6G 
MR. H. N. DICKSON ON THE CIRCULATION OF THE 
numbers of observations are dealt with, as in the admirable charts published bv the 
Meteorological Office; the second must still be regarded as of doubtful value except 
whei'e its indications can be very fully controlled by observations of another 
kind (17). 
2. Observations of temperature; important at all depths, but especially at a 
considerable distance from the surface, where seasonal variations need not be taken 
into account, and where differences of salinity are extremely small. Temperature 
observations give as yet by far the most trustworthy information about the slow 
vertical or horizontal “creep” of waters in the greater depths. In shallow waters 
near land, temperature observations may be exceedingly difficult to interpret; the 
oceanic waters tend to have a small annual range of temperature which may vary 
greatly according to the degree of mixing of the surface waters by sea disturbance, 
tidal streams, &c., while the normally greater annual range of the land waters is 
affected by the source from which tliey ai-e derived, e.y., glaciers, &c. 
3. OlDsei’vations of salinity. In the cases just mentioned, and indeed in all surface 
waters, the amount of salt affords a surer guide as to movement than does the tempera- 
tui-e, provided the determinations of salinity are made with sufficient exactness, chiefly 
because it is not liable to considerable local or seasonal variations, except by active 
mixing of w'aters from diffei'ent sources. Tims in the open ocean, <')bservations taken 
in a small area and witliin a few days of each other may show considerable irregulari¬ 
ties of temperature, but the coi'responding salinities will agree very closely ; hence it 
is not necessary in the latter case to “ generalise ” the observations in showing the 
isohalines, and thereby introduce risk of erroi' due to misinterpretation, to the same 
extent as in drawing the isothermals. 
4. Dissolved Gases. The results of analysis of the gas ct)ntents have recently been 
employed with marked success I)y Pettersson (18) and Knudsen (19) in traciug the 
source of origin of different ocetiiiic waters. It is unnecessary to refer to this in more 
detail here. 
5. Plankton. Qualitative and quantitative examinations of the plankton contained 
in samples of sea-water have recently, in tlie opinion of Cleve and others, proved a 
valuable assistance to the physical and chemical methods mentioned (20). 
In attempting to investigate the movements of water within a given area by 
applying the synoptic method to tlie distribution of tlie elements just enumerated, 
the limitations imposed are determined by the extent of that area and the nature and 
rapidity of the changes from the observation of which the movements are to be 
inferred. Wliere the changes take place slowly, observations distributed over a 
considerable period may be regarded as having been made simultaneously in the 
middle of that period, and treated as in all respects comparable with each other; 
and further, if with a large number of observations distributed fairly evenly over 
the “ period of observation,” we find tliat the condition revealed by successive sets of 
observations can be shown to follow iiaturally Iroin h^qiothetical continuous changes 
