Magnetism of Basalt. 517 



the Sgurr nan Gillean and Storr specimens have a marked 

 resemblance to the temperature curve found for magnetite 

 by Barton and Williams *. Curves are given for all the bars 

 except Dunvegan, which is similar to that found for the 

 Storr bar. In character they appear to be modifications or 

 combinations of two types, one the curve for magnetite f, to 

 which that for the Stromness bar has the greatest similarity, 

 and second, the curves IX., X., and Xll.f for Dattenberg 

 bars. In the one type there is a gradual increase of suscep- 

 tibility up to about 500° C, followed by a rapid descent to a 

 state of minimum susceptibility ; in the second type the 

 maximum susceptibility is reached at a comparatively low 

 temperature, and this is followed by a gradual loss of power 

 as the temperature rises. Whether the change exhibited in 

 these curves between the temperatures of 15° and 550° is 

 due entirely to the appearance of a new magnetic constituent 

 as a result of chemical change, or to the effect of change of 

 internal stress, is a point which cannot be decided at present. 

 The temperature of minimum susceptibility had values for 

 the five bars ranging from 430° for the Dunvegan bar to 

 710° for the Stromness bar, the average being about 565° C. 

 The temperature curve for the Faule Ader bar may be noted 

 as one in which a large increase of susceptibility took place 

 during; heating. 



To confirm the effect found in this bar an irregular strip 

 of rock from the same block was roughly ground to pass 

 into the furnace, and was heated to 350° C. or thereby, 

 and again cooled. Initially the bar had no observable 

 magnetization, and gave a deflexion of 2 mm. in a reversed 

 field of 35 c.G.s. units. On being cooled, the bar now 

 gave a deflexion of 6'5 mm. in the same field as before, and 

 was now magnetized so as to be capable of affecting a pocket 

 compass. Further heatings appeared to increase the effect 

 slightly. 



After the heating of the bars, they were tested under the 

 same conditions as those given in Table I., and the following 

 deflexions were noted (p. 578) : — 



* B. A. Report, 1892. 

 f See previous paper. 



Phil. Mag. S. 6. Vol. 17. No. 100. April 1909. 2 R 



