82 HABIT A]!TD INTELLIGENCE. [chap. 



Tlie occurrence of compound crystals is in some species only- 

 occasional, but in others, as in Herscliellite, it is general. 

 Both crystalline and organic structures sometimes assume 



Branching, a branching form ; as for instance in those combinations of 

 ice-crystals called frost-work ; in sea-weeds, in coral, and 

 in trees. But this is characteristic of imperfect crystals 

 and of low organisms. The ice-crystals that constitute 

 frost-work do not present perfect crystalline forms, as is 

 shown by their curved ou.tlines. Sea-weeds and corals are 

 low organisms ; and though a tree is not so, the trunk and 

 branches are less highly organized than the leaves and 

 flowers. 



Another mode of combination that belongs to compara- 

 tively low organisms, and, I believe, to imperfect crystals, 



Chain-like, is by a chain-like succession, as in the successive similar 

 segments of worms, and in those molluscoids called Salp^. 



Double, "Twin crystals," or crystals found in pairs united by 



two of their planes, are common, and are characteristic of 

 some species. Such a binary combination is not common 

 among organisms, but it occurs in the calyx leaves of the 

 poppy, and of that beautiful flower the dielytra. 



Circular. Many crystals unite circularly round an axis. I have 

 already mentioned Herschellite as an instance of this. 



Snow. ^ Snow consists of minute six-rayed stars, or, what is as 

 good a description, six-leaved flowers, each of which is 

 composed of six rhombohedral crystals. And there is 



Staurolite. a species of right prismatic crystal called Staurolite, 

 from the tendency of its crystals to unite into groups 

 of four, shaped like a cross. The resemblance of flowers 



Eesem- to such crystalline forms as these is obvious : and it is a 



• flowers.*° resemblance not only of form but of structure ; for flowers 

 consist of modified leaves arranged round an axis, 



Interpene- Tetrahedral crystals of species belonging to the cubic 



trating system are sometimes found interpenetrating each other, 



crvst&ls 



in such a way that the middle part belongs equally to 

 both. These cannot be compared to any normally com- 

 pound organisms : they can only be compared to those 

 monstrosities in which, for instance, one body equally 

 belongs to two heads. 



