Determination of salinity, as the total measure of inorganic dissolved matter, is by evaporation of the water and weighing of the residue. This is a difficult process, because some carbon dioxide and hydrogen chloride escape during the evaporation process and corrections must be made for this. Furthermore, at sea methods involving weighing cannot be used. So the methods to be applied on board a ship have to be indirect ones. In the past century only two major methods have been used in the oceanography for measurement of the seawater samples salinity: chlorinity titration and conductometry.

1. Chlorinity Titration

Chemical determination of halide content by titrationwas for many years the usual routine method for determining the "sea salt" content of sea water samples.

Karl Mohr

This method, which known as Mohr method (Mohr, 1856), consist of titrating a sample of seawater with silver nitrate solution of known concentration to the point where all halides (chloride plus a small amount of bromide ) have been precipitated as silver halide, as detected by suitable indicators or electrode systems. A 15ml Knudsen pipette is used to measure the sea water sample into the titration vessel. This pipette differs from the standard type in that, after filling, the volume of sample is defined by rotation of a 3-way stopcock fitted at the upper end.

The titrant used in Mohr method is silver nitrate:
Cl- + Ag+ = AgCl


Usually most of the silver nitrate is added as a strong solution, to just short of the end-point , and then the titration is completed with a more dilute solution of silver nitrate. The other halides present are similarly precipitated. Potassium chromate is added as an indicator so that, when the halides have been titrated to a low level at the end point, silver chromate is precipitated:


2 Ag+ + CrO42- = Ag2CrO4


The titrant is normally added from the Knudsen bulb burette. Again, this is designed so that it can be used for routine analysis at sea; the burette is filled by gravity from a large storage container, and a zero mark is defined by a three-way stopcock at the top, as in the Knudsen pipette. Since most open-ocean samples lie in a relatively small chlorinity range, the burette is designed so that much of its capacity is in the bulb. This allows the scale to be graduated in increments of 0-0.02 ml to improve precision. Chlorinities between 16‰ and 21‰ can be determinent using silver nitrate solution of 36.75g/l; outside this range the titrant strenght must be adjusted to that the titre falls on scale.


Before titration commences, five drops of 10% potassium chromate solution are added to the sample. The silver nitrate is then added as a fine stream from the burette, with strong magnetic stirring to break up the silver chloride flocs. The endpoint is indicated by the precipitate turning pale brick-red for more than 30sec after addition of the last drop increment of silver nitrate.
The silver nitrate solutions are calibrated against IAPSO Standard Seawater certified in chlorinity (nowadays only inK15)

Knudsen Burette

The chlorinity of the unknown seawater is calculated as follows:

Clu = Cls ×Tu × Ws \ Ts × Wu

where:

Clu and Cls - chlorinity of unknown and standard,

Wu and Ws - weight of unknown and standard,

Tu and Ts - silver nitrate titre for unknown and standard.

A modification of the Knudsen titration has been suggested by the Grasshoff and Wensk (1972). This uses a Metrohm incremental piston burette in place of Knudsen burette; the absence of greased stopcocks from the system is stated to improve markedly the convience and accuracy of shipboard analysis.
Potentiometric end point determination has been utilized by several workers to enhance the precision of the silver nitrate titration. Reeburgh and Carpenter (1964) used a differential electrochemical system for end point detection.
One of the major drawbacks of manual titration methods lies in the time taken per sample and the operators skill required. A semi-automatic method of chlorinity titration has been described (Jarner and Aren,1970) which reduces the time per sample to 5min while retaining high precision (0.004‰ in chlorinity).
Chlorinity was then converted to salinity by means of equation [1] or later equation [2] , prior to the introduction of the Practical Salinity Scale 1978. Chlorinity is now regarded as an independent chemical parameter to describe the properties of seawater and has no defined relationship to salinity.
2. Electrical Conductivity.

In the past fifty years, the chlorinity titration, which was time-consuming and required a certain degree of analytical skill, has been largely replaced by the measurement of electrical conductivity as a mean of estimating salinity.
The laboratory method consist of comparing by use salinometers the electrical conductivity of the sample with that of a standard (IAPSO Standard Seawater) of known salinity at the same temperature.

The measured conductivity ratio is than converted to the practical salinity by means of the equation of the PSS-78