Subsurface Laboratory Methods 245 



ent equipment yields curves that are reproducible to a degree or so in 

 peak temperature and to five percent of the peak amplitude under nor- 

 mal conditions. 



Since all the differential thermocouples print at zero millivolts where 

 there is no reaction taking place, it is desirable to spread the six records. 

 This is done by a simple potentiometer circuit (fig. 102), which places the 

 base line of each record about one-sixth of a millivolt from its nearest 

 neighbors. The exact separation desired is achieved by adjusting a 200- 

 ohm resistance in the battery circuit. It is also desirable to have certain 

 sensitivity scales available, since some of the reaction minerals such as 

 alunite, jarosite, kaolinite, and carbonates may extend beyond the chart 

 on high sensitivity. Because this type of recorder measures the electromo- 

 tive force of the thermocouple, a simple voltage divided with proportionate 

 resistances is efficient for obtaining one-half, one-third, or any other pre- 

 determined fraction of the generated electromotive force. 



Finally, there are two solenoid pens in series, one attached to the 

 edge of each recorder. By means of a button switch, the solenoids are 

 simultaneously activated, thus marking both records at the same time. 

 Since the temperature at that instant can be read from the program- 

 controller record, the temperature of the six records is also known and 

 can be written on the multirecord chart at the completion of the run. 



The advantages of this equipment are worthy of note. One of the 

 greatest is the multiple-record feature, by means of which with three 

 runs eighteen samples may be tested conveniently in an eight-hour day. 

 Also significant is the reduction in the number of potential variables in 

 using six samples under the same heating conditions. This is important 

 when runs of quantitative mixtures are compared. The unit is compact, 

 it does not require a darkened room for operation as in the photographic 

 recording methods, and the results are immediately observable. The chief 

 disadvantage lies in the necessity for applying minor corrections to each 

 curve. 



Procedure 



The samples to be tested by the differential-thermal-analysis appa- 

 ratus are passed through a 50-mesh screen and packed to finger tightness 

 around the differential thermocouple. No pretreatment is given for an 

 ordinary run. It has been found by experimentation, as reported by 

 others, that any attempt to attain equilibrium with a specified humidity 

 merely alters the initial absorbed-water peaks (100°-200° C), the ampli- 

 tudes of which are usually not used for quantitative analysis. Ordinarily 

 weighing has been found to be unnecessary, and reproducible curves may 

 be obtained for the same pure substance with finger-tight packing with a 

 close-fitting metal plunger to a constant level. In special cases attention 

 must be given to the problems of particle size, weight, and humidity. 



After the samples are loaded, the cover is placed on the specimen 



