8 Mr. William Sutherland on the 



For the first six compounds the ratio is the same as the 

 average in the last Table ; but in H 2 S, H 2 0, and NH 3 it 

 falls to a lower value, the reason for which appears to be the 

 presence of the H atoms operating like the two terminal JET 

 atoms in the paraffins. To see whether this is the case, let 

 us suppose 8, 0, and N carry the values already found for 

 them into (M 2 /)* and M/3, then we get for 



H 2 in H 2 S. H 2 in H 2 0. H 3 in NH 3 . 



B 7 '■' 9 6 



F 16 19 1-2 



B/F 44 4-7 50 



In each case the ratio comes near to the value 4 found for 

 the terminal H in the paraffins ; though both in (M 2 /)* and 

 M/3, H 3 in NH 3 appears smaller than H 2 in H 2 S, H 2 0, or the 

 paraffins. 



Returning to the values of the ratio in Table VII., the 

 values for the compounds of P aud As seem also to come near 

 to the value 10 ; but this has no particular significance, as 

 we shall see later on that a different principle is at work 

 amongst some of these compounds ; even at present we can 

 see that (M 2 /)* for PC1 3 has a value 6'6 } which is exactly three 

 times the value adopted for CI, leaving nothing for P ; and 

 when we turn to the value for CC1 4 , namely 6'8 ? which is a 

 good deal less than that for 4C1, we must recognize clearly 

 that we are getting into a type of compound in which the 

 pure additive principle ruling in the carbon series no longer 

 applies. Of course, in connexion with CC1 4 it may be said 

 that it might be expected to be exceptional because CH 4 is so ; 

 but then CH 3 C1 and CH 3 I are not exceptional, a fact well 

 worthy of attention in view of the exceptional nature of CH 4 . 

 Indeed, if we take the series of ratios 1) for CI, 12 for 

 CHC1 3 , and 12 for CC1 4 , with 13 for SiCl 4 , and J 3 for SnCl 4 , 

 we see an increasing departure from proportionality between 

 (WHy and volume M/3, which will become comprehensible 

 when we have dealt with inorganic compounds as a whole. 

 An attempt to trace the intermediate cases, such as that of 

 CHC1 3 , would lead us into too much detail, though promising 

 light on the great old controversy in chemistry between the 

 dualistic and unitary theorists and on the transition from 

 electrolyte to non-electrolyte. 



But without going into transition cases, there are a few 

 more values which we ought to discuss before leaving the 

 carbon serial compounds, as they bear upon the large value 

 of F which belongs to the H atom when apart from the 



