ON A NEW SELF-RECORDING THERMOMETER. 335 



3. Desert Sandstone, Adelaide River, 2*59. 



4. Gorge of the Katherine, 2-54. 



5. Coarse friable sandstone from head of Mary River, North 

 Australia, and under 40 feet of magnesite, 2*57. 



6. Desert Sandstone, McMinn's Bluff, 2 -51. 



7. Desert Sandstone, McMinn's Bluff, 2-36 - 2-37. 



8. Limestone, Tableland, Katherine River, 2*69. 



The specific gravity of the magnesite was about 2*90, but at 

 present I have no specimen to confirm this. 



Explanation of Plates. 



Plate xix., fig. 1. — Prismatic basalt, Glasshouse Mountain, 

 Moreton Bay, Queensland. 



Plate xx., fig. 2. — Core of prismatic basalt, Glasshouse Mountain, 

 Queensland. 



Plate xxi., fig. 3. — Various specimens of volcanic dust. 



Plate xxii., fig. 4. — Desert Sandstone, Mary River, North 

 Australia, x 70 diam. 



Plate xxii., fig. 5. — Ash from Bromo crater, Java, (active 

 volcano) x 70 diam. 



Plate xxii., fig. 6. — Sahara Desert Sand x 300 diam. 



Plate xxiii., fig. 7. — Hyalite from lava, Mount Bramble, 

 Springsure, Queensland. 



Plate xxiii., fig. 8. — Desert Sandstone, Yam Creek, North 

 Australia, x 50 diam. 



ON A NEW SELF-RECORDING THERMOMETER. 

 By H. C. Russell, B.A., F.R.S., &c. 



[One Diagram.] 



[Read before the Royal Society, N.S.W., 7 November, 1888.] 



I need not remind the members of the Royal Society that the 

 attempts to make a really satisfactory recording thermometer 

 have been very numerous, and that it is probable that a complete 

 solution of the difficulty in recording changes of temperature 

 accurately has yet to be discovered ; but I think I shall be able 

 to shew you that the one I have to describe this evening is in 

 many respects an advance upon those which have gone before it. 



