610 THURAS: DENSITY MEASUREMENTS 



requirements are simplicity and convenience of operation, free- 

 dom from disturbances due to the motion of the ship, rapidity, 

 accuracy, and limitation in the quantity of test water used. 



The apparatus is simple and rugged. All of the glass parts 

 except the thermometer and bobbin are stock articles, and the 

 bath is so constructed that it can easily be replaced if broken. 

 The heating and cooling elements and the stirrer are contained 

 in the copper base and are protected so that they can not easily 

 be injured. The operation of the bath is simple and convenient. 

 The test tube containing the liquid and bobbin is placed in the 

 bath and, by means of the heating coil and cooling tubes the 

 temperature is varied until the equilibrium point is reached. 

 The only readings necessary are the temperatures of the bath 

 for the final two movements of the bobbin. 



An instrument that is to be used on board ship should 

 operate rapidly. The bath and test tube containing the sample 

 were therefore made as small as possible, the heating coil was 

 placed close to the liquid of the bath and the cooling water was 

 passed directly into the liquid. With full voltage on the heat- 

 ing coil the bath can be heated 10° in less than two minutes, 

 and can be regulated by an external resistance to 0?01. The 

 test tube containing the sample is 18 mm. in diameter and has 

 such a small temperature lag that under ordinary circumstances 

 a single test can be made in less than ten minutes. 



The accuracy in density determinations demanded in oceano- 

 graphic work is 1 or 2 units in the fifth decimal place; this re- 

 quires that the equilibrium temperature be read to 0?05 C. In 

 figure 2 are shown for each of four bobbins the curve connect- 

 ing the equilibrium temperature and the salinity and density 

 of the sea water in which it is immersed. Repeated tests on 

 the same sample over several days gave consistent results, and 

 no two tests of the equilibrium temperature had as large a dif- 

 ference as 0?05. 



From some subsurface samples only a limited quantity of the 

 water can be used, as most of the sample is required for bio- 

 logical purposes. With this apparatus 30 cc. is sufficient for 

 a test, and with a little more difficulty a test can be made with 

 only 15 cc. 



