Now, the hollow centre of 475,116 miles in diameter would have a volume of one-sixth of the whole volume of the sun, which, filled with gases, would diminish all these densities just in proportion to what may be considered the degree of compression and condensation the gases might be subjected to. That there should be gases in the interior hardly requires to be more than stated, as there can be no doubt that the degree of heat to which the shell had arrived by the time it came to have the dimensions above mentioned, would be amply sufficient to excite chemical action among the elements of which the sun is composed; and the gases or vapours produced by that action would flow as naturally towards the interior of the hollow centre as towards the space beyond the outer surface of the shell, until they were stopped by increase of pressure, which of course would mean increase of density in this case. We see then that if the hollow centre has a volume of one-sixth of the whole volume of the sun and we multiply this volume by 6, we have a mass equal to the whole mass of the sun, were its mean density only the same as that of water. Consequently, if we multiply the said volume by 6 and by 1·413, that is by 8·478, we get a mass equal to the whole mass of the sun at its known mean density. Again, were we to suppose the hollow centre to be filled with gases of the same specific gravity of air, condensed to a pressure of 6560 atmospheres--which would correspond in density to 8·478 times the density of water--we should have in the hollow centre alone a mass equal to another sun, in addition to the one made up by the dimensions and densities stated above. We see then that if we fill the hollow centre with gases at the pressure, and with the density just stated, we have a sun of twice the mass it should be. But if we leave the specified gases in the hollow with one-half of the above density, and deduct the equivalent mass of the other half density of the gases from the shell, as estimated for the hollow centre, we should have a sun of the mass required by astronomy. In this way we should have the three specified densities reduced from 0·471, 1·70 and 0·71 to 0·236, 0·85 and 0·355, for the outer surface, the region of greatest density, and the inner surface of the shell, respectively; and the pressure and density of the gases in the hollow centre reduced to 3280 atmospheres. Thus, from what has just been shown, which at first sight may be thought very irrelevant matter, we discover that it is not necessary that there should be any matter in the sun even so dense as water. And still we have to think of what an insignificant pressure three or four thousand atmospheres would be in the centre of the sun.