126 MASS. EXPERIMENT STATION BULLETIN 166. 



original Hubl method under control conditions, ^ should be given prefer- 

 once, notwithstanding American practice to the contrary. Furthermore, 

 Wijs has shown 2 that his solution yields practically theoretical results 

 with a number of pure fatty acids. The solution is far more stable than 

 that of Hubl and more rapid in its action. 

 The structure of unsaturated acids, particularly the position of the 

 double bond in relation to the carboxjd group, influences the iodine absorp- 

 tion. ^ If the double bond is located at a considerable distance from the 

 carboxyl the results are generally normal, but when relatively close to- 

 gether the iodine number is likely to be below theory, although lengthen- 

 ing the absorption period increases the results. 



Wijs solution is said to yield high and variable results with cholesterol, 

 and low results with rosin and rosin oils, increasing with the excess of 

 iodine and the length of the absorption periods. 



Olein is the principal unsaturated glyceride in non drying oils and fats, 

 linolin constitutes a considerable proportion of drying and semidrying 

 oils and to some extent of nondrying oils and solid fats, and linolenin 

 occurs in large amounts in all vegetable drying oils. Clupanodonin ap- 

 pears to be the characteristic constituent of fish, liver and blubber oils. 



Linolic and linolenic acids and their glycerides absorb oxygen from the 

 air at ordinary temperature and dry to a hard elastic layer, and to this 

 property drj'^ing oils owe their value. On drying, such oils * absorb oxygen 

 ^vith oxidation of unsaturated acids, resulting in the formation of oxj'- 

 acids and corresponding loss of iodine number; polymerize with an increase 

 in specific gravity and viscosity and additional loss in iodine number; 

 partially deh3'di'ate with formation of stearolactone; and Uberate carbonic 

 acid. These changes result in considerable molecular rearrangement. 

 Oils that have been blown or boiled absorb less iodine. It is stated that 

 for practical purposes drjdng oils should have an iodine number of at 

 least 140, preferablj^ 170 or higher, and nondrjdng oils of 90 or imder. 

 For purposes of classification Ilolde limits semidrying oils to a range of 

 95 to 130. Certain fish oils have a high iodine number and will absorb 

 oxj'gen, but they do not dry to a hard layer. The fat of marine animals 

 has a higher iodine number than that of terrestrial animals. 



Calculated Data from the Iodine Number. 

 Theoretically, the unsaturated fatty acids belonging to the oleic, chaul- 

 moogric and ricinoleic series absorb 2 atoms of the halogen; linohc and 

 tariric series, 4 atoms; linolenic series, 6 atoms; clupanodonic series, 8 

 atoms, etc. The members of the chauhnoogric series are cycUc compoimds, 

 and contain only 1 pair of double-Unked carbon atoms, while open chain 

 acids of the same empirical formula would in most instances contain 2 

 pairs. Tariric acid contains triple-bond carbon atoms, and although it 



' Lewkowitsch, Analyst, 24, p. 259 (1899). 



= Chem. Rev. Fett u. Harz Indus., 1899, p. 1. 



' Lewkowitsch, Analysis of Oils, Fats and Waxes, 1, p. 400 (1913). 



« Holde-Muellor, Examination of Hydrocarbon Oils, pp. 297-298 (1915). 



