samples did not require preservation. The 

 latter, however, were "baited" with a pinch 

 of thorium carbonate (Harvey 1948) to pre- 

 vent the accumulation of phosphorus on the 

 sides of the container. An alternative 

 method that we found satisfactory and have 

 used on occasion consisted of adding the 

 amount of H2SO4 called for in the analytical 

 procedure immediately after collection. 

 Sulfide, CO2 , and O2 samples were analyzed 

 immediately after collection. We found the 

 practice of fixing O2 samples and titrating 

 them later resulted in an error because the 

 amount of free iodine in the supposedly 

 fixed sample increased with time. 



The rate of reactions, which includes 

 the development of color, depends among 

 other things on temperature. For this rea- 

 son the analyst must be sure that all sam- 

 ples and standards are brought to the same 

 temperature before adding reagents. Stand- 

 ard samples were interspersed among the 

 regular samples rather than being run as a 

 separate group. We have found that this 

 practice reduces operator bias. 



semipermeable membrane under freezing 

 conditions. Samples frozen in this type of 

 container have shown as much as a 2-percent 

 decrease in salinity concentration within 

 a year. Water from the same samples stored 

 at room temperature have shown a salinity 

 increase of over 1 percent during the same 

 period. 



Sample evaporation can virtually be 

 eliminated by sealing containers with poly- 

 seal screw caps. Salinity samples, for 

 example, have remained constant for as long 

 as a year and a half when sealed in this 

 manner, and we strongly recommend their use 



ANALYTICAL METHODS 



The colorimetric procedures that fol- 

 low refer to the Fisher A. C. electrophoto- 

 meter, the instrument used for the determi- 

 nation of the majority of samples collected 

 from the M/V Kingf ish and also the MA Alaska . 

 The colorimetric analyses of samples col- 

 lected from the M/V Pompano , however, were 

 carried out visually with Nessler tubes. 



SAMPLE STORAGE CONTAINERS 



The selection of proper sample con- 

 tainers is of considerable importance. We 

 have found, after considerable experimenta- 

 tion, that pyrex is equal to or superior to 

 soft glass, particularly for the storage of 

 total and inorganic phosphate and nitrate- 

 nitrite samples. During our survey work 

 (M/V Pomp ano ) , those samples that were not 

 analyzed on board ship (salinity and total 

 phosphorus) were stored in citrate bottles. 

 During the initial collection programs in- 

 volving the M/V Alaska , frozen samples were 

 stored in "Kimble" type screw-cap 25 by 200 

 mm. (Culture tubes. Salinity and unfrozen 

 total phosphate samples were stored in 4- 

 ounce prescription "Duroglas" bottles. It 

 was during the latter phase of the Alaska 

 program that pyrex was found to be superior 

 to the softer glass, particularly for the 

 analyses mentioned. Subsequent collections 

 (M/V Kingf ish ) employed pyrex storage con- 

 tainers for all samples except salinities 

 which were stored in 4-ounce prescription- 

 type bottles. 



We did not use polyethylene containers 

 because the material acts as an adsorbent, 

 reducing the reliability of an analysis. 

 Further, polyethylene becomes an effective 



Standard Samples 



With a few exceptions, the methods 

 described below require the use of labora- 

 tory-prepared standard samples. Initially 

 we employed distilled water in the prepara- 

 tion of these. However, it became evident 

 that the percent recovery depended, to some 

 extent, on the type of water. For this 

 reason we endeavored, when possible, to era- 

 ploy used portions of the samples for the 

 preparation of standards. If a sufficient 

 volume cannot be obtained for this purpose, 

 it is recommended that the percent recovery 

 be determined by "spiking" a few samples 

 with the standard solution. 



Standard samples were prepared from 

 compounds of known purity and formula 

 weight that contained the element, nutrient, 

 etc., under consideration. The compounds 

 selected were readily soluble but were not 

 hygroscopic or deliquescent to the extent 

 that they absorbed large amounts of atmos- 

 pheric moisture. The compounds used were 

 either bone dry or of a known moisture 

 content. This enabled all weights to be 

 reduced to a bone-dry basis. Final amounts 

 used in standard samples were obtained by 

 dilution of the relatively concentrated 

 stock solution. 



