Tlic por cont saturation of dissolvod oxygen wcls calcul.:itad from 

 tables proparcd at tho Ocoanographic Laboratories. These tables are based on 

 those of V/hipple and V/hipple (1911). 



Soluble Phosphatss . Soluble phosphates wore dotormincd by the 

 ceruloraolybdato method of Donigos (1920) according to tho modification of 

 Truog and Meyer (1929). In order to componsatG for the salt error, phosphate- 

 free soa water xi-.s used in the preparation of all comparison standards. So 

 that no error would be introduced by storage, analyses were made on board 

 ship as soon as possible after the samples had attained the temperature of 

 the laboratory. This same procedure was followed in the determination of 

 silicate and nitrite. 



Determination of Silicates . The soluble silicates were detemiined 

 by means of the silico-molybdatementod using picric acid standards as out- 

 lined by Thompson and.Houlton (1933). 



Nitrlte-rlitrogon . The detemiination of nitrites was iiadeby Ihe 

 Golorimotric method of Griess (1879) as modified by Ilosvay (1889). Nitrite* 

 free soa v/ator was used in the preparation of comparison standards to 

 compensate for salt effects. 



Determination of pH . A Hellige comparator was used with crosol rod 

 as the indicator. The data in the tables are not corrected for temperature' 

 or salt error. 



D irect Current . Measurements .' Tho CIIELAN vms anchored- at sereral 

 shallow stations and the water current measured directly by moans of a current 

 meter of tho latest Ekraan type (1932). The current meter had boon calibrated 

 just before use by Professor Y, V/dlfrid Ekman of Lund, Sweden. Tlic cum ic 

 magnitudes listed in tho tables arc expressed in knots, and directions in 

 degrees magnetic. Ro attempt has been made to correct the directions for 

 local variation or the deviation due to the iron masses of the ship. The 

 deviation caused by the ship's magnetism varies with the ship's heading and 

 the depth of the current meter. In most cases it is probably not more than 

 10° at 10 meters below the surface. 



Dynamic Computation s and Calculated Currents . The densitie&wof ".■ 

 the water sainples at atmospheric pressure and the temperatures at which they 

 occured in the soa were ascertained from the taiperature and chlorinity 

 data by the use of Knudsen's Hydrographlcal Tables (1901). The densities 

 in situ and reciprocal quantities, the specific volumes in situ, dynamic 

 depths, .and relative velocities were then obtained by methods based on tho 

 Bjerknes theory (1910) similar to those outlined by Hesselberg and Sverdrup 

 (1915) and Saith (1926). Tables by the latter authors have been modified at -^ 

 the Ocoanographic Laboratories to give chlorinity-prossure corrections 

 directly rather than salinity-pressure corrections thus eliminating frcm 

 dynamical computations the unnecessary step of obtaining salinity. Currents 

 as obtained by moans of Bjerknes theory arc relative, the currents of the 

 upper water level being obtained in reference to tho lowest observed level, 

 where it is assmaod that thero is 'no movement of the water, or, expressed 

 differently, the current near tho bottom is taken as zero. The values for 

 calculated currents appearing in the tables arc based on this assuiaption. 

 Those relative values arc alraost identical with tho absolute values at • 

 stations taken in deep water, but in shallow water in which there is an 



- 7 - 



