KOhig.] 51b f j„„. 19, 



inch apart, and between which the bead is made to pass with slight friction. 

 In order to dissolve the entire 2 milliifrs., I bring the powder from the 

 scale i)an on a square piece of Swedish filter-paper quarter inch square, 

 and place this upon a dark polished surfoce (bottom of an agate mortar). 

 With the last particles of the oxydes the paper is made to adhere to the 

 bead, burnt and thus no loss is possible. I keep a series of ten beads (from 

 10-10 per cent.) hermetically sealed in a glass tube for a primary com- 

 parison, and having thus established approximately the position in the 

 scale, it is easy to select one directly above and another below, with which 

 to compare. Any person accustomed to make colorimetric observations 

 with liquids will be able to use this method at once, without any special 

 practice. As the reflection from the spherical surface interieres to a cer- 

 tain extent with the adjustment of the color in the scale, I prefer flattening 

 the beads upon a healed steel anvil with a polished hammer, also heated, 

 to prevent a too rapid chilling of the glass. 



The thickness of the wire determines the thickness of the glass. I place 

 the several glass plates alongside each other upon a flat porcelain slab, by 

 which means the slightest difference in shade and depth may be recognized. 

 The intensity of color is much greater towards the pure titanium than 

 towards the zero ; the blue rays being entirely absorbed by the glass at 

 25 p. TiO.^, the glass being then purely pink. I am now experimenting 

 upon the feasibility of extinguishing the color of the titanium by a gradu- 

 ated scale of green, representing per cent of titanium, by which means I 

 am led to believe this colorimetric method, before the blow -pipe, will be 

 made easier and more perfect. 



2, Arfvedsonite. 



Crystallograplucal properties. The crystals are elongated prisms with 

 the clinopinakoid. As the faces are of equal width the cross section forms 

 an equilateral hexagon. No terminal faces could be observed. The 

 obtuse prismatic angle was found 124° 30' with tiie hand goniometer. The 

 habitus is altogether like that of common hornblende. In one large crystal, 

 which I obtained loose, measuring nearly 4 inches in length by one inch 

 in width, I noticed a disaggregation similar to that observed in uralite, 

 that is to say, the crj-stal breaks up, or crumbles into an infinite number 

 of small individuals. The principal direction of this jxirting appears to be 

 a plane parallel to the orthopinakoid, a face not developed in the crystal. 

 These small individuals exhibit brilliant faces and give a prismatic angle 

 of 1240 5' on the reflexion goniometer. On the smaller and more compact 

 crystals, which are imbedded in quartz. I observed a cleavage parallel to 

 the orthopinakoid quite as marked as that parallel to the prism. I could 

 not measure the angle accurately, but it appeared larger than the angle of 

 the prism. The cohrr is raven black (after removing superficial impurities 

 by hydrochloric acid), the lustre submetallic, and the color of fine powder 

 is lavender blue. 



Hardness = 6. 



