TRANSACTIONS OF SECTION B. 459 
dual character of the sodium light, which forms the principal standard, 
and the extreme feebleness of the hydrogen lines, which are used as 
‘secondary standards’ when dispersive coefficients are required. In order 
to secure a proper correlation of optical measurements of various kinds it 
is suggested that the above series should be abandoned in favour of lines 
selected from the series— 
Li Cd Na Hg Cd Cd Hg 
6708 6438 5893 5461 5086 4800 4359 
the green mercury line replacing the sodium doublet as ‘ principal standard.’ 
In the measurement of rotatory dispersion it has been found that the ratio 
Hg 4359 to Hg 5461 is sufficient to characterise the dispersive power of a 
substance ; in refractometry similar conditicns may be expected to prevail, 
and it is pointed out that these two lines can be read even more easily than 
the hydrogen lines which they displace by using a vacuum tube containing 
a drop of mercury and gently warmed by means of a flame. If additional 
data are desired, the cadmium lines can be read with great readiness by 
employing a tiny silver-cadmium arc, or the sodium and lithium lines may 
be read in the ordinary way by making use of their flame spectra. 
MONDAY, AUGUST 30. 
Joint Discussion with Section K and Sub-Section K (Agriculture) 
on Wheat.—See Appendix A. 
TUESDAY, AUGUST 31. 
Joint Discussion with Section I on the Chemistry of Food. 
Opened by Professor H. E. Armstrrona, F.R.S. 
(i) Proteins: the Relations between Composition and Food Value. 
By HE. Franguanp Armstrona, Ph.D., D.Sc. 
It has been customary hitherto, when investigating the protein content 
of a food material, to determine the amount of nitrogen in it and multiply 
this by the factor 63. The quotient is spoken of as protein without any 
reference to its nature, although it has long been realised that proteins of 
different origin are not the same. The practice was perhaps justifiable so 
long as the chemistry of the proteins remained an almost unexplored field ; 
recent work, in particular that of Emil Fischer, Abderhalden, Osborne, and 
others, makes it possible to take up a more scientific position. 
The proteins have been proved in the main to be built up of amino acids 
belonging both to the aliphatic and aromatic series or derived from cycloids 
containing nitrogen, of oxyamino acids and of diamino acids. The careful 
analytical investigation of a large number of proteins has shown that these 
various structural units are present in very different proportions in the 
different proteins; some of them may be altogether absent from a par- 
ticular protein. Cereal proteins, for example, are characterised by a high 
percentage of glutamic acid, amounting to about 30 per cent. in most cases, 
the legumes yield about 20 per cent., oil seeds about 17 per cent. ; meat 
proteins give only 8-11 per cent. glutamic acid but are characterised by 
containing much glycine. 
‘Ihe amino acids are so very different from one another in their chemical 
