из 



the following mineral would more nearly indicate that both of 

 them occupy a peculiar position among the other minerals. 



According to Winther^) britholite has the following com- 

 position: 



3[4 5гОо -iRoO.-Z'RO- ЩО -NaF] -îlPo О, • R.,0.,] 



If we do not take the fluorine and the water into con- 

 sideration we get also here something very like an orlosilicate 

 and an ortophosphate, and the quantity of water is here com- 

 paratively so slight, that it cannot in any essential degree alter 

 this result. 



The ardennite has the empirical formula blhO • 8MnO, 

 AAhOi'VoOr,- SSiO^; this composition is too basic to corre- 

 spond to the common form. 



The last mineral, the epistolite, is still less than the pre- 

 ceding ones to be referred to the common form; its composi- 

 tion is after Bog g ild 2) 



l9SiO., . 4 ГгО. • ЬШоО, 'RO- \0Na._O- 2\H._0 -ANaF 



Contrary to what was the case in the two preceding mine- 

 rals the quantity of acids is here quite predominant. By refer- 

 ring almost all the water to the basic ingredients, we may get 

 a compound formed consisting for the two third parts of 

 H^SiO.j, HoTiO^j and H^Sb^O^^ and only for one third part 

 of the sodium salts of these acids; such a compound, however, 

 is not very probable. It must then be supposed, either that 

 the mineral is a compound of very acetous silicates, tilanates, 

 and niobates , and in this case some of the water must even 

 be referred to the bases in order to saturate the acids, or 

 something of Тг'Оо or iVè.^Og, or both, must act as bases; in 

 both cases the composition will differ very much from the 

 common one. 



') Medd. om Grønl. XXIV p. 195. 

 ') Medd om Gionl. XXIV p. 189. 



XXVI. 



