150 



The Ohio Journal of Science [Vol. XVIII, No. 5, 



IV. Various Corrections. 



The sets of curves for Figure 3 for different values of y 

 cannot be directly compared with one another for several 

 reasons. First, those of 7/ = 8 cm. (together with the fourth 

 and sixth sets shown in Table I) were taken with a different 

 thermo-couple from the rest. Second, the different sets were 

 taken often several days apart and the oscillator deterioration 

 as well as possible change in sensitivity of the thermocouple 

 needed to be eliminated or corrected for. Third, the change 

 in room temperature, while not great, undoubtedly affected the 

 dielectric constant of the olive oil used in the spark-gap and 

 consequently the energy as read at the receiver. 



y**»-^ y^^£z 



9^.-^ 

 L/13.7 

 "K 9Z.O 



= 69.1 



^ri d§e T^osi ti on 



Figure 4. 



In the light of our experiments there seems to be a distinct 

 deterioration in the sensitivity of our thermocouple even though 

 soldered. For with a pohshed spark-gap and other conditions 

 the same the intensity was often distinctly less for all the tones 

 of a set for a given y than when the same set was taken earlier. 

 Any figures we could give in illustration of this fact would not 

 be worth much, for manifestly the other factors, such as oscillator 

 deterioration and temperature change, are not readily differ- 

 entiated from the sensitivity changes of the thermocouple. 

 The fact that these three factors cannot be readily differentiated 

 one from another shows, too, the futility, for the object we 

 had in mind, of employing a check receiver. Moreover, while 



