salt passing up inside. The tubes must be very narrow in 

 the case of such sails as cadinic sulphate, which gives 

 filaments almost as fine as hair. These wave about in the 

 solution but do not thicken or break. Very fine detached 

 filaments also grow very fast. While one may offer sug- 

 gestions as to the possible agency producing some of the 

 growths, I know of no explanation to account for the very 

 different forms assumed by the various salts, and especially 

 why salts of metals generally supposed to be related 

 chemically, behave quite differently. Thus cadmium gives 

 long slender filaments, zinc soft thick shoots. Cobalt grows 

 freely in long thivads; nickel gives fungoid or short thick 

 shoots; ferrous salts grow very freely; ferric in thick 

 fungoid forms. 



It occurred to me that possibly the growths might push 

 themselves between the grains of stone. One might thus 

 obtain food for speculation as to whether something of the 

 kind might not take place in nature. I accordingly placed 

 some silicate solution, together with a crystal, on top of a 

 small cube of sandstone. The crystal swelled but gave 

 little satisfaction, T then placed the cube in solution just 

 deep enough to cover it, with crystals beside it. Copper, 

 iron and cobalt sent out shoots which became firmly adher- 

 ent to the stone but did not appear to penetrate. I then 

 tried a specimen of shale in silicate solution. Copper again 

 grew, but did not penetrate between the laminae. I hope 

 to pursue the subject further. A method which is said to 

 be in use for preserving wood, viz., to saturate the wood 

 with silicate of soda and then place in a solution of iron, 

 which forms a coating of silicate of iron, might be tried, 

 but it is doubtful if true shoots would form. 



I am conscious that the results I have brought forward 

 are in many ways incomplete, and that much more remains 

 to be done in connection with the subject. I have only 



