147 



slide than originally tested. This represents a difference in salinity 

 of 0.5 Voo. 



Further experimenting- Avith the apparatus revealed the fact that 

 a difference of 0.1° in the temperature of the bath causes a differ- 

 ence of 3.5 divisions in the reading of the slide wire or approxi- 

 mately 0.035 Voo difference in salinity, for a difference of one di- 

 vision of the slide Avire units is approximately equal to 0.01 Vgo 

 in salinity. From this it can be seen that the temj)erature of the 

 bath should be carefully watched, a variation of 0.03° causing 

 0.01 Voo variation in the salinity of the sample being tested. The 

 warmer the bath the lower will be the salinity record obtained. 



It is deemed advisable to recalibrate the apparatus at least once 

 a week and at more frequent intervals if there is any reason to 

 suspect evaporation or change in the Y cell. In this connection 

 there should be a sufficient supply of three standard known solu- 

 tions kept on board at all times. By such means a cross check can 

 be made at any time to locate a possible change occurring in any 

 one of the three solutions. 



