34 ROYAL SOCIETY OF CANADA 



ditfusivity due to pure thermal conduction per se, and that more than half 

 of the average annual value is, therefore, due to the effect of percolation. 



The remarkably sudden fall of temperature which occurred between 

 November 27th and November 28th, in a period of eighteen hours, if 

 worked out on the same method as the longer periods, gives a value of 

 the diffusivity 0*323, which is about 200 times greater than the value, 

 •0015, due to thermal conduction, apart from percolation. Such cases are, 

 of course, exceptional. It is rarely that so large a quantity of cold water 

 can accumulate, and then be absorbed in so short a space of time. 



Eesults derived from similar observations with respect to the con- 

 ductivity of the soil have often been applied to estimate the possible age 

 of the earth. Geologists have, as a rule, complained that these estimates 

 did not allow then as much time as was required by other evidence for 

 the evolution of the earth. Since, however, these estimates of conduc- 

 tivity have generally neglected the effect of percolation, which must be 

 very great in the surface soils and in rocks which are permeable to water, 

 it is plain from the evidence which we have adduced that such estimates 

 of the age of the earth may require to be largely increased. 



The method of comparing the conductivity at different depths, 

 which was mentioned last year, by supplj'ing a known quantity of heat 

 through one of the thermometers, and obseiwing the resultant rise of 

 temperature, has been applied in a few instances during the past year. 

 The results are not as yet sufficiently exact and extensive to warrant any 

 important conclusions, bvit so far as they go they are in agreement with 

 those previously obtained. The conductivity of the lowest stratum, at a 

 depth of 108 " in the blue clay, was found to be greater than that of the 

 sand in the ratio of 41 to 36. The conductivity of the sand was found 

 to be very uniform, as had been previously inferred. 



Automatic Eecording Apparatus. 



After spending a considerable amount of time in experiments, a 

 satisfactory method has at length been evolved of making the apparatus 

 self-recording. This is accomplished by means of an electro-magnetic 

 mechanism of a complicated nature, which writes the record with pen 

 and ink on a revolving cylinder. It has been found possible to make the 

 apparatus work satisfactorily on a scale as large as one inch to the degree 

 Fahrenheit, and it is hoped that in special cases the records may be made 

 accurate to the hundredth of a degree. The same electrical thermometers 

 which have been in use for the last three yeara have been arranged in 

 such a way that any one of them can be connected at pleasure in a few 

 seconds, either to the automatic recording apparatus, or to the galvano- 

 meter with telescope and scale, which has been in use hitherto. It is thus 

 possible to check the readings of the recorder with very considerable 

 ecu racy. 



