67 



ture. The aim of the hydrographic cast is to obtain representative distributions 

 of temperature, and to subdue any erratic variations, of periods less than one- 

 half minute. The roll of the ship, for instance, will cause temperature oscilla- 

 tions in the environment of the thermometer. Because of the slow response of 

 the thermometer, however, these oscillations will be damped out. 



Accuracy of Thermometric depths - The accuracy of the thermometric meas- 

 urement of depth by means of the (U)-(P)* pair of thermometers is severely 

 limited in the region of the thermocline by virture of the motion of the wire and/ 

 or internal waves. The presence of a temperature oscillation in the environ- 

 ment of the (U)-(P) pair, together with the fact that the structure of the (U) ther- 

 mometer is such that its thermal response time is much smaller than that of the 

 (P) thermometer, brings about errors in depths by virtue of the disparity of the 

 thermal response components in the scale readings. This can partially be elim- 

 inated by using some sort of standardized pressure transmitting plug for the (U) 

 thermometer which will bring its response time up to a value more nearly cor- 

 responding to that of the (P) thermometer. 



Reliability of Reversing Thermometers - A system of classification of re vers - 

 ing thermometers derived from the history of the individual thermometers is 

 useful in the processing of data. It has been found useful in the work of the 

 Marine Life Research Program at the Scripps Institution of Oceanography, as 

 well as in the Gulf Survey Program at Texas A. &M. to classify thermometers 

 according to the percentages of failures (a record of which is made on the hydro- 

 graphic data sheets). In addition, the statistics of the discrepancies in readings 

 between two (P) thernnometers placed on the same bottle is useful in assigning a 

 reliability value to the thermometers. 



DISCUSSION: Theodore R. Folsom** 



The importance of new instruments to the advancement of science is often 

 mentioned and theoretical and experimental problems involved during develop- 

 ment are frequently stressed -- but it is seldom stressed that plain manufactur- 

 ing difficulties frequently limit availability. It is true that sometimes a long 

 wait is required before a valuable flash of design inspiration occurs -- but all 

 difficulties do not then simply melt away. There follows a painful period of 

 trial and test -- and a sometimes much more painful period when physical con- 

 struction is carried out. 



Many scientific instruments require craftsmanship of exceptional quality, 

 a commodity not so abundant in recent years; but whether skilled or unskilled, 

 present costs are extremely high. To minimize this ever-growing expense two 

 expedients appear necessary; an increased investment must be made in produc- 

 tion tools of the labor saving type, and a search and procurement organization 

 must be set up so that maximum number of mass-produced components can be 



* - (U) - unprotected; (P) - protected. 



** - Dr. Folsom 's discussion of Allyn Vine's paper has suggested to the 

 editors a broad and important area not covered at the 1952 symposium -- the 

 actual manufacture of research instruments, which is certainly a very large 

 part of the whole instrument problem. The editors, therefore, have asked Dr. 

 Folsom to generalize his original discussion with this point in mind. It is the 

 editor's opinion that another meeting profitably could be devoted entirely to this 

 subject. 



