P. G. H. Boswell — Petrology of Suffolk Box-stones. 253 



The brown matrix is often present in considerable amount, measure- 

 ments of relative volumes of sand-grains and cement-filled interspaces 

 being for extreme values — sand 19-8 per cent, matrix 80 - 2 per cent, 

 and sand 80-1 per cent, matrix 19-9 per cent. (These values were 

 calculated from the areal distribution in thin sections, e.g. Figs. 1 and 2.) 

 An analysis of material from the mouth of the Deben [203] gave the 

 following proportions by weigbt : sand 57*43 per cent, matrix 42 57 

 per cent, from which the proportions by volume, sand 62 per cent, and 

 matrix 38 per cent, can be calculated roughly on the assumption that 

 the sand is quartz and the matrix calcium phosphate. 



The cement sometimes exhibits concentric banding due to growth 

 round the quartz-grains. On treatment with dilute acid a large 

 proportion of phosphate is found, 1 and an analysis of the rock kindlv 

 carried out by my colleague, Mr. G. E. Kirby, A.R.C.Sc, B.Sc*, 

 shows 20" 1 per cent of calcium phosphate, 6-5 per cent of calcium 

 carbonate, and - 84 per cent of soluble silica. 



This percentage of phosphate is far too great to be accounted for 

 by the little detrital apatite present, hence it probably exists in the 

 cement as limonite-stained phosphorite. Glauconite may also have 

 played a part in cementation, but it has not been detected as such. 

 Surprisingly few grains of this mineral occur in the residues. 

 Limonite itself undoubtedly acts as a cement, and prolonged boiling 

 with acid is required to clean the crushed sandstone. Cementation 

 has also been effected by the deposition of secondary silica, partly as 

 quartz and partly in an opaline form. The latter appears to be 

 soluble in acid, and is found on analysis of the matrix; the finding 

 of opal also among the detrital mineral grains suggests that some of 

 the apparently isotropic pellets and matrix may be this form of 

 silica obscured by iron-staining. Other pellets may be iron-stained 

 calcareous material. Marked effervescence with cold dilute acids 

 indicates that calcium carbonate also is present as cement. 



Mineral constitution. — The sandstones, being fairly hard, were 

 crushed in a steel mortar, and, after sifting to 1 mm. the compound 

 grains were removed. The material was subjected to prolonged 

 boiling with dilute hydrochloric acid to remove the matrix from the 

 grains. A clean, white, micaceous sand was obtained, but it is 

 noteworthy that the muscovite is not strikingly apparent in the 

 untreated box-stones. The sand was then treated with either 

 bromoform or Sonstadt solution (mercuric potassium iodide), and 

 crops obtained of density < 2 - 59, which contained potash felspars, 

 opal, etc., between 259 and 2*8, which contained quartz and 

 plagioclase felspars, and > 2-8. The last crop was then farther 

 separated by electro-magnetic action (after minerals such as magnetite 

 had been removed by means of a bar-magnet), and the non-magnetic 

 portion treated with methylene iodide (density 3-33). The crop of 

 density between 333 and 2"8 contained andalusite, apatite, mica, etc. 

 Mineral grains were mounted temporarily in clove oil (R.I. 1-534), 

 as being very similar in refractive index to Canada balsam, and in 

 other oils of known and differing refractive indices, for examination 



1 There is sufficient in the rock for a fragment to yield the ordinary blowpipe 

 reactions for a phosphate. 



