except that given us by the experiments of Bischoff.1 Bischoff, experimenting on trachite, granite, and basalt, found the solids in these three cases about 10 per cent. denser than the liquids.

This is a subject on which we ought to have better information, and I do think that physicists are at fault for not having given information regarding it to geologists. Speaking, on the other hand, in the capacity in which you do me the honour to place me, we, the geologists, are at fault for not having demanded of the physicists experiments on this subject. This time last year I urged that some one should take up the subject. I now renew the suggestion. We have a great Government fund for scientific inquiry. Why do not physicists and geologists unite in an inquiry like this, and apply for assistance out of the Government grant fund? They should then repeat Bischoff's experiments and

1 Since this Address was delivered, some important experiments have been carried out at the request of Dr. Henry Muirhead, by Mr. Joseph Whitley, of Leeds. His experiments were made on iron, copper, and brass, and on whinstone and granite, and the general result, seemingly (but I believe not at all surely), indicated is that these substances are less dense in the solid than in the liquid state at the melting temperature.

test his results, but on a far larger scale, and extend the investigation to many materials besides those which he used.

In the meantime, the only experimental evidence we have, being Bischoff's, is to the effect that the melted rock is of less specific gravity than the solid rock, and that the solid rock would therefore sink down when frozen. Without insisting on this as being proved, I shall just call your attention to what would be the probable history of the solidification of a molten globe if this were the case. As soon as the surface began to freeze, and to freeze in sufficient quantity not to be floated up by mere superficial solidified foam, the mass of rock would fall down towards the centre. More would then solidify at the surface. This also would fall down, and the same thing would go on again and again. Gradually a sort of honeycombed solid would be formed. By and by a skeleton or frame-work through the whole would mount up to an extent sufficient to build up piers, as it were, to the surface, and the spaces between these piers when close enough would, in the con

VOL. II.

X

tinued freezing of the lava, be bridged across by solid rock thick enough in proportion to breadth not to break down and sink. There would again be breaking away of the piers and upheavals of the liquid material below; but by degrees the honeycombed mass would become nearly like a solid throughout with comparatively small interstices of liquid lava. This is in part the view of Hopkins. On account of arguments drawn from the phenomena of precession and nutation he was led to regard the earth as rigid on the whole; and looking to the existing evidence as to geological phenomena he was led to hold a view of the probable history of the earth agreeing with that which I have sketched, at all events so far as it would lead to the supposed present honeycombed

condition.

Hopkins considered other very important questions, and among these was the question of the influence on their condition as to solidity and fluidity, of the great pressure to which the interior parts of the globe are subjected. He made experiments as to the influence of pressure on the

melting points of some fusible substances, though he was not able to extend them so far as to test the effect of pressure in altering the melting point of rock materials.

Without making direct experiments, however, as to the influence of pressure on the melting point of rocks, we are able to infer by thermo-dynamic reasoning what must be the effect of pressure on the temperature of fusion, if we know whether rocks. expand or contract in the act of solidifying. My brother, Professor James Thomson, has pointed out that it is a necessary consequence of thermo-dynamic laws that the temperature of the melting-points of substances which expand in solidifying should be raised by the application of pressure, while the opposite would be the case with substances which contract in becoming solid. Water, for instance, expands when it freezes: and he calculated the amount by which the freezing-point of water ought to be lowered by the application of an additional atmosphere of pressure. He found it to be about of a degree Fahrenheit; and his theoretical result was afterwards verified by experi

ments made in the incipient Physical Laboratory in the old buildings of the University of Glasgow. Now, if the result announced by Bischoff, that the rocks he examined contract during solidification, be correct, it would follow as a consequence of thermodynamic laws that these substances melt at a higher temperature under high pressure than that at which they melt under low pressure. The conclusion to be drawn respecting the internal condition of the earth is, that we are not to infer liquidity of the interior, even if we should find evidence of a much higher internal temperature than that which would melt the rocks under ordinary pressure.

The view which I wish to put before you just now, however, is of a very different nature. I mean to deny altogether the intensely high temperature which Hopkins accepted, and the liquefying effect of which he endeavoured to obviate by introducing considerations regarding the solidity of the rocks, in consequence of intense pressure even at very high temperature. It is too generally supposed that the rate of increase of underground