802 TRUEBLOOD. 



likely that Griessmann at least must have had a not inconsiderable 

 outward heat-leak, the effect of which would have been to make his 

 observed low-side temperatures lower than they would otherwise 

 have been. Nevertheless, Griessmann's result for n at 165° C. is the 

 least of the three. It must be remembered in this connection that, 

 because of the subtraction method employed in calculating /x from the 

 data of all three observers, a considerable heat-leak mav easilv have 

 been present without a proportionate effect, or even, necessarily, an 

 effect of the same sign, on the calculated ^u. But it is hardly conceiv- 

 able that all of the discrepancy between the results of Griessmann 

 and of Peake, or, still more, those of Griessmann and of Grindley, 

 at 165° C, can be due to errors of observation; or, what amounts to 

 the same thing, that the width of 1° C. cm-'/kgni. of the band of points 

 of Davis' figure 6 can be due to this cause. It is not unlikely that the 

 high value of Grindley's result is in part a low-side kinetic energy 

 effect. His plug consisted of a quarter-inch glass plate through which 

 a single ^g-inch hole was drilled, and his low side thermometer (a 

 thermo-junction) was located in the jet from the orifice, about two 

 inches from the orifice. He made tests to determine whether altering 

 the position of the thermo-junction would affect the apparent low-side 

 temperature, and concluded that, on the whole, the indication of the 

 thermo-junction was independent of its position in the low-side 

 channel. His tests do not seem absolutely convincing, however. 

 Peake, who also used a single small hole (in a mica plate) as his main 

 throttling device, found on one occasion, when he accidentally omitted 

 the several layers of copper gauze regularly placed between the orifice 

 and his low-side thermometer, that, under the most unfavorable 

 conditions, the temperature registered by the low side thermometer 

 was thereby depressed l)y 7° C. Peake's maximum flow was nearly 

 twice Grindley's, however. 



Aside from these probabilities, one must ascribe a part of the lack 

 of agreement among Grindley, Griessmann and Peake to heat-leakage 

 effects about which it is impossible to be in any degree specific. It 

 happens that the result of tiuit one of tlu> three observers who made 

 tests showing that his heat leakage effects were negligible is the most 

 nearly in agreement with the value which the writer's work leads him 

 to believe correct. For the rest, the most that can be said is that the 

 eft'ects of heat leakage were probably small and may Im^e been of both 

 signs, and thus minimized in a mean of the results. 



Goodenough's empirical equation (equation (5)) gives, for ^t at 

 y = 4.0 kgm./cm.,2 T = 165 + 273° C. abs., the value m = 3°.34 C. 



