ANALYSIS OF COMPLEX MATTER 131 



soluble body from one less soluble, even when they are so 

 associated as to be indistinguishable in other ways. In a simi- 

 lar manner, a substance may be shown to contain more than 

 one kind of matter by reason of its parts having different 

 boiling-points. Investigation must take many directions 

 before we are satisfied that a given substance is homogeneous 

 and individual. We have learnt also that a body may exhibit 

 another kind of complexity ; for a body which cannot be 

 resolved into dissimilar parts by some processes may yet yield 

 to others. For example, a body, perfectly homogeneous in 

 some respects, may be decomposed when raised sufficiently in 

 temperature, or when forming part of an electric circuit, or 

 when placed in contact with other substances. Such bodies 

 are called ' chemical compounds,' and are distinguished from 

 other complex bodies by the fact that the same body contains 

 always the same relative quantity of each component. When 

 the process of analysis has been carried to its extreme limit, and 

 when specimens of matter from all portions of the earth's 

 surface have been subjected to the test, it has been found, so 

 far, that there are over seventy different kinds of matter which 

 resist further decomposition. These are called ' elements ' ; and 

 their different modes of combination with one another give 

 rise to that great variety which the surface of the earth 

 presents to our senses. It must be remembered, however, that 

 fresh investigations are always leading to new knowledge, and 

 the future may show that there are more elements ; while, on 

 the other hand, it may show that there are fewer, by proving 

 that our elements are really compounds of a smaller group of 

 more elementary substances. It must also be remembered 

 that the distinction between chemical compounds and ordinary 

 mixtures is one which must not be adhered to without dis- 

 cretion. Although very serviceable in preliminary classifica- 

 tion, it will have to give way when we come to investigate a 

 large number of bodies which belong partly to one class and 

 partly to the other. The nature of alloys, the considerable 

 change of property in large masses produced by slight admix- 

 tures, the nature of many vegetable and animal products, all 

 seem to show that there are relationships between different 



K 2 



