September 30, 



1897] 



NATURE 



533 



linually increased. Finally, quartz appears like an insulator in the 

 same r6le as water in ordinary aqueous solutions. In all these 

 cases I wish to keep in mind the results of Alexeeff and their 

 recent repetition for metallic alloys, together with the interpre- 

 tation of these results due to Masson. In a crust subject to 

 variable magnetism, traversed by earth currents, sustained by 

 semi-metallic carbides of the Mendeleefif-Moisson type, contain- 

 ing piezo-electric and thermo-electric sources, who can say that 

 electric fields are absent ? Again, the character of the changes 

 contemplated in (iibbs'^ famous " phase rule," as interpreted by 

 Le Chatelier, would here be ionic rather than molecular. 



A question of somewhat allied interest is the action of hot 

 water under pressure on rock-forming silicates. Investigations 

 of this kind have been described in the well-known and fascina- 

 ting book of Daubree. Daubree's work, however, is qualitative 

 in character, like that of many others in the same line, and the 

 furtherance of the subject is to be looked for in the quantitative 

 direction. Some time ago, Becker suggested experiments on 

 a huge mass of granulated rock under the action of steam at 

 exceptionally constant temperature. But no thermal effect 

 of the action of water could be detected. True, the boiling 

 point of water is a temperature relatively low for the purpose ; 

 yet similar experiments made with liquid water at over 200' 

 under pressure were equally negative as to results. Experi- 

 ments of this kind are not very conclusive. The insufficient 

 sensitiveness of the measuring apparatus, the rate at which 

 heat is carried off compared with the rate of generation, and 

 other obscure causes mar the results. The question may, how- 

 ever, be approached in a somewhat different way : if water 

 is heated under pressure in glass tubes, the volume of water 

 contained decreases as the square, whereas the chemically 

 active area, i.e. the inside surface of the tube, decreases as 

 the first power of the diameter. Hence, in proportion as the 

 tube is more capillary, the action of water on the glass will 

 produce accentuated volume eTRTects. Thus it was shown that 

 the behaviour of hot water is profoundly modified by its con- 

 tinued action on glass, inasmuch as its compressibility increases 

 at a very rapid rate with the lime of action even at 180°, until, 

 with the approach of solidification, the observed compressibility 

 is fully three times its isothermal value at the inception of the 

 experiment. Even more striking is thesimultaneousand continual 

 decrease of the length of the column of water. Clearly, therefore, 

 the confined volumes of glass and included water must undergo 

 contraction at 180° in forming an eventually solid aqueous 

 silicate, while increasing compressibility is due to the increasing 

 quantity of silicate dissolved. Now, in nearly all cases the effect 

 of solution is a decrease of compressibility. Hence the increased 

 compressibility observed is to be referred to a precipitation of 

 the dissolved silicate, in response to the action of pressure, a 

 result borne out by the appearance of the tube and by varied 

 correlative experiments. It is, however, the volume contraction 

 which is particularly interesting, because of its far-reaching 

 geological application. In the first place, the measurements 

 show that about '025 cubic cm. of liquid water is absorbed per 

 square centimetre of glass surface at 180° C. per hour.i The 

 effect of this absorption is a contraction of bulk amounting to 

 18 per cent, per hour. So large and rapid a contraction 

 is presumably accompanied by the evolution of heat. Hence, 

 under conditions given within the first five miles of the earth's 

 crust, i.e. if water at a temperature above 200° and under 

 sufficient pressure to keep it liquid be so circumstanced that the 

 heat produced cannot easily escape, the arrangement in question 

 is virtually a furnace whose efficiency accelerates with rise of 

 temperature or increase of terrestrial depth. 



PlEZOMETRY. 



It is not feasible to make much progress in pyrometry with- 

 out feeling the need of a corresponding development in high 

 pressure measurement. This has already appeared in the pre- 

 ceding jiarts of my address. It vvill not be expedient to look 

 into the history of the subject so comprehensively as I did in 

 the case of pyrometry, partly because the literature is more 

 diffuse, and partly because the real development of piezometry 

 is of recent date and virtually begins with pressures of the order 

 of several thousand atmospheres. So understood, although we 

 gladly pay homage to Oersted, to Regnault, to Grassi and many 

 others, our historical summary may be abridged. 



