449 TRANSACTIONS OF SECTION B. 
The metallic carbonyls possess altogether remarkable properties: at present, 
these defy explanation ; nickel carbonyl in particular seems to be an excep- 
tion to all rules. The complex iron carbonyls made known by Dewar and 
Jones also have most fascinating attributes, the variety of colours they 
display being specially interesting. The marked individuality of the 
members of the iron group exemplified in their carbonyl derivatives is in 
striking contrast with the tendency they display to behave as related 
elements ; the deeper problems of valency are clearly exposed for considera- 
tion in such peculiarities. 
The discoveries of the special activity of magnesium as a synthetic agent 
and of the superior value of nickel as a catalyst in fixing hydrogen are 
other illustrations of the individuality of metallic elements. We are 
greatly indebted to the French chemists for the invaluable preparative 
methods they have based on the use of these two agents. 
Although satisfactory progress has been made in almost every direction, 
the nature of many nitrogen compounds is still not properly understood. It 
is clear that we are as yet in no way seized with understanding of the 
attributes of this element as we are of those of oxygen and carbon, particu- 
larly in the case of mixed carbon-nitrogen compounds: we can make nothing 
of the physical data such substances afford. Nitrogen, in fact, is an extra- 
ordinary element, far more remarkable than any other; its ‘temper’ 
appears to vary more than that of any other element according to the 
character of its associates—nothing could be more remarkable, for example, 
than the change in properties from ammonia, NH;, through hydrazine, 
NH..NH., to azoimide, N;H. No other element can be so poisonous, so 
immediately fatal to life. We lack a model symbolic of its functions—which 
means that we are unable to fathom its vagaries and reduce them to simple 
order. 
The oximes and the diazo-compounds in particular have given rise to 
much dispute. Stereo-chemical formule have been assigned to these, but 
probably they have little relation with the truth ; although they have been 
of service by supplying symbols which can be offered up at examinations, by 
confining attention they have served to sterilise inquiry. No better illustra- 
tion could be given of the truth of the remark made by my friend the Pro- 
fessor that man is an idolater by nature, a fact that chemists should always 
bear in mind. 
The compounds in question are difficult substances to handle, far too 
prone to undergo change without invitation—it is to be feared that many of 
the conclusions which have been arrived at are based on incomplete if not 
unsatisfactory evidence." When I think of the state of our knowledge, I 
am reminded of the father of diazo-chemistry, Peter Griess, a man of marvel- 
lous experimental gifts; there is great need of such a man to reinvestigate 
the whole subject. 
If we inquire as to the general effect of the increase of knowledge of 
organic compounds, it is clear that the lessons which emerge from all modern 
inquiries are such as to justify Larmor’s remark that our conceptions of 
structure must be granted more than analogical significance. Nverything 
tends to show that function and structure are most closely connected—odour, 
taste, colour, physiological effect are specific rather than general properties, 
each conditioned in its special variety by some special structure; we are 
* Since this was written, Thiele’s discovery of ‘ Azomethane,’ MeN : NMe, has 
been announced. This is described as being a distinctly coloured, very pale 
yellow substance in the solid state. There can be little if any doubt, there- 
fore, that, as Robertson and I have argued, the colourless so-called syn- and anti- 
diazo-salts cannot possibly be compounds of the —N : N— or diazene type : such 
compounds would all be at least yellow in colour. 
