468 



Prof. Church on Tasmanite, 



I. 



II. 



III. 



IV. '3668 grm. of substance gave *9796 grm. carbonic acid 

 V. 



'4288 grm.* of substance gave 1*1466 grm. carbonic 



acid and 0*0354 grm. ash. 



•3535 grm. of substance gave *3019 grm. water and 



•0:284 grm. ash. 



•3087 grm. of substance gave '2729 grm. water and 



•0248 erai. ash. 



and *3I5 grm. water 



>5 grm. of substance gav 



e *175 grm. of sulphate of 



bA 



47* 



barium. 

 VI. *5125 grm. of substance gave "041 grm. of ash. 

 -VII. 6*16 grains of substance gave 2*12 grains of sulphate 

 of barium. 



VIII. -337 grm. of substance gave *121 grm. of sulphate of 

 barium. 



IX. '654 grm. of substance gave '2302 grm. of sulphate of 

 barium. 



X. *421 grm. of substance gave -0525 grm. of ash. 

 L XL '482 grm. of substance gave '0580 



grm. of ash. 



These results correspond to the following percentage num- 

 bers : — 



i I. 



II. 



... 



9-58 



8-03 



III. 



... 



9-82 



8-03 



IV. 



72-83 

 9-54 



V. 



VI. VII. 



VIII 



IX. 



X. 



XL 



Carbon .... 

 Hydrogen . 

 Sulphur .... 

 Ash 



J 72-92 

 J 8-25 



4-97 



8-00 



4-72 



... 



4-93 



4-83 



12-47 



12-03 



The following are the mean percentages, rejecting the hydrogen 

 of analysis III. as evidently too high : — 



Carbon . . . 72-88 

 Hvdrogen . . 9'56 

 . . 4-90 



Sulphur 

 Ash (a) 



8-14 (b) 12-24 



Before these numbers can be taken to represent the centesimal 

 composition of pure Tasmanite, they require recalculation after 

 deduction of the ash, which is, without doubt, an accidental im- 

 purity. This ash mainly consists of silica and alumina; it con- 

 tains also a small quantity of ferric oxide and of some soluble 

 sulphate, this latter compound being derived in all probability 

 from the oxidation of the sulphur contained in the Tasmanite 

 proper. Since this ash has almost exactly the same composition 

 as the shaly rock in which the grains of Tasmanite occur, we may 

 directly subtract it without introducing any appreciable error. 

 The percentages deduced from analyses I. to IV. are given 



