Vivianite in a Fireclay. 357 



top, middle and right ; Fig. 2, top, middle). In some cases one of 

 the bands, or one part of it, may be seen to consist of two or more 

 very closely contiguous bands (see Fig. 1, lower left ; Fig. 2, 

 centre, and lower left). 



Some examples show bands which are excentric (especially Fig. 2, 

 upper left and bottom, middle). 



The bluish-black ]ugmenting material of these stains loses colour 

 on heating, gives an iron borax head, dissolves readily in hydro- 

 chloric or nitric acid, and gives a definite phosphate reaction. It 

 is presumed that it is the earthy amorphous variety of vivianite. 

 Under the microscope it is opaque, though a few flakes can be found 

 which show a pleochroism and birefringence which agree with those 

 of vivianite. 



An examination into the possible mode of formation of these 

 stains brings out several points of great interest. It is supposed 

 that the vivianite has been formed by precipitation as a result of 

 reaction of solutions carrying iron and phosphate, the iron solution 

 being dilute and deriving its iron from the general mass of the fireclay, 

 which contains only a trace of iron, and the phosphate solution being 

 more concentrated and derived from some phosphatic body in the 

 clay. The phosphatic body may have been bone, fish scale or 

 coprolite, if of animal origin, or plant root or stem, if of vegetable 

 origin (phosphates being essential constituents of plants). From 

 the information available at present no definite conclusion can be 

 reached with regard to the original character of the central phosphatic 

 bodies. 



From their general characters, as described above, there can be 

 little doubt that the stains under discussion have bee]i formed as a 

 result of diffusion of solutions in opposing directions, as the some- 

 what similar " rhythmic precipitates " are formed in gelatine, agar, 

 etc.^ In the opinion of the writer the colloidal material of the fire- 

 clay itself formed the medium through which the diffusion took 

 place. The vivianite bands are unlike Liesegang rings, which become 

 wider apart away from the centre of diffusion, and they are also 

 unlike the modifications produced by the writer, which are evenly 

 spaced under certain conditions, or becoming closer together away 

 from the centre of diffusion under certain other conditions (op. cit., 

 pp. 19-21 and 24). 



Examination of the accompanying photographs will show how the 

 spacing of the bands of vivianite varies and appears to follow no 

 law which has yet been worked out in connexion with banded 

 precipitates. In discussions of banded precipitates which have 

 appeared up to the present time the effect of the medium through 

 which diffusion takes place, upon the reaction, has either been 

 supposed to be very slight, or has been neglected. When dealing 

 with certain reactions in gelatine the effect of the medium appears 



^ Amer. Jour. Sci., Ser. iv, vol. xliii, 1917, pp. 1-26 



