Deposits in Meiastable Inorganic Hydrosols. 



33 



it has not before been examined microscopically at various periods during 

 this metastable existence. 



The actual setting into a thick jelly does not appear to yield microscopic 

 structures, probably because this change at its onset is rapid, and so yields 

 only ultramicroscopic structures. But the jelly, at first clear, gradually 

 shows an increasing opacity after setting, and then, if broken up and 

 examined under the microscope, shows flat membranous scales or plates like 

 those described above for the more coarsely precipitated colloidal ferric- 

 hydrate. 



The most interesting appearances, however, are those found before jelly 

 formation in the perfectly mobile colloidal solution, when the solution is so 

 dilute (1-2 per cent, of Si02) that it remains fluid for some weeks. 



Since these growths form slowly, as great care as possible was taken, by 

 autoclaving the mother-solutions of sodium silicate and hydrochloric acid 

 and the dialysing apparatus and all glass apparatus and containers, to prevent 

 infection adventitiously from without. 



The growths obtained with colloidal silicic acid are illustrated by the 

 microphotographs of figs. 3 and 4. No special care was taken in auto- 

 claving the mother-solutions in the experiment shown in fig. 3, but it 

 is to be remembered that these are strong hydrochloric acid and strong 

 sodium silicate, neither of them suitable media for the growth of organic 

 germs. 



After mixing and dialysing, the colloidal silicic acid was run off into a 

 stoppered bottle which had been sterilised by blowing live steam through 

 it. The growths shown were observed after standing for about three weeks 

 at laboratory temperature. The solution contained just over 2 per cent, of 

 SiC>2, had been dialysed for 48 hours, and coagulated in about five weeks' 

 time. 



The coarser growths shown in fig. 3, A, B, c, and d, are obtained occasionally 

 when a stronger solution (about 3 per cent.) of colloidal silicic acid is left for 

 two or three days ringed round with gold size. They appear to form when 

 slow massive formation of a gel phase occurs in a fairly strong solution 

 accompanied possibly by slow concentration due to slight evaporation.* 

 These forms yield often very beautiful patterns with ramifying and branching- 

 fibres. 



The coarser fibres seen in the micro-photographs are not cracks in a 

 contracting jelly, but a more solid phase actually growing in a more fluid 

 phase ; this is well seen at the growing edge of such a tuft of fibres, as also 

 from fine structural details in the individual fibres. 



The appearances seen in fig. 3, e and F, and in all the photographs of 



VOL. LXXXLW — B. J) 



