— XXIX — RAPPORT 1905—06 ANNEXE A 
and from the surface map can be determined their points of intersection with the 
surface-line of she section. 
Plates XII—XV contain the sections indicated on Plate XI; on the sections 
are to be found, besides the isohalines, by way of further information, some actual 
mean values for certain points. 
The salinity in the depths is also represented by level-maps (Plates IIT—VIII) 
for the depths of 20, 40, 60, 80, 100, and 200 meters and by two bottom-maps 
(Plates IX and X). 
The level-maps are drawn on the basis of the sections, the surface-map and 
the previously mentioned station-curves, these last being used in the case of those 
stations through which no section passes; on the other hand the level-maps have 
been of use in improving the sections and have occasioned correction of the sta- 
tion-curves. 
The bottom-map (Plate IX) over the mean salinity has been drawn ex- 
clusively on the basis of the sections and the level-maps and, so far as the Channel 
and the southern part of the North Sea are concerned, with the aid of the sur- 
face-map. 
The bottom-map (Plate X) gives for each station the highest and the lowest 
salinity found at the station. 
As is evident from the above remarks, the maps and sections are not a 
rude reproduction af the mean salinity but are so far as possible subjected to 
graphical adjustment. This process has been especially effective in the construction 
of the surface-map on account of the great mean deviations and the great irre- 
gularity in time that characterize the surface observations, but it has been easier 
to carry out in this case on account of the closeness of the observation-places. 
For the subjacent layers of water the adjustment is less pronounced, and 
_ this for two reasons. In the first place the mean deviation of the observations is 
less and accordingly the scope for adjustment less; but in the second place even 
within this scope it has heen impossible to carry out the adjustment as far as de- 
sirable for lack of sufficiently close observation-material. For the stations are as 
a rule so far from each other that a mean value at one station does not get much 
support from mean values from other stations. An exception to this are the Scotch 
stations, a circumstance which has been made use of in the case of the Scotch 
sections XI, XII and XIII, which have been constructed on the basis of the monthly 
Scotch stations. Here it has been possible to adjust with support from neigh- 
bouring Scotch stations, which are found throughout the whole period from August 
1902 to May 1905, so that section XI, XII and XIII do not need to be considered 
as especially representing the period September 1904—May 1905, but may be taken 
as parallel to the other sections. 
