ox THE GEOLOGY OE EAEBADOS. 183 



portion (A) ceased to attack the sample it was removed by elutria- 

 tion and filtration, the insoluble residue heated in a water-bath for 

 several hours with a strong solution of sodium carbonate to which 

 some caustic soda had been added, and washed, with the assistance 

 of a Bunsen's pump, (1) with the same solution, (2) with water, 

 and (3) with dilute hydrochloric acid. The residues obtained were 

 dried, ignited, and weighed, and consisted of the clay and of other 

 siliceous minerals. After weighing, this residue was reduced to an 

 impalpable powder in an agate mortar, and a weighed portion boiled 

 repeatedly with strong sulphuric acid until the clay was completely 

 decomposed. The excess of acid was then driven off and the residue 

 heated with dilute hydrochloric acid to dissolve the basic sulphates 

 of alumina and iron formed, and after removal of the dissolved 

 portions it was heated with a solution of sodium carbonate to dis- 

 solve the silica set free. The final residue, after being washed and 

 ignited, consisted only of crystalline siliceous minerals, and in the 

 majority of cases only of quartz. 



The solution obtained from the second portion (B) was evaporated 

 to dryness and, as well as the undissolved matters, kept at a tem- 

 perature of from 120° to 130° C. for an hour or longer to render all 

 the silica insoluble in water. The residue and undissolved matter 

 were treated with warm hydrochloric acid, taken up with water, 

 filtered and washed. The residue from that was dried, ignited, and 

 weighed, and consisted of the colloid silica originally present, of a 

 small quantity of colloid silica arising from the decomposition of 

 certain hydrous silicates present in the rocks in small quantity, and 

 of the clay and other siliceous minerals. The difference between the 

 proportions of this and of the clay and siliceous minerals as deter- 

 mined in sample (A) gave the amount of colloid silica present. The 

 alumina, iron peroxide, &c., were determined in the hydrochloric 

 acid solution from (B) by the usual methods, and the carbonic anhy- 

 drides in another portion of the sample. At first, acetic acid was 

 used as the solvent ; but doubts arising as to its partially attacking 

 certain zeolitic minerals present, it was decided to use hydrochloric 

 acid and to estimate the small quantity of silica set free from these 

 minerals as colloid silica. As this was probably derived originally 

 from siliceous organisms the method may be considered fairly 

 accurate. "Where the colloid silica was principally present in an 

 amorphous form, as in the foraminiferal chalks, it entered into 

 solution with far greater ease and rapidity than when it was in the 

 form of remains of siliceous organisms, when it not unfrequently 

 offered considerable resistance to solution. Is this difference in the 

 solubility of the silica due to differences in its state of hydration or 

 to purely molecular causes ? 



The samples of argillaceous earths and of volcanic sands and mud- 

 stones were examined by the methods of analysis usually employed 

 for clays and for siliceous minerals, the soluble silica present being 

 determined as described. In these samples the alkalies were sepa- 

 rately determined, whilst in the other samples, as they were deter- 

 mined by difference, the errors of analysis fall upon them. In the 

 Q. J. G. S. No. 190. o 



