1823.] improved Method of making Coffee, 31 



diminishing expenditure on one article, they allow of some other 

 enjoyment which was before unattainable. A reduction on 

 quantity permits indulgence in superior quality. In the present 

 instance, the importance of economy is particularly great, since 

 it is applied to matters of high price, which constitute one of 

 the daily meals of a large portion of the population of the earth. 

 That in cookery also, the power of subjecting for an indefinite 

 duration to a boiUng heat, without the sHghtest dependition of 

 volatile matter, will admit of beneficial application, is unques- 

 tionable. 



Article VII, 



On Ultramarine, and the Methods hy which its Purity may he 

 ascertained. By R. PhiUips, FRS. L. and E. 



Before the time of MargrafF, whose analysis of lapis lazulj 

 was published in 1768, the colouring matter of this mineral was 

 supposed to be copper ; according to the chemist just mentioned, 

 as quoted in Klaproth's analyses, vol. i. p. 163, lapis lazuli con- 

 sists of oxide of iron, silica, lime, and its sulphate, omitting any 

 notice of the alumina which it contains in very considerable 

 quantity, and without stating the proportions of the ingredients 

 enumerated as its constituents. 



Rinmann and Cronstadt have also mentioned the composition 

 of this mineral, but their statements are so inaccurate as to 

 require no further notice. According to Klaproth, lapis lazuli 

 consists of 



Silica 46-0 



Alumina 14*5 



Carbonate of hme 28*0 



Sulphate of hme Q'^ 



Oxide of iron 3*0 



Water 2-0 



lOO-O 



With respect to the colour of this substance, Klaproth observes, 

 that " though the researches of Margraff" have refuted the 

 opinion formerly received, that the blue colour of the lapis 

 lazuli originated from an admixture of copper ; and though it has 

 been demonstrated that the colour of this fossil is owing only to 

 iron, yet its other constituent parts have not yet been deter- 

 mined with due accuracy."' 



Now as neither the protoxide nor peroxide of iron could ba 

 suspected of imparting a blue colour, it is singular that Klaproth 



