54 



NATURE 



^March 21, 19 1 8 



that the difference is of serious importance. The 

 cosine should be squared and the moment of inertia of 

 the balance should be replaced by the time of its swing ! 

 A little thought will show that K must come in equally 

 on both sides of 'the equation and so be eliminated. It 

 is somewhat surprising to find T in the numerator, for 

 this would seem to indicate that if the balance did not 

 turn at all there would be^as measured by 

 its tangent — ^an infinite deflection — i.e. 90°. Of 

 course, the real meaning is that while the deflecting 

 couple becomes less as 1" is greater, the sensibility be- 

 comes greater in the proportion of the square of the 

 time, and (the deflection goes on getting greater with 

 increasing slowness of rotation until the whole thing 

 becomes unmanageable on account of its too great 

 delicacy, or until the decrement, by its consequent 

 increase, more than compensates for the diminished 

 stability. It is not clear what numerical results, if any, 

 were obtained by Eotvos. By the formula now given, 

 taking T as 60, K as 300 or thereabouts, and (/> as 45°, 

 the amplitude should only come out about one-seventh 

 of the amount that the published formula would re- 

 quire. 



It may be worth while to point out that the centri- 

 fugal force of the balance about its vertical axis, if the 

 beam is 20 cm. long, and turns once a minute, is about 

 720 times as great as the alteration of weight at the 

 equator, so that if the. beam were exactly in neutral 

 equilibrium when stationary and pointing east and west 

 it would have, in virtue of its rotation, a stability given 

 to it under which the change in weight could not pro- 

 duce a steady deflection exceeding about 1/12°. No in- 

 formation is given as to how fe was determined, nor 

 is centrifugal stability mentioned. .As in any system 

 the logarithmic decrement becomes less as the stability 

 is greater, it would be useless to determine fe with any 

 but the correct stability. The only method apparent 

 to the present writer would be the addition of a 

 stability bob equal in effect to the calculated centrifugal 

 stability and a determination with the rotation stopped. 



No mention is made of the most interesting feature 

 In the scheme of the experiment. If the balance is in 

 perfectly neutral equilibrium when not rotating, then 

 the centrifugal stability is the only stability, and perfect 

 synchronism is obtained whatever be the speed of rota- 

 tion, whereas if there had been any initial stability or 

 instability it could never be attained at any speed. 



If the direction of rotation is such as to make the 

 north end heavier than the south end, then with very 

 small damping this end should be in nearly its highest, 

 jiot in its lowest, position, as might at first be expected, 

 at each turn. 



This experiment, which, like those with the gyro- 

 static compass, ■ and unlike Foucault's pendulum ex- 

 periment, is best done in the tropics, is one of such 

 interest and beauty that it is to be hoped, even in these 

 difficult times, it may be set up and exhibited in some 

 physical laboratory. 



It is unfortunate that the author has not done justice 

 to Eotvos, but he has prepared somewhat of a tangle 

 which it has been a pleasure to unravel. 



C. V. Boys. 



RESULTS OF VOLCANO STUDY IN 

 HAW AIL 

 'X'HE Hawaiian Observatory was founded in 1912 

 -»■ by the Massachusetts Institute of Technology, 

 and financed in large measure by business men in 

 Hawaii. Its publications have been systematic volcano- 

 logic and seismometric bulletins, and two larger re- 

 ports, as well as numerous special articles. The scien- 

 tific work has been done by Mr. T. A. Jaggar, director 

 of the station, and Mr. H. O. Wood, associate. Pre- 

 NO. 2525, VOL. lOl] 



liminary announcement of results ^ at the end of the 

 first five years of work reveals discoveries which may 

 be of interest to science at large, and some of these 

 discoveries are briefly reviewed here. 



Nature of Hawaiian Gases and Flames. 



The gas collected from a blowing-cone in the lava 

 pit of Kilauea in 1912 by Day and Shepherd =* con- 

 tained dominantly sulphur dioxide, carbon dioxide, and 

 nitrogen, subordinate amounts of the combustible 

 gases, sulphur, carbon monoxide, and hydrogen, and 

 only 4 per cent, of water vapour. The 79 per cent, 

 of SO2, CO2, and H2O could not, to the writer's 

 thinking, be juvenile, but must in part result from 

 union with -atmospheric oxygen. Day had suggested 

 that heat-producing reactions betw^een such gases as 

 free S, CO,, and H, rising through the lava, would 

 raise the surface temperatures so that the lava column 

 might be at its hottest above instead of in the depths. 

