REPORT ON THE PETROLOGY OF THE ROCKS OF ST. PAUL. 17 



This phosphate is dull, and slightly resinous on fresh surfaces, with a sub-conchoidal 

 fracture, and a hardness exceeding that of fluorite. It is opaque in thin splinters, but the 

 edges are slightly translucent : small fragments heated in the test-tube decrepitate loudly. 

 The latter characteristics taken together with those already mentioned points to a close 

 connection between this substance and that erroneously described as pyroclasite by 

 Shepard. 1 The particles of the phosphate of St. Paul that remained in the test-tube 

 blackened, and set free a small quantity of water. Splinters large enough to be held with 

 the forceps in the flame of a blow-pipe turn white, the edges fuse with difficulty to a 

 white enamel, and colour the flame a pale yellow. When treated with borax it yields a 

 colourless glass. I have not been able to discover traces of fluorine in it. 



Thin sections of this phosphate present almost all the microscopical characters 

 described as belonging to the concretionary silica of the Carboniferous limestone 

 cherts, 2 except, perhaps, that the phosphates never exhibit the brilliant chromatic 

 polarisation peculiar to the quartz of the chert. The agate-like structure is very 

 marked. The concretions, more or less circular in form, either stand side by side or run 

 into one another. Their outer edges have zones of deep brown or yellow, while at the 

 centre they are generally colourless. When they merely touch one another at tangential 

 points, intervals filled with a dark opaque substance occur (see fig. 4). Some of these con- 

 cretions are fibro-radial, and in polarised light show the black cross like chalcedony. The 

 concentric zones are made up of lines of extreme delicacy, only distinguishable under very 

 high magnifying powers. The outer zones contain no enclosures of foreign matter, but 

 accumulations of greyish intertwined filaments are noticeable towards the centre. Seen 

 with a high power, these fibres assume a branching dendritic disposition that strikingly 

 resembles vegetable structure. They resist the action of hydrochloric acid much longer 

 than the surrounding mass. Reflected light sometimes shows in the central parts of the 

 concretions small sections of a steel grey colour, bearing a strong resemblance to enclosures 

 of carbonate of lime, Large black or yellowish-brown undulating lines which cross the 

 slides and separate the concretions owe their colour to the presence of hydroxyde of 

 manganese and iron. 



Many of the specimens showing alteration, or impregnated with the phosphate just 



1 He used the word pyroclasite to indicate the manner in -which the mineral acted when subjected to heat. 

 To the phosphorite we are describing should be referred the substances to which Shepard has given the general name 

 of pyro-guanite (American Journal of Science and Arts, IL ii, 1856, p. 96). The minerals thus grouped together were 

 pyroclasite, glaubapatite, and epiglaubite. According to Shepard, these three varieties are found at Monk's Island, in 

 the Carribbean Sea, where they cover the soil to the depth of many inches. They are hard and stony, and on the parts 

 exposed to the action of the atmosphere are cream coloured, while the interior is of a yellowish-brown. Shepard 

 admits the disappearance of the ammonia, and strives to explain it by contact with trap rocks, whose heat might have 

 fused the phosphate. I will return to this theory later on. According to Dana (System of Mineralogy, p. 535), what 

 Shepard calls pyroclasite is only compact guano. 



1 A Renard, Becherches lithologiques sur les Phthanites du calc. carb. de Belgique, BulL Ac. Roy. de Belgique 



Serie, t xlvi, Nos. 9 and 10, 1878. 



C 



