the deep currents, 
while the information obtained from the western areas, though 
scanty, is of much value in affording evidence of the easterly trend 
of the deep Atlantic water towards and through the channel lying 
between the Shetland and Orkney Islands into the North Sea 
