MODE OF ANALYSIS. 471 



the gr,pen grains of this earth are chlorite, or chlorite blended with mica, 

 and rarely specks of serpentine, I cannot but express the opinion, that as a 

 mineral manure the efficacy of the green particles has been greatly over- 

 rated. As these particles are very little liable to decomposition, their ac- 

 tion, whatever it may be, must be slow, and, I should infer, nearly imper- 

 ceptible. Indeed, I am rather disposed to regard its favourable operation, 

 if indeed it has any, as flowing from a mechanical agency, after the man- 

 ner of a clay, than as arising from the liberation of its potassa through 

 chemical decomposition. Not that I would call in question the usefulness 

 of the earth taken as a whole, for happily this is too well established. But 

 when I find a decided content of sulphate of lime, with carbonate and phos- 

 phate of lime in addition thereto, together with distinct traces of organic 

 matter, it appears to me unnecessary to look any farther in order to account 

 for the phenomena in the case. 



" I now proceed to state my method of examination, together with the re- 

 sults obtained. 



" The specimens were kept in a dry room, exposed to air in shallow dishes, 

 for several weeks ; after which, portions free from crystals of sulphate of 

 lime visible by the naked eye, and large fragments of shells, were heated 

 in a platina capsule to 300, Fah., in order to expel hygrometric moisture, 

 and subsequently to low redness, to decompose organic matter.* The or- 

 ganic matter is very inconsiderable, and was in no instances rigidly de- 

 termined. 



" Having ascertained by experiment that the iron-pyrites was not decom- 

 posable by tepid dilute hydrochloric acid, the following method was resorted 

 to for the determination of the phosphate of lime. Two hundred grains 

 of the triturated earth were suffered to stand (with occasional agitation) in 

 contact with a dilute hydrochloric acid for three hours. The whole was 

 then transferred to a filter, and the earth well washed thereon, with abund- 

 ance of tepid water. The clear fluid and washings thus obtained were 

 super-saturated with ammonia, and the precipitate subsequently digested 

 in a warm potassic solution for the removal of the silica and the alumina. 

 The per-oxide of iron and phosphate of lime now remaining, after being 

 well washed, were treated with a cold, dilute acetic acid, whereby the 

 phosphate alone was taken into solution. It was then precipitated by 

 ammonia, dried, ignited, and weighed. Having found reason to believe 

 that the proportion of finely divided phosphate of lime was pretty uniform 

 in the different specimens of the green-sand, I was only at the pains to 

 determine its exact proportion in specimen No. l.f Having ascertained 

 how much per-oxide of iron each sample contained, this amount was de- 

 ducted from that yielded by the treatment of the same specimen with nitro- 

 hydrochloric acid (aided by gentle heat), whereby the sulphuret of iron 

 was decomposed. Thus the exact quantity of iron which was engaged by 

 the sulphur (and consequently the amount of bi-sulphuret of iron) was 

 ascertained. 



" The carbonate of lime was determined in the usual way, viz., by treating 

 the first obtained solution in hydrochloric acid with ammonia, whereby the 



*This last step was always attended with the extrication of a little sul- 

 phur. 



f I will here observe that, by the process now described, it was ascer- 

 tained that had the whole of the precipitate by ammonia from the hydro- 

 chloric acid solution been taken for phosphate of lime, it would have in- 

 volved the error of an over-estimate of the phosphate by nearly 800 per 

 cent. 



