184 



MR THOMAS HEATH ON 



The values of 



J-rrC 



VI 



deduced from the epoch and amplitude of the annual term 



agree amongst themselves much better than the values derived from the semi-annual 

 term. In the case of the latter, however, the coefficients are so small that any better 

 agreement could hardly be expected. It will be seen from these tables that all the 

 four mean values derived from the new thermometers are somewhat greater than those 

 from the old, and that the two values from the old and new derived from the 

 amplitudes of the semi-annual term are both greater, and by nearly the same amount, 

 than those derived from the retardation of phase. The results from the annual term 

 specially show also a distinct tendency to decrease as the deeper thermometers are 

 approached, a tendency which was pointed out by Lord Kelvin in his paper referred to. 

 This will be more clearly seen from the following table, showing the results from both 

 sets of thermometers and from both the terms of the equations, arranged in the order 

 of the means of the depths of the two thermometers concerned in each case. 



VALUES OF 



/ttC 



VT 



IN OEDEE OF MEAN DEPTH. 











Mean 

 Depth. 



Annual Term. 



Semi-Annual Term. 









From 

 Phase. 



From 

 Amplitude. 



From 

 Phase. 



From 

 Amplitude. 



Means. 



Old t x and t 2 

 Old t x and t 3 

 New t x and t 2 

 Old t x and t 4 

 New tj and t 3 

 New t l and t 4 

 Old t 2 and t 3 

 Old t 2 and i 4 

 New t 2 and t 3 

 New t 2 and t 4 

 Old t z and t 4 

 New t 3 and t 4 









19-2 



16-0 



16-0 



14-4 



13-0 



12-0 



9-6 



8-0 



8-0 



7-1 



4-8 



4-0 







107 

 109 

 113 

 110 

 115 

 115 

 115 

 114 

 117 

 116 

 114 

 116 







102 

 106 

 107 

 108 

 111 

 112 

 115 

 117 

 117 

 119 

 123 

 123 







107 

 103 

 131 



103 

 124 

 118 

 098 

 099 

 113 

 102 

 101 

 064 







099 

 110 

 084 

 112 

 107 

 113 

 133 

 130 

 144 

 149 

 125 

 168 







104 

 107 

 109 

 108 

 114 

 115 

 115 

 115 

 123 

 121 

 116 

 118 



The tendency of the value of / — to decrease with increasing depth is very apparent 



from about 7 feet deep, to the lowest depths in question. The shorter thermometers 

 of each set, however, show strange irregularities, which seem difficult to explain, 

 except on the supposition that owing to surface conditions different from those assumed 

 as the foundation of the harmonic law, they do not follow that law with so great a 

 degree of exactness as the deeper ones. In the case of the results from new t 3 and t A , 

 some part of the irregularities may also be due to the sinking of these thermometers, 

 to which I have already referred, since their establishment in the rock. The general 

 tendency of the deeper thermometers to give smaller values is, however, apparent, and 

 seems to suggest that the conducting power of the rock increases with depth. 



