September 28, 1906.] 



SCIENCE. 



401 



to determine the tangential stress at F in- 

 volved. It follows that F = 10^^ dynes/cm." 

 or 10^ kg./cm.'; or since in a shear the tan- 

 gential and the normal stresses are equal per 

 square centimeter, about 100,000 atmospheres, 

 even for the excessive strain in question. 

 Now in a region where differential stresses of 

 this value abound, the pressure itself must be 

 at least of the same order, and hence if we 

 compute such pressures hydrostatically (a 

 case most favorable to shallowness of the seat 

 of reaction) with ten feet of rock to the at- 

 mosphere, this would be equivalent to a depth 

 below the surface of one million feet or 190 

 miles, where a shear of the value of one half 

 is surely out of the question. Imagine the 

 earth radii all flexed by this amount at their 

 outer ends. Besides we are much too far 

 down for practical vulcanism. Of course, we 

 can get nearer the surface with bigger strains 

 and smaller stresses, or we may imagine the 

 energy of a fault all spilled upon the surface 

 of rupture; but even in this case while the 

 work done will depend on the volume dis- 

 placed, it will also in a large measure be dissi- 

 pated within that volume and by no means on 

 the surface of separation alone. 



The picture as a whole is not alluring be- 

 cause it is vague, to me at least, who am all 

 the while fondling my own little notion. In 

 fact I once came as near being a physical 

 geologist as Elihu Thomson, though nobody 

 seems to have found it out. Yet in the days 

 when I still deluded myself into thinking such 

 things interesting I happened upon an as- 

 tonishing result in the endeavor to dissolve 

 hot glass in hot water.^ It did in fact dis- 

 solve to an eventually solid substance, which 

 for hardness and optical character was not 

 distinguishable superficially from the igneous 

 glass; and it did this completely (in water, 

 not in steam) at a temperature certainly much 

 helow 200° C. and in such a way (this is the 

 point) that the system of igneous glass and 

 liquid water contracted on combination as 

 much as 20 or 30 per cent, of the initial total 



-American Journ, XL VI., p. 110, 1891; VII., p. 

 1, 1899; IX., p. 161, 1900. Phil. Mag., XL VII., 

 p. 104, 461, 1899. 



volume. Think of this; the contraction of 

 concentrated sulphuric acid upon admixture 

 with water is but 2 to 3 per cent., and even 

 granting that molecular changes and not vol- 

 ume contractions are the truly important fea- 

 tures, the case for water-glass can not be so 

 easily dismissed. Whoever has tried to coerce 

 a solid-liquid system (as I did) knows that he 

 has a task on his hands; and whoever tries to 

 diminish bulk 20 or 30 per cent, (which I 

 didn't do) is destined to fall very short of his 

 hopes. I argued, therefore, that the solidifi- 

 cation of water in glass, under the exceptional 

 conditions stated could hardly take place with- 

 out the evolution of heat such as accompanies 

 any solidification of the liquid. All efforts to 

 prove the inference directly miscarried; but 

 in case of a reaction which proceeds very 

 gradually, in small compass and under high 

 pressure, failure is almost a foregone conclu- 

 sion. Nature, as Elihu Thomson truly states, 

 in her ideal laboratory can garner the heat of 

 a slow process, while such heat slips irrecov- 

 erably through our fingers in the workshop. 

 At this point then my argument is based not 

 on direct but on circumstantial evidence, and 

 if I were the reader, and not the writer, I 

 would merely grant a fused glass at a tem- 

 perature presumably somewhat higher than 

 the melting point. 



To me the picture obtainable from the bear- 

 ing of these experiments on vulcanism is more 

 attractive than any other with which I happen 

 to be acquainted, even if I have to chaffer 

 uncomfortably for the excess of heat above 

 mere fusion which seems to be present, as if 

 my glass had gnawed its way convectively 

 into the higher temperatures of deeper iso- 

 therms. The idea^ here is important: it is 

 probable that the water in a magma at 200° 

 will diffuse into a magma at 300° (and higher 

 in turn), across the surface of contact. The 

 region of fusion is, therefore, essentially sinh- 

 ing in character in its avidity for magmas at 



•' Inferred because the rate of solution increases 

 rapidly as temperature rises. Moreover the higher 

 temperatures of deeper isotherms are being con- 

 tinually brought from lower to higher levels by 

 convection, as solution proceeds. 



