130 BEPORT— 1892. 



so that the points of one series fell midway between those of the other 

 eeries, thus touching every 125 ft. or thereabouts. When it is considered 

 that these two series were taken a month apart, and yet nowhere differ 

 by more than 0°-3 F., it seems pretty good evidence of the accuracy and 

 constancy of the results. Many of the points, as at 1,590, 2,500, 3,000, 

 3,500, 4,000, and 4,460, were checked as many as four or five times.' 



The thermometei'S were raised and lowered by steel wire, which also 

 served to measure the depths. They were lowered to place about 11 a.m., 

 and left till next morning, though trials showed that they had practically 

 i-eached a stationary condition by 5 or 6 P.M. Sometimesthey were left in 

 for several days. They were of the Negretti maximum kind, used upside 

 down. They were carefully compared, and corrections were applied not 

 only for their index errors, but also ' for the contraction of the recording 

 column in cooling from the maximum tempei'atnre to the temperature at 

 which they were read. They were put two (sometimes three) together, 

 and hence always checked each other, agreeing very satisfactorily. They 

 were hung upon a long spiral spring in the zinc bucket in water, or upon 

 the wire 250 ft. from the bucket at the end.' 



The well is cased with iron tubes to the depth of 1,750 ft. There 

 are three of these tubes, the first reaching from the top to the depth 

 of 400 ft., the second from the top to 1,200 ft., and the third or innermost 

 from the top to 1,570 ft., the internal diameter of tliis last being 4| inches. 

 No reliance was placed on the temperature of the cased portion. 



The strata are nearly horizontal, their dip being only 50 ft. to the 

 mile ; and the uncased portion consists almost entirely of shale, the only 

 important exception being an oil sand at from 2,900 to 3,000 ft. The 

 conditions are eminently favourable for the purposes of the Committee. 



In view of the disturbance of temperature by convective circulation 

 of water, which was proved to exist in the deep bores at Sperenberg and 

 Schladebach (see Reports for 1876 and 1889), and was obviated by the 

 use of special plugs for isolating any required portion, Mr. Hallock made 

 careful experiments on the effect 'of isolation. Two series of woollen 

 wads were placed ten feet apart, and thermometers in the intervening 

 space gave the same indications as other thermometers above and below 

 the wads. ' In fact a thermometer simply tied on to the steel wire 

 recorded the same as an isolated one.' Nevertheless, as a measure of pre- 

 caution to prevent circulation, long blocks nearly filling the hole were 

 nearly always placed above and below the thermometers, except when 

 they were in the long buckets. 



In endeavouring to account for the absence of convection in this dry 

 bore, we are confronted with the fact that air is much more expansible 

 than water, and therefore more easily put in motion by difference of 

 temperature. On the other hand, the process of 'jumping ' which was 

 employed in the present case leaves the sides much rougher than diamond 

 boring, which was the method employed at Sperenberg and Schladebach ; 

 and this roughness tends to hinder circulation. 



The main difference to which we must look for the explanation is the 

 difference of thermal capacity ; this capacity, for equal volumes, being 

 about 3,000 times as great for water as for air. It would thus appear 

 that the air is so easily warmed and cooled, that the heat whicli it gives 

 and takes does not sensibly affect the temperature of the walls, surrounded 

 as they are hy a practically infinite extent of rock tending to maintain 

 them at their original temperatures. 



