Ill] INDEPENDENT SIMILARITIES OF STRUCTURE. 107 



Surely, however, there can be no real parity between tlie 

 relationship of existing minerals to nebular atoms, and the 

 relationship of existing animals and plants to the earliest 

 organisms. In the first place, tlie latter have produced 

 others by generative multiplication, which mineral atoms 

 never did. In the second, existing animals and plants 

 spriug from the living tissues of preceding animals and 

 plants, while existing minerals spring from the chemical 

 affinity of separate elements. Carbonate of soda is not 

 formed, by a process of reproduction, from other carbonate 

 of soda, but directly by the suitable juxtaposition of car- 

 bon, oxygen, and sodium. 



Instead of approximating animals and minerals in the 

 mode suggested, it may be that they are to be approximated 

 in quite a contrary fashion ; namely, by attributing to 

 mineral species an internal innate power. For, as we 

 must attribute to each elementary atom an innate power 

 and tendency to form (under the requisite external condi- 

 tions) certain unions with other atoms, so we may attribute 

 to certain mineral species — as crystals — an innate power 

 and tendency to exhibit (the proper conditions being sup- 

 plied) a definite and symmetrical external form. The dis- 

 tinction between animals and vegetables on the one hand, 

 and minerals on the other, is that, while in the organic 

 w^orld close similarity is the result sometimes of inheritance, 

 sometimes perhaps of direct production independently of 

 parental action, in the inorganic world the latter is the con- 

 stant and only mode in which such similarity is produced. 

 When we come to consider the relations of species to 

 space — in other words, the geographical distribution of 

 organisms — it will be necessary to return somewhat to the 

 subject of the independent origin of closely similar forms, 



