56 



ning as quantitative physical oceanography did, let us start out with the use of a 

 bucket and a thermometer. The purpose of emphasizing such a simple piece of 

 equipment is because much good work has been done with it. Bucket tempera- 

 tures can be either correct or misleading depending upon the degree of thermal 

 layering near the surface and depending upon the skill of the observer and inter- 

 preter. The simple bucket and thermometer will probably continue to be a 

 mainstay for calibration purposes for a long time. In the hands of a good ex- 

 perimentalist it is also probable that good research work will be done with it in 

 the future. For example, a modern version of the bucket is to use a hose and 

 to pump water with it from a given depth to the surface. 



Water Intake Thermometers - Much of our knowledge of horizontal gradients in 

 the ocean has come from a study of intake temperatures on ships. This has 

 seldom given data which permits the study of fine horizontal structure but it 

 would seem that this could and should be done. The sensitivity of the apparatus 

 should be improved as well as the time constant. Practice has indicated that 

 an observer usually has to go with the thermometer to insure that the necessary 

 auxiliary data be obtained. 



Reversing Thermometer - From 1870 until 1938 the two principal tools in phys- 

 ical oceanography were the reversing thermometer, to determine temperature, 

 and the Nansen bottle to obtain water samples for salinity and oxygen determina- 

 tions. This thermometer combined a simplicity of operation, a high absolute 

 precision, and was operable over a wide depth range. Another excellent fea- 

 ture of the reversing thermometer is that when used in protected and unprotected 

 pairs they constitute their own depth gage. These desirable characteristics 

 have made it a standard oceanographic instrument and one which will not be en- 

 tirely replaced in the forseeable future. 



It does not seem necessary to describe' the reversing thermometer in de- 

 tail, as it is well known among oceanographers, and has been described several 

 places in the literature. For the moment let us say that, under ideal conditions 

 reversing thermometers can give temperature to about 0.02° C and depths to 

 about 10 meters. The reversing thermometer, however, suffers from a slow 

 speed of response and from being able to give data at only a limited number of 

 depths. 



Because the reversing thermometer has become a standard tool in ocea- 

 nography, it will pay us to critically examine its limitations so that we use them 

 on experimental work for which they are best suited and do not improperly inter- 

 pret results on experimental work for which they are unsuited. 



Because of the nature of the use of reversing thermometers it is doubt- 

 ful if they will ever be well suited for studying small scale or rapid thermal 

 changes in the ocean. They are excellent, however, for studying the gross fea- 

 tures of the ocean and for measuring stable temperatures at great depth. An- 

 other virtue of the reversing thermometer is that it has been paired for many 

 years with the Nansen bottle and hence the general utility of each device is great- 

 ly enhanced by the presence of the other. 



Bathythermograph - For many years there had been a need for a practical de- 

 vice which would make a continuous temperature measurement. In 1938 Spill- 

 haus designed the bathythermograph which was an early practical temperature 

 depth recorder. Furthermore, it was possible to use it from a slowly moving 

 vessel. 



In 1940 and 1941, Ewing, Worzel and the writer modified the bathyther- 