As is often the case in physics, the great advances in the 



.1 This is an initial rate of about 180 kilograms per s juare metre per year. 



subject are associated with the name of 'one man ; for though 

 many able investigators have contributed effectively to the 

 progress of piezometry, the overshadowing importance of the 

 results of Amagat have superseded all researches coextensive 

 with his own. For over twenty years Amagat has been labour- 

 ing on this definitely circumscribed subject. Year after year 

 his prolific experimental ingenuity has put forth results, each of 

 I which in its turn constituted the highest attainment in accuracy 

 : and the greatest breadth of scope which high-pressure measure- 

 ment had reached at the lime. It is impossible to give any 

 i adequate view of this sustained labour in an address. The 

 subject is highly specialised and demands special treatment ; 

 but we owe to Amagat the bulk of our knowledge of the 

 \ properties of a gas regarded not as an ideal fluid, but as a 

 j physical body ; some of the most far-reaching results in the 

 I thermodynamics of liquids and some of the best data in the 

 ; elastics of solids. 



! Amagat investigated gases within an interval of pressure 

 1 which at times reached 4000 atmospheres, with a view to 

 interpreting their divergence from the laws of ideal gaseity. 

 I Indeed we may note in passing that, just as the advanced 

 \ astronomy of the day is being enriched with unexpected dis- 

 coveries from a discussion of mere errors of observation, so 

 refined physical measurement gleans new harvests in carefully 

 tracing out the all but rigorous sufficiency of established laws. 

 The product of pressure and volume, nearly constant in the 

 ordinary isothermal behaviour of gas, shows, under higher 

 pressures, a well-marked passage through a minimum in the 

 case of all gases except hydrogen. Hence below a certain 

 definite pressure, varying with the character of the body (say 

 40 aim.), gases are more compressible than Boyle's law asserts, 

 and above this pressure they are continually less compressible 

 and begin to resemble hydrogen in this respect. The sharp- 

 ness of the minimum diminishes as temperature increases and^ 

 probably ultimately vanishes. Cailletet, it is true, had 

 undertaken a study of the same subject simultaneously, but 

 his results were not in the same degree correct. Again, the 

 coefficient of expansion of gases considered in its isopiestic 

 j behaviour for temperatures not too far above the critical point, 

 1 increases with pressure to a maximum, which seems to occur at 

 the same pressure for which the volume-pressure product is a 

 I minimum. This thermal maximum also decreases with lem- 

 \ perature and finally vanishes. To specify the conditions 

 further than this would be to exceed the limits beyond which 

 verbal statement ceases to be lucid. The value of Amagat's 

 , work, however, is not merely the formulation of such general 

 ! laws for gases as a whole, but rather the investigation of sharp 

 I and specific results for each gas individually. Thus if one uses 

 these data for a given gas to compute the constants in Van der 

 Waal's law, one is actually able to predict remote critical 

 conditions of the gas in question with a fair degree of accuracy. 

 Whenever pressure measurements are to be made through 

 such large intervals as are here in question, the elastic 

 constants of the apparatus become of increasing moment. 

 Amagat, however, treated these incidental measurements as- 

 of like importance with the rest of his labours. The starting 

 point of his investigation into high pressures was the open^. 

 mercury manometer first erected along a staircase near Lyons,, 

 finally in the shaft of the St. Etienne Mine, about 380 metres- 

 deep. This apparatus was used for graduating the closed 

 manometer, preferably containing nitrogen. In later experi- 

 ments for excessively high pressures, the closed manometer was 

 replaced by the " manometre i pistons libres," a sort of 

 inverted Braniah press, in which the small pressures of an open 

 mercury manometer acting on a large pi.ston compensate the 

 relatively large pressures of the piezometer acting on a small 

 piston. The ingenious feature of Amagat's apparatus is the 

 rotation of both pistons just before measurement, a device by 

 which friction is rendered harmless. Equipped with this instru- 

 ment, direct determination of the bulk modulus for glass and 

 metals was actually feasible. In the case of glass no serious 

 variation of the compressibility could be ascertained within 2000 

 atmospheres and even 200°, an observation of great value in 

 practical research. Poisson's ratio was similarly determined, and 

 the data used in computing Young's modulus. But the most 

 important result of these researches, a result to which Prof. 

 Tait also contributed, is the datum found for the ab-solute com- 

 pressibility of mercury. This will enable all future observers in 

 piezometry to standardise their apparatus with ease and nicety. 

 Time prevents me from dwelling ujwn the remaining invesli- 



NO. 1457, VOL. 56] 