 Continuous recording and observation of flames, with 

 experimental measurements of temperature and sound- 

 ings of the lava for viscosity differences, show that 

 this generalisation is well founded, and, in addition, 

 that atmospheric oxygen is brought in contact with 

 the magmatic gas so as to produce abundant flames of 

 different colours. Air is sucked down at the convec- 

 tional whirlpools and cascades. It is carried downward 

 in the liquid lava lakes by foundering of porous crusts 

 which cannot melt in the superfused lava glass. Air 

 is also carried down in broken wall rock, in avalanches, 

 and by burial of old talus. Lastly, with 33 per cent, 

 volume shrinkage due to such gas reaction within the 

 lava column as 2H2+02 = 2H,0, even at high tem- 

 peratures (1100° C. more or less), and with convec- 

 tional gas pumping, a Bessemer furnace effect through 

 the liquid lava may be created by indraught of air 

 from the walls. 



Of the three combustible gases H, CO, and S, 

 sulphur is most in evidence as surface flames, carbon 

 monoxide along with impurities may be represented 

 by rare flames, while hydrogen probably flashes mostly 

 to water-vapour in depth. There are whitish flames 

 occasionally seen, and intensely hot bluish to violet 

 flames play at all times from the glowing grottoes and 

 chimneys. Some work has been done in an effort to 

 photograph the flames with colour filters and pan- 

 chromatic plates, and there is a promising field here 

 for the study of flame spectra. 



Nature of a Lava Column. 



While it was known many years ago that some of 

 the Hawaiian lava pools were shallow, few observers 

 have imagined that the liquid lava rising 600 ft. 

 during a year within a pit much deeper than that 

 would be found by sounding at the end of the period 

 to be only 45 ft. deep, though still fully liquid at the 

 surface. This was the case at Halemaumau, the 

 inner lava pit of Kilauea, in January, 1917 (Fig. 1). 

 Sounding was accomplished by dunging a steel pipe 

 into the lava lake at several different locations, and 



1 " The Outbreak of Mauna Loa, Hawaii, 1914," by T. A. ]a.gsar, Amer. 

 /own. Sci., vol. xxxix., February, TO15, pp. 167-72. "Activity of Mauna 



• oa, December, 1914-January, 1915," by T. A. Jaggar, Amer Journ. Set., 

 vol. xl., December. 1915, pp. 621-39. " Lava Flow from Mauna Loa, 1916," 

 by T. A. Jaggar, Amer. Journ. Sci.. vol. xliii., April, 1917. pp. 25S-88. 

 " Seismic Prelude to the 1914 Eruption of Mauna Loa," by H. O. Wood, 

 Bull. Seie. Soc. America, vol. v.. No. i, Marcn, 1915, pp. 39-50. " Notes on 

 the 1916 Eruption of Mauna Loa," by H. O. Wood. Journ. ofGeol., vol. xxv., 

 Nos. 4 and 5, 1917, pp 322-36 and 467-S8. " Volcanologic Investigations 

 at Kilauea," bv T. A. laggar. Amer. Journ Set., vol. x'iv., September, 1917, 

 pp. '61-220. "Li<e Aa Lava at Kilauea," by T. A. Jaggar, Journ. Wash. 

 Acad. Sci., vol vii., No. 9, May 4. 1917, pp. 241-43. "On the Terms 

 Aphrolith and Dermolith," by T. A. Jaggar, Journ. Wash. Acad. Sci.. vol. 

 vii., No 10, May ip, 1917, pp. 277-81. "Thermal Gradient of Kilauea 

 Lava Lake," by T. A. Jaggar, Journ. Wash. Acad. .Sci., vol. vii., No. 13, 

 July 19, 1917, pp. 397-405. "On Cyclical Variations in Eruption at 

 Kilauea," by H. O. Wood, Second Report Hawaiian Vol. Obs. (Cambridge, 

 Mass., 1917). 



2 "Water and Volcanic Activity," by A. L. Day and E. S. Shepherd, 

 Bull. Geol. Soc. Amer., vol. xxiv., 1913, pp. 573-606. 



