Deposited from Metastable Solutions of Inorganic Colloids. 21 



were carefully cleansed w ith chromic acid and alcohol, and were flamed and 

 examined with the microscope before using them. All pipettes and lifters 

 were also flamed. In the manufacture of the tubes the formation of the 

 rounded bottom of the tube must have heated the glass red-hot for a 

 distance of 3 or 4 cm. at least from the bottom, and the drawing out at 

 the top must similarly have heated all the remainder. We are accordingly 

 quite satisfied that the deposits or growths we are dealing with were 

 actually formed in the tube-contents after the tubes were sealed off and 

 left standing. 



Each tube after filling was hermetically sealed and then sterilised by 

 heating in a steam autoclave for 15 minutes at 110° C. The tubes have a 

 capacity of about 100 c.c. and were about one-third filled with the selected 

 mixtures of colloids. The filling was carried out by nipping off at the narrow 

 end, warming the tube, and dipping into the prepared mixture of colloids 

 contained in another sterilised tube (as in the process of filling a thermo- 

 meter bulb) ; the tube was then resealed and autoclaved as stated above. 



The solutions used were those described by Dr. Bastian as the "yellow" 

 solution and the " colourless " solution, and the same solutions were also 

 employed with the addition of two drops of o-per-cent. sodium carbonate 

 solution in 30 c.c. of total fluid. This small amount of sodium carbonate was 

 added in order to provide material for formation of organic carbon compounds 

 if there was any tendency towards such a growth of organic matter. In no 

 case, however, have we been able to assure ourselves that there was any 

 formation of organic carbon compounds. We feel certain that there was no 

 appreciable growth of such compounds, but would like to leave this question 

 open because it is so difficult of solution. 



At the expiration of nearly seven months some of the tubes were opened 

 and examined microscopically, when the growths shown by the microphoto- 

 graphs (Plate 1) were observed. 



Some of the growths observed closely resemble exceedingly fine vegetable 

 fibres, such as filaments of cotton fibre, but finer in structure. Some are 

 rounded like silk fibres, others show flat bands like cotton fibres. Other 

 growths show branched and sometimes transversely divided filaments like 

 hyphae of moulds. These inorganic growths, indeed, so closely resemble 

 vegetable fibres that we were assured by two competent histologists, to whom 

 we showed them, that they were fibres obtained by leaving cotton filaments 

 on the slides in the process of preparation, and that the use of the reagents 

 for cellulose would certainly demonstrate this fact. 



This suggestion was valuable because it showed quite clearly, on applica- 

 tion, that the growths were not organic, or, at least, were not cellulose ; for 



