[106] It would appear that Davy latterly preferred zinc to iron, as the protecting metal. In a letter, dated October 1826, addressed to a ship-owner, who had made some enquiries of him upon the subject, he says--"The rust of iron, if a ship is becalmed, seems to promote the adhesion of weeds; I should therefore always prefer pieces of zinc, which may be very much smaller, and which, in the cases I have heard of their being used, have had the best effect." The bottom of the Glasgow was examined twelve months afterwards, when the discs of iron were found oxidated throughout, presenting in their appearance the characters of plumbago. The copper on the starboard side was preserved, but covered with weeds and shell-fish. The sheets on the larboard had undergone the usual waste, but were clean. The Zebra was docked four years after the experiment had commenced, when the zinc protectors were perfect, and it did not appear that they had exerted any influence in preserving the copper, as it had wasted equally on both sides. It may be presumed in this case that the Voltaic circuit had by some fault in the arrangement been interrupted. The apparent conversion of iron into a substance resembling plumbago, by the action of sea-water, has been frequently noticed. The protectors thus changed[107] were, to a considerable depth from the surface, so soft as to be easily cut by a knife; but after being exposed for some time to the action of the atmosphere, they became harder, and even brittle. A portion of this soft substance having been wrapped in paper for the purpose of examination, and placed in the pocket of a shipwright, gave rise to a very curious and unexpected result: at first, the artist, like Futitorious with his chestnuts, thought he perceived a genial warmth; but the effect was shortly less equivocal; the substance became hot, and presently passed into a state of absolute ignition. Various theories have been suggested for its explanation: Mr. Daniell has advanced an opinion which supposes the formation of silicon, and thus accounts for the spontaneous ignition by the action of air.