496 report — 1878. 



centre, and this displacement was towards the-analogous-pole ; showing that while 

 the temperature was rising, the conductivity in that direction was greater than in 

 the opposite direction. When, however, the experiment was done slowly with a 

 thicker crystal, so that thermal equilibrium was gradually attained, no such 

 unilateral effect could be observed. Rough preliminary experiments snowed the 

 unequal semi-axes minor to have a ratio of about 10 to 12, but there was con- 

 siderable discordance between the various results. 



A calorimetric method was next adopted to measure the flow of heat across a 

 thin wall of tourmaline cut normally to its crystallographic axis. The thin slice 

 was fixed between two similar portions of glass tubing, either end of which could 

 therefore be made to hold a weighed quantity of mercury whilst steam was blown 

 up into the other. In this way the heat which passed upwards through the crystal 

 when one surface was maintained at 100° could be measured in either direction. 

 Experiments were made alternately, the times required to heat the mercury 

 through a given range of temperature being compared in the two cases. To 

 eliminate error, after half the experiments had been made the crystal slice was 

 itself reversed between the glass tubes. The results, which exhibited as fair 

 agreement with one another as could be expected, showed, as before, that the con- 

 ductivity for heat was greater towards the analogous pole so long as the temperature 

 of the crystal was rising. 



In respect of the electrical conductivity, time had only permitted a few preli- 

 minary experiments. The slice of crystal was heated in a steam bath. A five 

 microfarad condenser was charged through the crystal for one minute with 10 or 

 12 Daniell's cells, and the condenser was then discharged through a sensitive 

 Thomson galvanometer of 7000 ohms resistance. The limit of the very slight swing 

 was accurately observed, and then the operation was repeated with the tourmaline 

 electrically reversed. This was repeated alternately. When the temperature was 

 rising a difference between the two swings was perceived ; also when the tempera- 

 ture was falling there was a difference in the other direction. But these must have 

 been chiefly due to the electromotive force, so-called, of the crystal. When the 

 temperature was steady not the slightest difference could be perceived. The authors 

 would wish, before being satisfied with this result, to heat the tourmaline to higher 

 temperatures, and to try a much higher electromotive force — say that of 1000 cells. 



10. On Gaussin's Warning regarding the Sluggishness of Ships' Magnetism. 

 By Sir William Thomson, F.B.S. 



(Practical Bule and Caution.) 

 1. After steering for some time on westerly courses, expect 1 («) westerly 

 error if you turn to the north, 1 (b) or easterly error if you turn 

 to the south. 



2. After steering for some time on easterly courses, expect 2 (a) 

 easterly error if you turn to the north, or 2 (6) westerly error if 

 you turn to the south. 



The diagram representing £ase 1 (a) illustrates the physical 

 explanation: N. and S. representing the north and south points of 

 the compas3 card (or " true south" and " true north " poles of its 

 S needles), and the small letters s, s, s, true southern polarity, and 

 n, n, n, true northern polarity, induced in the port and starboard 

 ends of deck beams, and port and starboard sides of ship, while 

 s steering east, and remaining for some time after she has been 

 turned to north. 



In the Admiralty ' Compass Manual,' Gaussin's warning is given 

 with reference to the direction of swinging, in correcting the com- 

 pass by magnets, according to Airy's first method. In the Reports of the Liverpool 

 Compass Committee, and in Mr. Towson's 'Information for Masters and Mates 

 regarding Ships' Magnetism,' instances of perplexing changes in the compass are 

 given, and are referred to the same cause. The " sluggishness" of ships' magnetism, 



