762 



MR J. D. HAMILTON DICKSON ON 



temperature were equal. By equating each of the three pairs of adjacent values of t, 

 three values of ^ t would be obtained, and, inserting the mean of these in the equation 

 for<5 = 2mm., a fair value — probably the best attainable — would be got for t. These 

 corrections could not be neglected for cadmium. Having regard to the value of the 

 correction, so long as tan 6 was positive (the present case) the correction for the same 

 given value of 8 increased as the temperature fell, and was additive. At 62°, the 

 temperature of the highest measurement, the correction for $ = 8 mm. would be 1"52° ; 

 and at — 148°, the temperature of the lowest measurement, it would be 2*56°. 



Ten measurements of dE/dt were made on the A-curve, and these were corrected in 

 the sense just indicated. The adjoining Table VIII. gives a couple of sets of readings 

 as they appear in the note of the observations. The columns are : — 



1. The deviates in mm. 



2. The lower temperature. 



3. The higher temperature. 



4. The readings of the tangent, as previously explained. 



5. The means of columns two and three. 



6. The tangent calculated from column four. 



7. The temperature finally adopted. 



8. The tangent on the same scale as equation (4(5). 



Table VIII. 



1 



2 



3 



4 



5 



6 



7 



8 



6 

 4 

 2 

 8 

 6 

 4 

 2 



41 

 37 

 29 

 -63 

 -69 

 -78 

 -89 



- "J) 



: f f 



-191 \ 

 -182 ( 

 -169 

 -150 ) 



170 + 348 

 100 + 200 



+ 2951 

 100 + 500 



13f j 



14 1 

 -127 . 

 - 125|- ( 

 -123H 

 -119i) 



1-73 

 0-49 



14J 



-119 



345- 



98- 



The ten values of the tangent thus obtained were 



Table IX. 



Temperature 

 Tangent/100 



62° 

 5-04 



37° 



4-28 



141° 

 3-45 



-7° 

 2-82 



-35° 

 2T6 



-53° 

 1-81 



-;<6° 

 1-33 



-101° 

 1-13 



-119° 

 0-98 



-148° 

 0-81 



Along the Tait-line, on fig. 10, which was drawn by means of equation (46), these 

 ten observed values of dE/dt were marked by small circles. Considering the difficulty 

 in measuring these values, and also that they were got from the free-hand curve of the 

 observations (the A-curve), their closeness to the calculated Tait-line affords a strong 

 presumption of the correctness of the theory. 



