TEANSACTIONS OF SUCTION C. 567 



80 saturated with the suhstance under examination at a high temperature, that it 

 became opaque either on cooling or when slowly re-heated. It was again fused so 

 as to be quite transparent, and then very slowly cooled over the flame. If pro- 

 perly managed, the excess of material held in solution at a high temperature 

 slowly crystallised out, the form and character of the crystals depending on the 

 nature of the substance and on the presence of other substances added to the bead 

 as test reagents. By this means I proved that in a few exceptional cases small 

 simple solid crystals are formed. More frequently they are compound, or occur 

 as minute needles, but the most characteristic peculiarity is the development of 

 complex skeleton crystals of extreme beauty, built up of minute attached prisms, so 

 as to give rise to what would be a well-developed crystal with definite external 

 planes, if the interspaces were all filled up. In many cases the fibres of these 

 skeletons are parallel to three different axes perpendicular to one another, and it 

 might be supposed that the entire skeleton was due to the growth of small needle- 

 shaped crystals all uniformly elongated in the line of one crystalline axis, so that 

 the resulting mass would be optically and crystographically complex ; but in some 

 cases the different systems of fibres or needles are inclined obliquely, and then the 

 optical characters enable us to prove that the separate prisms are not similar to 

 one another, but developed along diflfereut crystalline planes, so as to build up one 

 definite crystal, mechanically^complex, but optically and crystallographically simple, 

 or merely twiuned. In a few special cases there is a well-pronounced departure 

 from this rule, and truly compound groups of prisms are formed. In the centre 

 there is a definite simple prism ; but instead of this growing continuously in the 

 same manner, so as to produce a larger prism, its ends, as it were, break up into 

 several smaller prisms, slightly inclined to the axis of the first ; and these secon- 

 daiy prisms, in like manner, break up into still smaller, so as ultimately to give 

 rise to a curious complex brush-like growth, showing in all positions a sort of fan- 

 shaped structure, mechanically, optically, and crystallographically complex. 



I have done my best to describe these various kinds of crystals seen in blowpipe 

 beads as clearly as can be done without occupying too much time, but feel that it 

 is impossible to make the subject as simple as it really is without numerous illus- 

 trations. However, for the purpose now in view, it will, I trust, suffice to have 

 established the fact that we may divide the crystals in blowpipe beads into the 

 following groups, which on the whole are sufficiently distinct, though they neces- 

 sarily pass one into the other. 



1. Simple crystals. 3. Fan-shaped compound groups. 



2. Minute detached needles. 4. Feathery skeleton crystals. 



It must not be supposed that crystals of one or other of these groups occur 

 promiscuously and without some definite relation to the special conditions of the 

 case. Veiy much depends upon their chemical composition. Some substances 

 yield almost exclusively those of one group, and other substances those of another, 

 whilst in some cases a difference in the rate of cooling and other circumstances 

 give rise to variations within certain limits; and, if it were possible to still 

 further vary some of the conditions, these limits would probably be increased. 

 Thus, for example, the earliest deposition of crystalline matter from the glassy 

 solvent is sometimes in the form of simple solid prisms or needles, but later on in 

 the process it is in the form of compound feathery tufts ; and if it were possible to 

 cool the beads much more slowly whilst they are very hot, I am inclined to 

 believe that some substances might be found that in the early stage of the process 

 would yield larger and more solid crystals than those commonly met with. This 

 supposition, at all events, agrees with what takes place when such salts as potas- 

 sium chloride are crystallised from solution in water. Some of my blowpipe beads 

 prove most conclusively that several perfectly distinct crystalline substances may 

 be contemporaneously deposited from a highly heated vitreous solvent, which is an 

 important fact in connection with the structure of igneous rocks, since some 

 authors have asserted that more than one mineral species cannot be formed by the 

 slow cooling of a truly melted rock. The great advantage of studying artificial 



