166 



APPENDIX. 



A quantity of the substance was prepared for analysis by elutriating away from the 

 heavier admixtures, washing several times with distilled water, collecting on a iilter and 

 drying. Treatment with dilute acid is not only unnecessary, since calcium carbonate is 

 absent, but harmful, inasmuch as the substance is very sensitive to even dilute acids. It is 

 a hydrated silicate, and has the typically zeolitic property of assuming whatever degree of 

 hydration brings it into equilibrium with its environment. Hence constant results cannot 

 be obtained by drying it either in air or in the oven. Roughly, there are about 18.6 per 

 cent of moisture in the air-dried material, and about 7.6 per cent after drying at 110° C. 

 Heated to 300° C. it is not completely dehydrated, and reabsorbs moisture on exposure. 

 At red heat it turns buff-colored and ceases to be hygroscopic. Even then, however, it 

 still retains 1 per cent or so of moisture, the last traces of which are only driven off by 

 heating to incipient white heat, when the substance sinters to a light brown, opaque glass. 



The analysis, reduced to terms of anhydrous material, runs as follows : — 



Since all deep-sea deposits formed in situ retain a certain ballast of undecomposed rock- 

 silicates, these were separated by treating the material with boiling sulphuric acid, fol- 

 lowed by dilute alkali, when a residue of 7.1 per cent of rock-silicates was found. This 

 contained 69.3 per cent of SiOj and 19.7 per cent of AljOj, so that it cannot affect the 

 bulk analysis seriously. 



The analysis is chiefly remarkable for the high content of alkalies, such as is unknown 

 in any secondary deep-sea silicate except palagonite, glauconite, and phillipsite. Alto- 

 gether the substance approximates to phillipsite in composition, as is seen by comparison 

 with the foregoing mean of three phillipsite analyses from Murray and Eenard (Challenger 

 Report on Deep-Sea Deposits, page 404). 



Final proof of the zeolitic nature of the substance is afforded by its behavior towards 

 dilute acids, to which it readily gives up alkalies and alumina. Even when it was 

 attempted to clear the substance of manganese for analysis, by treatment with cold dilute 

 acid and Soj, the figures for the bases were found too low. 



An extraction of a sample containing 18.36 per cent of total water was carried out by 



heating up 2 grams and boiling for ten minutes with 50 cc. of l — n. hydrochloric acid, the 



process being thrice repeated. The residue was collected and the extract analyzed, with 

 the following results : — 



