18 MARION AND GENERAL GREENE EXPEDITIONS 



In detei'iiiining the depths of the observations in 1983 a combina- 

 tion of meter-wheel readings and unprotected reversing thermometers 

 was used. The deepest bottle of a series carried one protected and 

 one unprotected thermometer. At stations where two series were 

 necessary, unprotected thermometers were attached to the upper- 

 most, deepest, and middle bottles of the deep series. The depths 

 indicated by these unprotected thermometers were used in conjunc- 

 tion with the meter-wheel readings to determine the depth of re- 

 versal for all the bottles. Whenever conditions seemed favorable, 

 that is when there was little wind and a small wire angle, oppor- 

 tunity was taken to check the pressure coefficients of the unprotected 

 thermometers. The pressure coefficients so obtained w^ere based on 

 the assumption of an accurate meter wheel and were consistently 

 higher by about 3 percent than the coefficients given in the test 

 certificates. These experimentally determined pressure coefficients 

 were used in deriving the depths of reversal. However, it is probable 

 that the pressure coefficients given in the test certificates are more 

 accurate than the meter wheel. The listed depths of the observa- 

 tions therefore are probably too shallow by about 3 percent. 



During July 1934 the General Greene ran two lines of oceano- 

 graphic stations across the shelf northeast of Newfoundland and a 

 complete traverse of the Labrador Sea from southern Labrador to 

 Cape Farew^ell, Greenland. Nansen water bottles and Richter & 

 Wiese reversing thermometers were again used. The same time in- 

 tervals were allowed for the thermometers to attain temperature 

 equilibrium, and the same messenger speeds of travel were used as 

 in 1933. 



A brief description of the details of the methods employed has 

 been given by Soule (1935) (pp. 49-58). Provision was made for 

 the determination of salinities by either the silver nitrate titration 

 method or the electrical conductivity method. A new model Wenner 

 salinity bridge was received during the season and w^as calibrated 

 with titrated samples as described by Soule (1935). This new model 

 embodied many of the improvements in construction recommended 

 by Wenner, Smith, and Soule (1930). All routine measurements 

 of salinity were made with the neAv salinity bridge and each sample 

 was so measured twice. During the season, and on the cruise under 

 discussion, a total of 2,570 measurements were made of half that 

 number of samples. No tAvo measurements of the same sam])le 

 differed by more than 0.015Voo in salinity, so it was not necessary 

 to measure any of them a third time. All measurements were re- 

 ferred to Copenhagen standard water of the batch P13, the same 

 batch being used throughout so that any variation in salt ratios 

 which might possibly exist between different batches would not 

 invalidate the calibration curve of the bridge. Copenhagen standard 

 water was used for every series of measureuients, and either Copen- 

 hagen standard or a substandard water was used in each cell once 

 every 10 or 12 measureuients. All titrations and the routine bridge 

 measurements were made by the oceanographer's assistant. 



As a result of careful comparisons of the sinuiltaneous measure- 

 ment of samples by both titration and new mod(d salinity bridge 

 methods the conclusion has been reached that at least under condi- 

 tions existing on board the General Greene at sea the titration 



