METEORIC AND ARTIFICIAL NICKEL-IRON ALLOYS. 
37 
there was no asbestos box surrounding the ring within the beaker, and the heating- 
coil was between the other two. But in every case, for temperatures above 300° C. 
and below 800° C., the relation between temperature and current is almost exactly 
linear and can be expressed as 
(6+6 0 ) = mC, 
where m is the tangent of the angle which the 6c line makes with the current axis, 
and 6 is the temperature reckoned from 0° C. The values of m for the four series of 
experiments are in the approximate ratio :— 
Meteoric ring (1st winding) .... 1 
,, ,, (2nd winding) .... 1‘25 
Nickel-iron ring. 1'35 
Iron ring (see § 6, p. 39 below)' ... 1 '60 
whilst the corresponding values of 6 () are 275°, 250°, 285°, and 270° respectively. 
From the latter numbers the curious result follows that the absolute temperature, 
between the limits specified, was directly proportional to the current producing it. 
In fact, it was seen from the data that the actual temperature between 300° C. and 
800° C. produced by any current could be obtained correctly within a few degrees by 
determining a single temperature corresponding to a single given current, and then 
assuming in further experiments that the absolute temperature would be proportional 
to the current applied. 
There were only a few measurements at temperatures above 800° C. The data 
seem to show that the 6c curve slowly departs from linearity and becomes concave to 
the axis of temperature. 
Below 300° C. the 6c curve becomes convex to the axis of current. It would thus 
appear that the curve possesses an ill-defined point of inflexion somewhere between 
300° C. and 800° C. 
For temperatures below 100° C. the relation between 6 and c is approximately 
parabolic. 
§ 5. In practically all of the observations the resistance of the ring primary was 
measured in addition to that of the secondary. Corresponding values of the heating 
current were also tabulated. As it is impossible from want of space to give these data 
in the tables printed on pp. 42 to 47, the specimen curves shown in fig. 8 are reproduced 
to exhibit at a glance the degree of consistency attainable in the measurements. 
Fig. 8a, exhibiting the results of observations extending over three months, shows 
the extent to which the same heating current produced the same steady temperature 
on different occasions. Fig. 8b shows similarly the constancy of the ratio of the 
resistances of the primary and secondary at different steady temperatures, both coils 
being inside the heating coil and constructed of wire from the same reel. 
§ 6. The dimensions of the rings used have already been given, In the first 
