1 2 STROMEYER, Grouping of Chemical Elements. 



reversed would be to endow the above empirical formula 

 with a scientific value which it would be absurd to claim 

 for it. 



The mean positions of the groups are given in Table I., 

 and have been summarised in Table II., which also 

 contains the intervals between the groups, expressed in 

 average group intervals. 



It will be noticed that the (VII.) Fluorine, or rather 

 Manganese, group and the (Villi.) Iron group, as well as 

 the (VIII 2 .) Nickel and the (VIII 8 .) Cobalt groups, fall 

 almost together, whereas the (VI.) Oxygen and (VII.) 

 Fluorine groups, the (VI II 3 .) Cobalt and the (IX.) Sodium 

 groups, and the (X.) Magnesium and the (XI.) Aluminium 

 groups, are separated by about three half average group 

 spacings, and the distance of the mean (VIII.) Iron, 

 Nickel, Cobalt groups from the (IX.) Sodium group is 

 i "89 average intervals. It will also be noticed that the 

 (VI.) Oxygen group falls at 5*27, and the (XI.) Aluminium 

 group falls at 11 "34, so that between these two groups 

 there are &i spaces for five groups and only 99 spaces 

 for eleven groups. The (IV.) Carbon and the (XII.) 

 Silicon groups both seem to split up naturally into two 

 sub-divisions, as indicated in Table I. These four might 

 be called the Carbon, Titanium, Silicon, and Tin groups. 

 The (XIV.) Sulphur group, as will be seen from Table I. 

 and from the diagram, slants so considerably that the 

 elements of which it consists are separated from one 

 another by whole group intervals, and it would almost 

 seem as if there should be not one group but three, for 

 each of which, however, only one element had been 

 discovered. 



A considerable difference is also noticeable between 

 the mean positions of groups which are separated by 8 



