192 CHEMICAL DISCOVERY AND INVENTION 



it develops two more units of valency and becomes quadri- 

 valent. 



If in common salt the chlorine atom owes its univalency to 

 the presence of one electron, and in water the oxygen owes its 

 bivalency to two electrons, it appears that the combination of 

 the chloride with water must be due to some other agency than 

 the presence or absence of electrons. This question is still the 

 subject of active debate, and restrictions as to the number of 

 valencies which may be assumed have of late been considerably 

 relaxed. 



The most interesting and practically useful assumptions are 

 those which within the last few years have been proposed by 

 Professor Alfred Werner of Zurich, as the result of investigations 

 into the composition and properties of the complex compounds 

 formed by the union of the salts of certain metals, such as platinum 

 and especially cobalt, with ammonia. Many of these have long 

 been known, and their constitution has remained a puzzle. 

 Werner has shown that only a portion of the groups of atoms 

 attached to the central metallic atom are capable of forming 

 ions, so becoming chemically reactive. Those which are not 

 ionisable, are connected by a peculiar subsidiary valency and 

 never exceed six. The number of these is known as the co- 

 ordination number. 



It is possible that the case of water of crystallisation, already 

 referred to, may be found to come within Werner's hypothesis. 



The question arises therefore whether valency, that is, capacity 

 to enter into chemical combination, is in all cases electrical, and 

 due to the intervention of electrons. It is of course well known 

 that some elements usually very readily ionisable show no signs 

 of the same property in certain forms of combination. Thus 

 chlorine when in the form of the chloride of hydrogen or a metal 

 is immediately reactive when brought into contact with other 

 ionised elements or groups of elements. But the compounds, 



CH 3 C1, CH 2 C1 2 , CHC1 3 , and CC1 4 , 



and many others in which the halogen is directly attached to 

 carbon, show no signs of response to the common reagents. 

 Thus pure chloroform, CHC1 3 , does not give, like other chlorides, 

 a white precipitate of silver chloride. It does, however, interact 

 with ammonia, exchanging its three atoms of chlorine for one 

 atom of nitrogen, producing hydrogen chloride and hydrogen 



