488 Prof. Odling on the Natural Groupings 



an acid teroxide, or ferric acid. Copper appears to form a 

 similar compound when its hydrate, diffused in an alkali, is 

 treated with chlorine. Many of the above-mentioned ill-defined 

 compounds serve to shadow forth the range of analogies. It is 

 evident, however, that iron, which is at one extremity of the 

 series, has a tendency to form hypcroxides, — copper, which is at 

 the other extremity, to form hypo-oxides, — while nickel is inter- 

 mediate both in position and tendencies. The parallelisms of 

 nickel with cobalt, and of iron with manganese, are most com- 

 plete. It is observable that the colour of manganous and cq- 

 baltous solutions is complementary to that of ferrous and 

 nickelous solutions. Cobalt has a greater tendency than nickel 

 to form hyperoxides. In this respect iron and manganese are 

 much on a par. The superior tendency of manganese is shown 

 in the ready oxidizability of its metal, and in its formation of 

 permanganic acid ; but, on the other hand, ferrous salts oxidize 

 much more rapidly than manganous. Manganese is doubtfully 

 magnetic. The binoxide of manganese is represented by the 

 bisulphide of iron. 



Chromium is associated with this group by marked analogies. 

 The atomic weight is sequential M'ith those of manganese and 

 cobalt : — 



Cr = 26-3 85-4 „^ „ 



Mn = 27-6 -3- =27-8 mean. 



Co = 29^ 



Sum 85-4 



The atomic volume assigned to chromium by Kopp is 69, a 

 number obviously too high. The equivalent of chromium being 

 328' 75 on the oxygen scale, we should have for the calculated 

 specific gravity of the metal — 



328-75 



328-75 



= 4-76. Kopp. 

 = 4-98. Miller. 

 = 7-00. 



44 



Now, according to Richter, the specific gravity of chromium 

 is 5*9, which is exactly intermediate between the results given 

 by Dr. Miller's number 66, and the iron number 44. However, 

 Richter's determination of the specific gravity of chromium was 

 made with a specimen of metal that contained carbon, and is 

 therefore, in all probability, somewhat lower than the actual 

 specific gravity of the pure metal. In the sesquioxides of iron 

 and cobalt, as in nearly all protoxides, we find the atomic volume 



< 



