guard formed a subassembly which was inserted in a plastic tube 
and collar after connections to the main cable were made. The 
plastic tube was threaded to fit a standard size A, brass stuffing 
tube that provides a watertight seal between the cable and the unit. 
The tube was then filled with epoxy resin, leaving the tip of the 
thermistor bead extending approximately 1/8 inch. 
The thermal time constant of the thermistor thus encapsulated 
was approximately 1/2 second. Since the temperature readings 
were to be taken at intervals of from 10 seconds to 1 minute, 
"aliasing" of the temperature signal could occur unless a thermal 
lag were introduced. For this reason, the thermal time constant 
was increased to about 16 seconds by applying several coats of 
Armstrong A-2 epoxy resin. Measurement of the time constant 
during this process was done by rapidly immersing the thermistor 
unit in a homogeneous water bath that was maintained several de- 
grees below room temperature. The thermistor unit being tested 
was used with a bridge circuit (to be described in the next section) 
whose output was fed to a Minneapolis-Honeywell potentiometer 
strip chart recorder with a full scale pen travel of 1/4 second. 
From this temperature trace as a function of time, the time con- 
stant was easily determined (fig. 4). 
Since the temperature of the probes was to be recorded 
electrically, a bridge circuit was needed to provide a de voltage 
output proportional to the temperature. The basic form of the 
bridge that has been used, called an asymmetric half-bridge, is 
shown in fig. 5B. In order to obtain a linear voltage output, € 9, 
with respect to temperature, the value of 7, is of prime impor- 
tance. For a given set of normalized thermistor sensors there is 
an optimum value of this resistor, 7, for linear output. A method 
for obtaining this optimum value is described in Appendix II. In 
order to adjust the bridge to measure a certain temperature range, 
both zero and span controls are used. Thus, the bridge can be 
balanced for zero output voltage at any desired temperature, and 
the span can then be adjusted to make the full scale temperature 
compatible with the limits of the recording system. 
In the equivalent circuit (fig. 5A) shown the ratio of #./7, 
adjusts the zero or balance point of the bridge, while the sum, 
